Disability and Rehabilitation, 2008; 30(11): 871 – 883

RESEARCH PAPER

The experience of men using an upper limb prosthesis following

amputation: Positive coping and minimizing feeling different

ADAM SARADJIAN1, ANDREW R. THOMPSON2 & DIPAK DATTA1

1
Northern General Hospital, Sheffield, and 2University of Sheffield, UK
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

Accepted May 2007

Abstract
Purpose. Psychosocial factors are likely to play a crucial role in adjustment to upper limb amputation and prosthesis use,
and yet have received only minimal exploration within the literature. This study therefore, sought to gain a rich
understanding of the experience of living with an upper limb amputation and of using a prosthetic arm and hand.
Methods. The qualitative method of Interpretive Phenomenological Analysis was used. Purposive sampling culminated in a
homogenous sample of 11 males with unilateral upper limb amputations, who wore a prosthesis at least weekly. Semi-
structured interviews were carried out, transcribed and analysed according to the methodology.
For personal use only.

Results. Participants identified a theme of ongoing awareness of difference in appearance and ability. Consequently,
participants described themes of psychosocial and functional adjustment to minimize this sense of difference. This was
facilitated by the participants’ prostheses and their positive coping style. Within this, participants also identified the personal
meanings of their prosthesis and highlighted the terms of its use. The minimization of their sense of difference resulted in
participants regaining a sense of worth.
Conclusions. The findings offer a greater psychological insight into adjustment from an upper limb amputation and the role
of prostheses. These findings have implications for both the clinical rehabilitation of patients who undergo upper limb
amputations, as well as for future research into the use and value of prostheses in facilitating the adjustment to this
experience.

Keywords: Amputation, upper limb prosthesis, coping, interpretative phenomenological analysis

and anger have been found amongst people with

Introduction
amputations [8].
Amputation of a limb has an extensive effect on Losing a limb has been found to dramatically
people’s lives, with people losing many physical change a person’s sense of body-image and conse-
functions and abilities that were once taken for quently self-image, which has, in turn, been asso-
granted [1]. It has been found to impact upon ciated with a person’s satisfaction with life [9].
people’s occupational status, leisure pursuits, social Related to this, social discomfort, perceived social
contacts and activities of daily living [2 – 4]. In stigma and public self-consciousness have been
addition, people commonly experience stump pain found to independently predict depressed affect in
and sensation following amputation and also phan- people with amputations [10 – 12]. In addition, the
tom pain and sensation of the absent limb [5,6]. The amputation of a limb and the use of a prosthesis
emotional reaction of the loss of a limb has been can change a person’s social life and the nature of
likened to the grief process of losing a spouse [7]. social interactions [1,13]. There is some evidence
Parkes describes the person who has lost a limb to that after lower limb amputation, men are more
initially experience shock, denial and numbness, concerned with the restoration of function and
then anxiety and depression before they are able to women with regaining a ‘feminine’ body image
‘adjust’. Indeed clinical levels of depression, anxiety, [14,15].

Correspondence: Andrew R. Thompson, Clinical Psychology Unit, Department of Psychology, University of Sheffield, Western Bank, Sheffield, S10 2TP, UK.
Tel: þ44 (0)114 2226570. Fax: þ44 (0)114 2226610. E-mail: [Link]@[Link]
ISSN 0963-8288 print/ISSN 1464-5165 online ª 2008 Informa UK Ltd.
DOI: 10.1080/09638280701427386
 872 A. Saradjian et al.

There are many factors that have been investigated the person more than any technical features of the
in moderating a person’s psychosocial adjustment to prosthesis itself [26,27].
losing a limb including patient demographics such as Murray [28] has argued that the concept of body-
age, gender and level of education and amputation image in relation to people with prostheses should not
related variables such as the level, cause and time only concern the affective response but also incorpo-
since amputation, and some psychological factors rate the concept of embodiment. Prosthetic embodi-
such as coping style but few consistent findings have ment [29] can involve a functional and emotional
been established [16 – 18]. The impact upon the identification with prostheses to varying degrees,
person and the process of adjustment to limb whereby the boundaries of the ‘self’ become ambig-
amputation is a highly complex and dynamic one uous [30]. Murray [28] delineated six themes in the
that varies across individuals [1,19]. perceptually embodied experience of upper and lower
Upper limb amputation (ULA) in the United King- prosthesis users, particularly noting that two main
dom is considerably less common than lower limb form could be differentiated; those who had a strong
amputation (LLA). Approximately eight times more emotional connection to it, experiencing the pros-
men than women undergo ULA, and it is more thesis as part of their body and those who merely
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

common amongst adult men of working age [19]. viewed it as a tool and a means to an end.
Successful rehabilitation is important economically in There is a dearth of psychological research on
facilitating return to active employment. The loss of an people with ULA. Little is understood about the
upper limb potentially has a greater impact than LLA experience of having one, how it affects a person’s
[19,20]. This is because people’s hands and arms are life, their body-image and their general attitude
not only particularly important functionally, allowing towards it. A psychological understanding of wearing
people to manipulate objects and carry out most of the prostheses is necessary in order to help facilitate
activities of daily living, but socially as well. They are rehabilitation. In addition to the lack of knowledge in
used for expression, communication and affection, with the area, considering the highly subjective and
hands considered to be the second most individual and personalised nature of an individual’s ‘body-images’
For personal use only.

personal part of the human anatomy after the face [31] and the complex and dynamic nature of the
[19,20]. However currently, there is insufficient em- adjustment process, qualitative investigations could
pirical research to support this assertion. yield a richer understanding of the issues. Such
The aim of a prosthesis is to restore body image and methodology also allows scope to include positive
improve functioning in a cosmetically acceptable way meanings, adaptations and coping mechanisms that
[19]. The fitting of an artificial limb is an integral part are made in response to such a traumatic life event
of such patients’ rehabilitation. Implicit in this is the [32,33]. As a result of these factors, a qualitative
expectation that the prosthesis facilitates the adjust- approach to obtain an ‘insider perspective’ has been
ment process. Along with the potentially greater advocated by a number of authors in the area of pros-
functional and personal/social importance of the theses research [16,34]. As no such research has been
hand and arm, comes a greater discrepancy between conducted, the present study aims to redress this,
the person’s wishes and what the prosthesis can offer through exploring the personal meanings of wearing a
[19,20]. Narang and Jape [21] have found that upper ULP in people who have experienced a ULA.
limb prostheses (ULP) are functionally and psycho-
logically less satisfying than lower limb prostheses.
Upper limb prosthesis usage rates vary between 37 Method
and 83% [22,23]. While investigations of satisfaction
Study design
with the prostheses and their rate of use have been
limited, the little research on this topic has found Interpretative Phenomenological Analysis [35] (IPA)
only a moderate association between satisfaction and was used to obtain an ‘insider perspective’ of wearing
use [14,24] and no correlation between satisfaction a ULP. Such a method was deemed the most suitable
and social discomfort [11]. Biddiss and Chau [25] because it is concerned with trying to capture the
recent review suggests that personal and contextual experience of individuals who share particular
factors play a critical role in prosthesis acceptance. characteristics within the context of the world in
As Murray and Fox [14] point out, the relationship which they live. It has been argued that IPA is
between a person and their prosthesis tends to be particularly suited to research within health psy-
narrowly considered in terms of use, acceptance and chology [36]. This is predominantly due to health
rejection rates with little understanding about the psychology being based within the social cognition
relationship between people and their prosthesis. paradigm, which IPA also shares. Namely, IPA
Such an understanding is all the more important assumes that people think about their bodies (and
considering that prosthesis use and acceptance is a the impact of illness) and that these thoughts may
complex process that lies within the psychology of be communicated. Importantly however, IPA
 Experience of men using an upper limb prosthesis following amputation 873

acknowledges the impossibility of gaining direct centre and were asked non-directive and open-ended
access to an individual’s experience, relying not only questions concerning what it was like to wear a pros-
on the individual’s ability to communicate it, but also thesis, how they used it and what it meant to them.
the researcher’s own interpretation of this experience They were also asked about their social lives and how
and the interaction between participant and re- they saw the future. More specific follow-up ques-
searcher. Thus within IPA, the themes identified tions were asked of participants depending on their
are the product of these factors and only one of a preliminary responses, encouraging them to elaborate
limitless number of possible representations. on their initial account. Finally, participants were
While IPA cannot be used to develop a general- invited to discuss any personally meaningful area of
izable theory of adjustment to amputation per se, their experience that had not already been covered.
through exploring patients’ personal meanings of The duration of interviews varied, but generally
ULP, theory and clinical practice in areas such as lasted approximately 45 minutes. Interviews were
adjustment to amputation, prosthesis use and the audio-taped and transcribed verbatim. The inter-
relationship of prosthesis to body-image and the self- viewer was a white, 27-year-old male with no
concept can be informed. disabilities or disfigurements.
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

Participants Data analysis

Characteristic of IPA, participants were selected Analyses of the transcripts were conducted according
through purposive sampling, the inclusion criteria to established IPA guidelines [37]. Transcriptions
for which were adult males who had undergone were analysed consecutively after completion of all
unilateral ULA at least 2 years previously and who the interviews in order to minimize bias. Analysis of a
wore their prosthesis at least weekly. Participants transcript involved first becoming immersed within
were excluded if they had a concurrent chronic illness the data through reading and re-reading the text and
or mental health difficulties and if they were unable to recording any initial reflections and observations
For personal use only.

speak English. Such a sample was selected due to before initial conceptual themes were identified and
ULA predominantly affecting the male population of labelled. After which the analysis took on a more
working age. Half (n ¼ 80) of the total population structured process, through identifying relationships
meeting inclusion criteria of the participating NHS between the emergent themes and organizing them
rehabilitation centre in Sheffield, UK were randomly into clusters and hierarchies. Finally, a summary
selected and sent an information sheet and consent table of the structured themes and their relationships
form. This culminated in a sample of 11 participants, to one another were produced. The master themes of
whose demographics and amputation characteristics each participant were then read together and those
can be seen in Table I. Pseudonyms have been used ‘rich’ in evidence and considered to form a coherent
to protect participants’ anonymity. representation of the group’s experience were com-
bined to produce the final thematic analysis.
Interview procedures
Quality control
A semi-structured interview schedule was developed
in line with an IPA approach [33]. Participants were As noted, IPA acknowledges the influence of the
interviewed at their homes or at the rehabilitation researcher and the interaction with a participant.

Table I. Participants’ demographics and amputation related characteristics.

Current Current Time since

Participant relationship employment amputation Cause of Amputation of
(pseudonym) Current age status status (years) amputation Level of amputation dominant hand

John 54 Married Retired 35 Industrial accident Below-elbow Dominant

Peter 45 Married Employed 26 Road traffic accident Shoulder disarticulation Non-dominant
William 55 Married Employed 28 Industrial accident Below-elbow Non-dominant
Simon 49 Co-habiting Employed 25 Industrial accident Below-elbow Dominant
James 64 Married Employed 39 Industrial accident Below-elbow Non-dominant
Terry 31 Single Employed 7 Fireworks accident Wrist disarticulation Non-dominant
Jess 54 Married Employed 28 Industrial accident Above-elbow Non-dominant
Dick 45 Married Employed 19 Industrial accident Above-elbow Dominant
Donald 55 Married Employed 48 Road traffic accident Above-elbow Dominant
Stan 55 Married Employed 28 Industrial accident Above-elbow Dominant
Dean 61 Married Retired 31 Industrial accident Below-elbow Dominant
 874 A. Saradjian et al.

However, steps are still taken to ensure reliability and Although there is also evidence that over time
trustworthiness in the research process. The general there is a process of desensitization to this:
guidelines for qualitative research proposed by
Elliott, Fischer and Rennie [38] were employed. Peter: I think in early years, there is a little bit of an
During the interview, a reflective listening manner embarrassment factor about it . . . but I think as the years
was adopted, thereby clarifying a shared under- go on you get used to it, you just don’t particularly
standing of participants’ experience and thus estab- notice it so much . . . I would imagine it still happens.
lishing data credibility [39]. In addition, a further
personal meeting or phone call was arranged in Despite this, for many there remains a significant
which participants’ individual themes were presented awareness of difference which often gives rise to a
to them with the purpose being for them to provide sense of body shame and self-consciousness:
feedback on the credibility of the analyses. All
participants’ confirmed that their individual thematic Jess: You’re very conscious of the fact that . . . I mean, me
analysis was an accurate summary and representation and my wife don’t find it off-putting but I think other
of their experience. A research diary was kept people would. I won’t even wear a short-sleeved shirt
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

including how the themes were identified and how because I’ve only got one arm showing and I’m
they may relate to each other, thereby establishing an obviously disabled.
‘audit trail’ that was shared with the second
author, who audited the analysis. Imperative in The change in ability left some participants with
qualitative results, participants’ quotes are presented the perceived need to prove their worth, particularly
so that the reader can appraise the fit between the in occupational settings:
data and the researchers interpretation and allow the
reader to consider alternative meanings and under- Dick: Getting back into work was quite difficult . . . trying
standings. to be accepted and prove yourself. ’Cos, when I went to
college for re-training, there was lads there saying, ‘Oh,
For personal use only.

nobody’s going to employ us. We’ve only got one arm’.

Results
The master themes and sub-themes that emerged Participants’ expressed strong feelings about how
from participants’ accounts are illustrated in Figure 1. not only are they aware of a difference from the ‘nor-
Participants recounted their experience in terms mal’ population, but also from the ‘disabled’ popula-
of adjustment from undergoing an amputation and tion, further excluding them from others in society:
having a prosthesis and accordingly the emergent
themes and their possible inter-relationships are Donald: I suppose that I live this sort of twilight
representative of this. The particular quotations existence, you know. I was disabled to able-bodied
used in this section were chosen as they were people but to disabled people, I wasn’t because I wasn’t
considered to best express the theme being dis- in a wheelchair.
cussed or demonstrated a significant nuance within a
theme. As a result of this awareness in physical difference
in appearance and ability, participants described a
process of psychosocial adjustment and functional
Impact of amputation: Awareness of physical
adaptation. The emphasis on adjustment in these
difference and ability
respective areas varied between participants depend-
The predominant theme concerning the impact of ing on what was of greater personal value to them.
the amputation was a change to participants’ self-
image as they became aware of a physical differ-
Psychosocial adjustment
ence in appearance to others and a change in their
‘ability’. This change in self-image was enmeshed As noted, for many an awareness of difference
within their social experience. In the early stages, remains, but participants identified a process of
participants reported greater emotional sensi- adjustment in their social interactions, activities and
tivity and greater sense of awareness prompted roles that minimized this difference and the negative
by a tendency towards hyper-vigilance to others feelings associated with it.
reactions:
Social interactions
Simon: You think you’re the only one that’s ever had a
hand off . . . people looking at you, you’re actually so An awareness of difference is ongoing because of the
conscious of that, I’d be watching for people looking at reactions of others encountered on a daily basis.
me. However participants’ prostheses offer some respite
 Experience of men using an upper limb prosthesis following amputation 875
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14
For personal use only.

Figure 1. Summary of the emergent themes and their possible interrelationships.

to this as they play an integral role in moderating expressed through staring but also through actually
others’ reactions through helping to conceal their asking participants about their amputated arm/
difference and thus facilitating a ‘normal’ appearance prosthesis. Most participants managed this curiosity
(discussed later). This can be very effective as by responding openly and honestly. This appeared a
participants remarked upon others’ surprise when highly effective strategy because once others’ curios-
they finally noticed: ity was satisfied, this allowed the interaction to
continue with attention no longer focused on the
John: A lot of people say to me . . . oh I didn’t know you amputated arm and/or prosthesis:
had an artificial limb, I never noticed it . . . . And I might
have known him for ages.
Jess: I just wait for it. Sooner or later . . . they’re going to
come and ask – what have I done? And you just go back
In meeting new people, others’ reactions were to the same story and tell them and there’s nought else
most commonly of curiosity, which was often said about it.
 876 A. Saradjian et al.

As this also illustrates, such an approach involves

Acceptance and support of others
accepting it is likely to happen:
Of fundamental importance to participants was the
Jess: I used to shy away at first. I’d try and keep it away acceptance and support of others, particularly their
from people, out of sight. But then I thought, well, family and friends:
people are going to ask anyway so let’s go for it. If they
ask, I’ll just answer them. Simon: They [family] all rallied round me. I’ve got a
sister . . . and she looked after me and fortunately I lived
This acceptance may also be contributed to by the across the road from my mum and dad so I got quite a
attainment of a more objective perspective of others’ bit of help from them.
curiosity, through not taking it personally:
William: People outside them [family], them that know
me, they just class me as the same. When I go across the
Jess: It’s human nature . . . they wouldn’t be normal if road to the pub, I’m just William from across the road,
they didn’t ask. full stop.
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

A further cognitive component of managing Peter: I think it’s helped being in a settled relationship as
others’ reactions is through making external attri- well. I think that’s been a contributing factor.
butions. While participants described very few
negative or demeaning reactions of others, to help
Adjustment in leisure activities and occupational
maintain self-esteem participants attributed any
roles
such comments or attitudes as the responsibility of
others: Most participants had to make some adjustments to
their leisure activities and occupation as a result of
Dean: If people want to have a dig, I’m not bothered . . . . their amputation. The majority returned to their
I think that they’re narrow-minded. place of work but due to many working in manual
For personal use only.

occupations, the roles that they adopted changed.

However there appears a fine line between what is Some participants altered the leisure activities that
perceived as an insult and what is acceptable. This they took part in, some due to physical limitations
seems to depend upon the relationship between and some due to a sense of body shame:
people:
Stan: It stopped me playing golf. I did try playing golf
Simon: I mean people used to say, oh, stumpy, how are with my prosthesis on but I found that if I played 18
you? Oh, it was a bit of kid, they were mates, it didn’t holes I would suffer for three days with marks across my
matter. But if somebody I didn’t know said it, oof, red back.
rag to a bull!
John: I won’t go swimming . . . my wife keeps telling me
Indeed, humour appeared an important charac- not to be so conscious but I can’t help it.
teristic of social interactions with people partici-
pants had established relationships with, particularly However, generally participants reported very little
friends. It seemed to be considered a normal part of restriction in their leisure activities, with all engaging
friendly banter and was thus socially inclusive rather in a wide range that were personally meaningful for
than excluding: them.

Simon: Lads used to take my arm off me and hit me with Cognitive adjustment
it, because I used to hit them, I’d whack them on the
knee, you know like, devilment, isn’t it? In addition to the cognitive components associated
particularly in assisting social interaction, downward
Terry: Now they’re [friends] careful when they turn social comparison was also evident amongst partici-
round and say, ‘oh Terry, just give us a hand’, because pants in helping to gain a sense of perspective of the
now I just unclip my hand and give it them. It’s like, oh impact of ULA on their lives and in the process
yeah, funny bugger, just because you can do it.
bolstering their self-worth:

Humour seems to facilitate social adjust- Jess: They sent me to a plastics hospital . . . when I went
ment through a process of normalization to the in there I was feeling pretty down and sorry for myself
prosthesis, desensitizing people to it making it and I saw these kids in there that were having operations
less threatening and in the process, others more and I thought, they were a lot worse than me, so what
comfortable. am I moaning about? And that did me a power of good,
 Experience of men using an upper limb prosthesis following amputation 877

it gave me a kick and I just thought after that, just get on Dean: Well it gives you self-confidence . . . you don’t
with your life. have to rely on other people, that’s the main crux of the
thing.

Functional adaptation
It was evident that participants considered functional Role of prosthesis and terms of use
adaptation as a process of re-learning:
Adjustment to a prosthesis and integration to self
Dean: People think . . . oh, it’s gone. But you’ve just got Participants described the process of adjusting to a
to be, like, life doesn’t end, it’s a new beginning . . . it’s prosthesis in terms of wearing it and using it. In
like a baby, you go through all the stages. You’ve got to adjustment to using a prosthesis, participants de-
learn different things.
scribed varying degrees of ease in doing so but one
which understandably was more challenging at the
Compensation
beginning:
The lost function of the amputated arm was
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

compensated for to varying degrees between partici- Dick: It seemed a bit strange at the time, when you go in
pants by an increased use of the intact arm, the use of and have one fitted and you think, ‘This is going to be
technology to assist in completing tasks and the pros- clumsy’ but you soon adapt to it, you soon realize that
thesis (the role of the latter will be discussed later): you need it. So you find it hard to start with but it all
works out in the end.
William: I just use my right hand all the time. I can get
on quicker by doing things and I’m adapted. Participants demonstrated how adjusting to using
a prosthesis is a very personal and individual process.
James: I would say in this modern day, a lot of power This often involves using it in ways and purposes it
tools do a lot, and make it a lot easier. was not necessarily intended for:
For personal use only.

Participants exhibited the importance of cogni- William: I use it as a tool but I don’t use the tools that
tive aspects in successful functional adjustment. This was meant for it . . . it’s quite handy for knocking poles
often involved a cognitive re-evaluation of functional into sand but you’ve got to go careful that you don’t
abilities acknowledging the loss and in so doing break it . . . it comes in quite handy when you’ve got a
adjusting expectations: wasp in your car because you can take it off and swat
it . . . you don’t get stung or anything . . . I don’t think I
Terry: I don’t do as well as I used to do, but I suppose for use it as I should do because I adapt.
somebody with one hand, yeah, I do well . . . I might not
be as quick as you perhaps but it gets done. It also involves using it in a way that is optimal to
the individual rather than what the potential of the
Another important cognitive facet involved prosthesis is, but that the prosthesis extends the
problem-solving abilities in facilitating the comple- options open to its user:
tion of tasks:
Stan: I’m not using it to its full capability I know, but I
James: It just takes a little bit more thought . . . even little am using it to mine . . . for what I want . . . I love driving
things, you have to think a lot more about them before and it allows me to still be able to drive because it allows
you set about things. me the freedom of not having to have an automatic car if
I don’t want.

Self-efficacy and independence

Thus the use of a prosthesis can be conceptualized
The importance of functional ability was evident as extending the potential of the individual and the
for many participants, seemingly being quite practi- options open to them who then adjust to using it in a
cally orientated people with a history of working in way that fulfils their needs.
manual occupations. As a result many appeared to In adjusting to wearing a prosthesis, participants
derive much of their self-worth from their self- described how this often involves individual ways of
efficacy and independence: making it more comfortable to wear:

Terry: It’s just that I like to be independent . . . that’s the John: I had this idea of putting a piece of string from my
kind of thing I like to put across that I am capable and belt to the underarm strap which kept it away from
ninety nine percent of the time I am able to function on under my arm and that helped . . . and I still use it to this
my own. day.
 878 A. Saradjian et al.

It also involves adapting to the different bodily to rub against participants’ skin making it sore. This
experience of wearing a prosthesis in terms of often prohibited prosthesis use during the summer
altering posture and learning how to ‘carry’ it: and on holidays, not only due to discomfort but also
because appropriate seasonal clothes undermine the
Jess: I got used to it, the weight of carrying it, sticking cosmetic utility of the prosthesis:
your shoulder up in the air like that . . . just carry it as
normal as I can. Stan: I don’t usually wear short sleeved shirts with my
arm, so I usually take it [prosthesis] off during the
As this participant suggests ‘successful’ adjustment summer.
to wearing a prosthesis could be equated to, as well
as facilitating a ‘normal’ appearance: For those who were socially dependent on their
prosthesis, they continued to wear it despite the
Jess: I know some guys, they’ve got an artificial arm and discomfort and other problems this causes:
they look terrible. The way they carry it, you can just tell.
Whereas generally, people don’t notice that I’ve got an Jess: It can be boiling hot in the summer, I still put it on
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

artificial limb. and it’s very uncomfortable. In fact, in the summer, I

send it back more times for repair. . . . Because it’s a
shiny leather lined socket and it gets soaked with sweat,
For most of the participants in this study adjust-
so does the harness that holds it on, and it smells.
ment to wearing a prosthesis culminated in it
Basically if I can smell it, other people can.
becoming integrated to their body image, which
may be conceptualized as meaning that the prosthe-
Ongoing support from services
sis is predominantly no longer in the person’s
conscious awareness: Despite the problems related to prostheses, partici-
pants described ongoing support from services which
Dean: . . . over the years you just get used to them. It’s met their prostheses related needs and thus mini-
For personal use only.

just like a part of your skin. mizing the impact:

William: I’ve never thought of it. I just think that it’s my Peter: The limb centre is brilliant anyway. They are really
arm. supportive. If ever you need anything, you only need to
pick up the phone and they will do anything they can for
you.
For many participants the prosthesis becomes part
of them like an item of clothing:
Facilitative role of prosthesis
Donald: I suppose it’s like wearing glasses or what have The role of the prosthesis varied in terms of provi-
you. You know, you choose a pair that are attractive that ding appearance or function for individuals depend-
you think enhance rather then detract from your image ing on what was of more personal importance to
and that you’re comfortable with.
them, but for most it fulfilled a role in both areas. As
noted earlier, the prosthesis plays an important role
Dick: I don’t think about it. Just go to work, put your arm
on and do your work and that’s it . . . cos really it’s like in facilitating a ‘normal’ appearance:
somebody going to work and putting a cap on . . . you
never think about it at work, it’s your other arm. Stan: I think from cosmetic point of view, I would much
rather wear my arm and look normal. I think it helps
Like an item of clothing, even participants who with how you look and how you feel. I don’t want people
to be aware that I’m anything other than me, you know I
found the prosthesis comfortable to wear would take
don’t want people to be sympathetic towards me or
it off when at home and relaxing, when no longer
anything like that, I’m wanting people to acknowledge
serving a use: me for who I am. I want to look to all intents and
purposes no different to them. And I think it [prosthesis]
Simon: I’m not saying this [prosthesis] is uncomfortable helps me, erm, to a degree.
but without it on, I’m more comfortable. I’m relaxed
[at home] and I sit and have a beer and watch telly. As well as serving a role for themselves, this
There’s no need for it to be on if I needed it on, I’d have
‘normal’ appearance is also perceived as important in
it on.
meeting perceived social demands, usually in more
formal settings:
Practical problems and limits of use
The most common problems related to wearing a Peter: These days the only time I feel I need to wear it is
prosthesis were those caused by hot weather which to go out anywhere. I feel it’s important ’cos people
caused sweating and the straps holding the prosthesis don’t want me on their wedding photographs with, well I
 Experience of men using an upper limb prosthesis following amputation 879

don’t think so anyway, with your jacket sleeve tucked percent. Even though I’m disabled, I’ve never classed
in your pocket. It looks a lot better with your prosthesis myself as disabled . . . because if I didn’t have it
on. [prosthesis] then I would feel disabled because I
wouldn’t be able to do or physically look . . .
As this participant goes on to suggest, the under-
lying reason being that: This ‘normal’ identity the prosthesis helps par-
ticipants assume illustrates how providing form,
Peter: I think that everybody wants to fit in. I think that’s appearance and facilitating function has a more fun-
a big thing. damental impact than these elements in isolation.
Essentially they have a profound impact upon a
Participants also alluded to the prosthesis provid- person’s confidence and self-esteem:
ing a more ‘intact’ body-image by creating a sense of
wholeness: James: When I go out I’m like a normal person, because
I’ve got two hands. I just don’t like going out without it. I
James: I feel more whole when wearing my limb. mean there might be like people seeing me, even though I
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

Because when I’ve got that on, I’ve two hands, haven’t I? would say ninety percent round about know that I only
got one (arm). But I feel better because I think I’ve got
two. I feel better with it on. Confidence, isn’t it really?
The facilitative functional role the prosthesis
fulfilled for participants was important in a variety
Evaluation of prosthesis
of ways, in completing the activities of daily living
and fulfilling an occupational role: Participants explicitly evaluated their prosthesis,
although this was not an isolated process. Satisfac-
Terry: It allows me to do more than what I would be tion with it varied within an individual participant
able to do if I didn’t have a prosthesis, which is depending on what they were comparing it to. In
basically why I’ve got one . . . without it, life would be a comparisons with the lost hand, as expected the
For personal use only.

hell of a lot harder . . . just even day-to-day life would be prosthesis was perceived less favourably:
trouble.
John: Although with this one [myoelectric prosthesis] I
Dick: I couldn’t go to work and not put it on ’cos I can open the thumb and finger, I haven’t got the full
wouldn’t be able to do my job. potential of a hand . . . I mean there’s quite a difference
between the prosthesis and my hand because I’ve got
The prosthesis also facilitated participants to quite veiny hands.
continue recreational and leisure activities. Driving
was a highly valued activity for many participants, for ULP were also evaluated poorly in comparison
leisure and for their occupation with many also with LLP:
stressing the importance of the prosthesis facilitating
their use of a ‘normal’ car: Peter: They are very antiquated to be quite honest. I
think that a lot of technology has gone into legs, ’cos
William: I drive an ordinary car, five gears, handbrake, people need them to walk about in. I think arms have
everything for a normal, able-bodied person. been left behind a little bit.

The facilitative functional role is also important in While most participants evaluated their prostheses
that it add to the presentation of a ‘normal’ positively as it fulfilled a purpose for them and was
appearance: certainly better than nothing, relative to their
perceived ideal there remained much room for
Terry: A lot of people have turned round and said, improvement:
well, I’ve not noticed, because of the way I go and get
the job done . . . and that’s the kind of thing I like to put Dean: They’re [prostheses] a means to an end . . . they
across. want summat a bit better . . . which is easier to wear,
more adaptable. And maybe something like instead of
It may be considered that prostheses therefore working on batteries, you want something that your
facilitate a ‘normal’ identity for participants, through mind can work on . . . they want something which is
more natural and they can interact with other people and
helping them to regain a sense of not feeling different
they don’t look out of place.
in physical ability or appearance:

John: I suppose it means I’m almost normal again . . . How and when a prosthesis is used, for whatever
without the [artificial] limb I wouldn’t be one hundred purpose is a wholly personal experience. Its use is a
 880 A. Saradjian et al.

continual cost-benefit analyses of the above factors take you a couple of goes to do it, but you’ll find out that
associated with the prosthesis: you can do it.

Stan: It’s just for, for purposes of feeling more One participant also highlighted the importance of
comfortable rather than being uncomfortable. engaging socially with others:

Dean: To me people take you for what you are. It’s like if
Coping style facilitating adjustment
you’re going to go out in a group and you’re not going to
Participants described their attitude to coping with get involved in what’s going on in the conversation . . .
the challenges they faced as a result of their over a period of time, they’re not gonna involve you. But
amputation and to life more generally. This attitude if you get into the conversation and what’s going on,
and approach to life permeated through the other they’ll accept you.
themes and was evident within their psychosocial and
functional adjustment. Participants also demonstrated an engagement
within the treatment process:
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

Positive attitude
John: The sister come round and said, I don’t think
Participants displayed a positive attitude to life and you’ll ever be able to write with your left hand again so
coping with the challenges that confronted them. I’d advise you to start practising, and I did. I got a
This was exhibited through accepting what had writing pad and a pen and learnt to write.
happened and being able to positively reframe
problems and ultimately make the most of their This also often involved an assertion of control
situation: within the process, with participants making their
needs and wants known to the service:
Dean: Disability is what you make of it . . . if you sit in a
For personal use only.

corner and brood over it and don’t get on with life, Peter: [The service] took a little bit of persuading that I
people won’t help you. . . . You can’t turn the clock back would be better off with one of them rather than a plastic
you’ve just got to get on with your life . . . you’ve just got one. Eventually they succumbed.
to make the best of what you’ve got.

Self-worth
For some, time passed before such an attitude was
able to be adopted but for others it was present from Participants reported high self-worth, which was
the start: evident from ‘successful’ functional and psychosocial
adjustment, which for these participants was in terms
James: The doctor came in and he said, we’ve had to of minimizing their sense of difference from others
amputate, I said that’s alright no problem . . . he came and thus being able to fulfil the activities and roles
back in, we’ve had to amputate your arm and I said that were personally meaningful to them. It also
that’s no problem. And then he came back again and meant that the difference did not make them any less
said we’ve had to amputate, I said it’s alright. So then he worthy than anyone else:
came next morning and said we’ve had to amputate it! I
said that’s alright. And he said, well you haven’t took it Dean: Only difference between me and you is that
as people normally would . . . and I said well once it’s [tapping prosthesis]. Nothing else.
gone, it’s no use me moping about it.
Peter: I’m as good as the next man.
Engagement in life and openness to experience
Pride in positive adjustment
This positive attitude was exhibited by participants’
engagement in life, being open to new experiences In fact participants even drew a sense of self-worth
and adopting a ‘can do’ approach. Within this, through a recognition of how ‘positively’ and success-
participants emphasized the importance of patience fully they had adjusted and overcome the challenges
and persistence in this process: they faced:

Jess: I still try and do things even if it goes pear shaped, John: I’ve done really well . . . I carried on, it’s so easy to
so what, I have a go at it. give up . . . . I proved it, haven’t I? I’ve proved what you
can do.
William: I’ve done various things . . . I’ve done archery,
football, crown green bowls, ten pin bowling, darts Simon: I’ve obviously helped myself, I mean it’s easy to
tennis . . . There are ways to do different things. It might crawl into a corner if you want, isn’t it? Or you just get
 Experience of men using an upper limb prosthesis following amputation 881

up and get on with it and I’m not being big-headed, but a wish to appear ‘normal’ to others and this often
I think I did the other; I got up and got on with it. being facilitated by prostheses. As with the present
research, this ‘normal’ appearance also led to social
interactions that included others’ surprise upon
Discussion and conclusion
discovering participants’ prostheses in Gallagher
This study illustrates how for the present group of and MacLachlan’s study. However, their participants
participants the process of adjustment from a ULA described such interactions as awkward, whereas the
involves attempting to regain lost physical form and participants in the present study used such instances
function, thereby minimizing the difference between as a positive indication of appearing ‘normal’.
current and former selves, but also attempting to Beyond this, due to the lack of detailed discussion
minimize a sense of difference from others. Partici- in this area, it is difficult to ascertain other variations
pants were highly invested in portraying a ‘normal’ in the social experience of the different populations.
appearance that served to maintain a ‘normal’ The present study highlighted the social nature of
identity and offset a self-image as disfigured and adjustment for people with ULA and prostheses that
disabled. Prostheses played an integral role in this is consistent with findings for people with LLA,
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

process through facilitating participants’ form, func- which have demonstrated a mediating effect of body
tion and body image, although the emphases of image and social discomfort to a person’s emotional
which varied between individuals. There is a process adjustment [10 – 12]. The findings of the present
of adjustment to prostheses, which also demon- study somewhat support Murray and Fox’s [14]
strated the individuality of a person’s relationship to research with people with LLA, which suggested that
it. For many it became integrated into their body men’s positive body image was more associated with
image. Participants also experienced practical pro- the function of prostheses. While the present study
blems related to their prosthesis and limits to its use, did demonstrate that functional ability facilitating a
which to some extent, could be minimized by the sense of self-efficacy and independence were of
support of services. Use of prostheses seemed to be a fundamental importance to participants, for many,
For personal use only.

continual cost-benefit analysis of the facilitative role so too was the aesthetic appearance.
prostheses fulfilled along with its problems and The dual role of facilitating appearance and
limits. An important cognitive element was found function that prostheses provided participants, along
in the psychosocial and functional adjustment to with their adjustment to it, meant that for many the
ULA, which helped to minimize the sense of prosthesis became integrated into participants’ body
difference. Participants also displayed a positive image. Thus prosthetic embodiment was found to
and accepting attitude, exhibiting an engagement in occur. Both the broad types of prosthetic embodi-
life and an openness and embracement of new ment that Murray [28] differentiated were identified.
experiences within their rehabilitation and life in However, it was apparent that those who appeared to
general. As a result, participants’ lives appeared emotionally identify with their prosthesis could also
minimally restricted, with all describing the fulfil- consider it a tool at other times. Therefore depend-
ment of personally meaningful roles, activities and ing upon the situation people may be able to
relationships. This contributed to the high self-worth transcend different embodied forms. It was clear
found amongst participants coupled with the ‘suc- that for the couple of participants who only appeared
cessful’ minimization of their sense of difference. to consider their prosthesis a tool, it was not
Participants also drew self-worth through recognition necessarily less important to them in the role it
of their positive adjustment. fulfilled as for those who viewed their prosthesis as
The areas of people’s lives which have previously part of their body.
been found to be affected following ULA such as Theoretically, much of participants’ psychosocial
occupational status, social relationships and activities adjustment can be understood from within a
of daily living [2,3,24] were found in our study. biopsychosocial model of body shame [41]. Partici-
Whereas quantitative studies have focused on the pants described ‘external shame’, implicit with which
negative impact on the individual, this study is a perception that they will be devalued, rejected
acknowledges that while such challenges are faced, and excluded as a result of their amputation. The
these can also be overcome and life need not be any prosthesis serves as a ‘defensive behaviour’ of
less fulfilling. concealment to this and in so doing helps to manage
The results highlight some of the salient issues for social interactions. It did this by facilitating a
people with ULA and prostheses, which are com- ‘normal’ appearance that generally allowed partici-
parative with qualitative findings with people adjust- pants to blend into crowds and when meeting new
ing to a LLA [1,40]. Such studies also highlighted people it often ‘bought time’ for them before the
the impact on a person’s self-image and how this was ‘difference’ was noticed. This allowed participants to
socially embedded, with their participants expressing be evaluated by others as people rather than being
 882 A. Saradjian et al.

immediately discredited based on their physical Problem-solving strategies could also be usefully
difference in line with stigma theory [42]. As adopted in helping to facilitate functional adjustment.
acknowledged by participants, the underlying moti- The fact that the interviewer in the study has no
vation of wanting to appear ‘normal’ was the ‘need to physical disabilities or disfigurement may have had an
belong’. In line with Baumeister and Leary’s [43] impact on the findings. It was apparent that partici-
standpoint concerning the importance of belonging- pants discussed their experience assuming the naivety
ness to adjustment, participants identified the value of the researcher. It is important to recognize that
of the acceptance and support of significant others. the themes identified in participants’ experience of
Participants’ positive perception of their social having a ULP following amputation cannot be
relationships and engagement in all aspects of their generalized and assumed to represent the experience
lives, exhibiting very little activity restriction may of other people with ULA or indeed other disabled or
have been particularly important in their positive disfigured populations as the participants formed a
emotional adjustment as Williamson et al. [44] homogenous sample. They were all men of white
found that poor social support and restricted activity ethnicity from an apparently similar socio-economic
are predictive of depressive symptoms. background, of a similar age who had had their ampu-
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

The findings of the present study generally seem to tation and prosthesis for many years. It is likely that
endorse the underlying aims of prostheses as out- people differing in these characteristics would have
lined by Ham and Cotton [19]. While not necessarily some different as well as overlapping experiences.
‘replacing’ the missing part, for participants, pros- Indeed, future similar qualitative studies may
theses helped to restore a person’s body image usefully explore the coping strategies of other
through being functionally useful and cosmetically ‘groups’ of people with upper, lower and combina-
acceptable, even becoming integrated to participants’ tions of amputations; although assumptions should
body image. As this study also shows, how prostheses not be made about such ‘groups’ as the partici-
can help an individual cannot be assumed and varies pants in this study illustrated. As may have been
between people with its utility depending on the expected with middle to older aged men generally
For personal use only.

person using it. from traditional working class, manually orientated

The findings of this study have important clinical backgrounds, great emphasis was placed on their
implications in the psychological therapy and rehabi- functional ability, self-efficacy and consequent in-
litation of patients who have undergone an ULA. It dependence, but for many appearance was also very
reinforced the assertion that rehabilitation is more important. Quantitative studies may be able to test
successful when individuals accept the impact of their hypotheses concerning whether the apparent person-
disability and engage in the process of adapting to the ality characteristics of the participants and the coping
multifaceted changes that disability involves [45]. It strategies elucidated in this study are indeed asso-
also shows the individuality of the rehabilitation ciated with positive adjustment and the alleviation of
process and the use and role of prostheses in the distress caused by ULA. In addition, in guiding
facilitating this. It is therefore important to engage future research on the use and satisfaction of pros-
the individual in this process as much as possible in theses, this study has illustrated that a wider view
order to maximize the potential of prostheses and needs to be taken. For example, the cosmesis of a
optimise rehabilitation. Participants’ also demon- prosthesis is usually evaluated in isolation whereas,
strated how an engagement in their social world perhaps it should be considered in terms of fulfilling
aided their adjustment. Much of this entailed learning a role through its success in allowing the wearer to
to manage social interactions, particularly in meeting conceal their difference. In addition, prosthesis use is
new people. Whilst we would not wish to be seen to usually considered in terms of acceptance or rejec-
endorse a deficit model, and recognize the need for tion but instead future research should be considered
community interventions aimed at changing societal in terms of patient outcomes and not purely on the
reactions, our findings do suggest a value in providing amount of time it is utilized for.
social skills training akin to those used with people
with disfigurements [46]. The significant role of the
cognitive aspects of psychosocial and functional References
adjustment exhibits the potential value of cognitive- 1. Gallagher P, MacLachlan M. Adjustment to an artificial limb:
behavioural therapy in assisting patients’ rehabilita- A qualitative perspective. J Health Psychol 2001;6:85 – 100.
tion. Such an intervention would help encourage 2. Demet K, Martinet N, Francis G, Paysant J, Andre J. Health
patients to recognize and challenge maladaptive related quality of life and related factors in 539 persons with
thoughts about disability and their sense of worth. amputation of upper and lower limb. Disabil Rehabil
2003;25:480 – 486.
In social situations it would help encourage greater 3. Jones LE, Davidson JH. The long-term outcome of upper
perspective taking and facilitate people to make limb amputees treated at a rehabilitation centre in Sydney,
more helpful attributions about others’ behaviour. Australia. Disabil Rehabil 1995;17:437 – 442.
 Experience of men using an upper limb prosthesis following amputation 883

4. Whyte AS, Carroll LJ. A preliminary examination of the 26. Fishman S. Education in prosthetic and orthotics. Prosthet
relationship between employment, pain and disability in an Orthot Int 1977;1:52 – 55.
amputee population. Disabil Rehabil 2002;24:462 – 470. 27. Herberts P, Korner L, Caine K. Rehabilitation of unilateral
5. Jenson TS, Krebs B, Nielsen J, Rasmussen P. Immediate and below-elbow amputees with myoelectric prostheses. Scand J
long-term phantom pain in amputees: Incidence, clinical Rehabil Med 1980;12:123 – 128.
characteristics and relationship to pre-amputation limb pain. 28. Murray CD. An interpretative phenomenological analysis of
Pain 1985;21:267 – 278. the embodiment of artificial limbs. Disabil Rehabil 2004;26:
6. Kooijman CM, Dijkstra PU, Geertzen JH, Elzinga A, van der 963 – 973.
Schans CP. Phantom pain and phantom sensations in upper 29. MacLachlan M. Embodiment: Clinical, critical and cultural
limb amputees: An epidemiological study. Pain 2000;87:33 – perspectives. Milton Keynes: Open University Press; 2004.
41. 30. Gow D, MacLachlan M, Aird C. Reaching with electricity:
7. Parkes CM. Psychosocial transitions: Comparisons between Externally powered prosthetics and embodiment. In:
reactions to loss of limb and loss of a spouse. Br J Psychiat MacLachlan M, Gallagher P, editors. Enabling technologies:
1975;127:204 – 210. Body image and body function. Edinburgh: Churchill
8. Frierson RL, Lippmann SB. Psychiatric consultation for acute Livingstone; 2004. pp 155 – 168.
amputees: Report of a ten-year experience. Psychosomatics 31. Pruzinsky T, Cash TF. Integrative themes in body-image
1987;28:183 – 189. development, deviance and change. In: Cash TF, Pruzinsky
9. Breakey JW. Body image: The lower limb amputee. J Prosthet T, editors. Body-images: Development, deviance and change.
Disabil Rehabil Downloaded from [Link] by Northeastern University on 11/05/14

Orthot 1997;9:58 – 66. New York: Guildford Press; 1990. pp 337 – 349.
10. Rybarczyk B, Nyenhuis DL, Nicholas JJ, Cash SM, Kaiser J. 32. Dunn DS. Positive meaning and illusion following disability:
Body image, perceived social stigma and the prediction of Reality negotiation, normative interpretation and value
psychosocial adjustment to leg amputation. Rehabil Psychol change. J Soc Behav Personality 1994;9:123 – 138.
1995;40:95 – 106. 33. Gallagher P, MacLachlan M. Positive meaning in amputation
11. Rybarczyk B, Nyenhuis DL, Nicholas JJ, Schulz R, Alioto RJ, and thoughts about the amputated limb. Prosthet Orthot Int
Blair C. Social discomfort and depression in a sample of 2000;24:196 – 204.
adults with leg amputations. Arch Phys Med Rehabil 34. Desmond D, MacLachlan M. Psychological issues in pros-
1992;73:1169 – 1173. thetic and orthotic practice: A 25-year review of psychology in
12. Williamson GM. Restriction of normal activities among older Prosthetics and Orthotics International. Prosthet Orthot Int
adult amputees: The role of public self-consciousness. J Clin 2002;26:182 – 188.
Geropsychol 1995;1:229 – 242. 35. Smith JA. Semi-structured interviewing and qualitative
For personal use only.

13. Burger H, Marincek C. The lifestyle of young persons after analysis. In: Smith JA, Harre R, Van Langenhove L, editors.
lower limb amputation caused by injury. Prosthet Orthot Int Rethinking methods in psychology. London: Sage; 1995. pp
1997;21:35 – 39. 9 – 26.
14. Murray CD, Fox J. Body image and prosthesis satisfaction 36. Smith JA. Beyond the divide between cognition and dis-
in the lower limb amputee. Disabil Rehabil 2002;24:925 – course: Using interpretative phenomenological analysis in
931. health psychology. Psychol Health 1996;11:261 – 271.
15. Parkes CM, Napier MM. Psychiatric sequellae of amputation. 37. Smith JA, Jarman M, Osborn M. Doing interpretative
Br J Hosp Med 1970;7:610 – 614. phenomenological analysis. In: Murray M, Chamberlain K,
16. Rybarczyk B, Nicholas JJ, Nyenhuis DL. Coping with a leg editors. Qualitative health psychology: Theories and methods.
amputation: Integrating research and clinical practice. Rehabil London: Sage; 1999. pp 218 – 240.
Psychol 1997;42:241 – 256. 38. Elliott R, Fischer C, Rennie D. Evolving guidelines for
17. Horgan O, MacLachlan M. Psychosocial adjustment to lower publication of qualitative research studies in psychology and
limb amputation: A review. Disabil Rehabil 2004;26:837 – related fields. Br J Clin Psychol 1999;38:215 – 229.
850. 39. Streubert HJ. Evaluating the qualitative research report. In:
18. Desmond DM. Coping, affective distress, and psychosocial LoBiondo-Wood G, Haber J, editors. Nursing research:
adjustment among people with traumatic upper limb amputa- Methods, critical appraisal and utilisation. StLouis: Mosby;
tions. J Psychosom Res 2007;15 – 21. 1994. pp 481 – 489.
19. Ham R, Cotton L. Limb amputation: From aetiology to 40. Sjödahl C, Gard G, Jarnlo GB. Coping after trans-femoral
rehabilitation. London: Chapman & Hall; 1991. amputation due to trauma or tumour – a phenomenological
20. Baumgartner RF. Upper extremity amputation and prosthe- approach. Disabil Rehabil 2004;26:851 – 861.
tics. J Rehabil Res Develop 2001;38:vii – x. 41. Kent G, Thompson AR. Models of disfigurement: Implica-
21. Narang IC, Jape VS. Retrospective study of 14,400 civilian tions for treatment. In: Gilbert P, Miles J, editors. Body
disabled treated over 25 years at the Artificial Limb Centre. shame: Conceptualisation, research and treatment. Hove:
Prosthet Orthot Int 1982;6:10 – 16. Brunner-Routledge; 2002. pp 106 – 116.
22. Alison A, Mackinnon SE. Evaluation of digital protheses. J 42. Goffman E. Stigma: Notes on the management of spoiled
Hand Surg 1992;17A:923 – 926. identity. New Jersey: Prentice-Hall; 1963.
23. Pinzar MS, Angelats J, Light TR, Izuierdo R, Pluth T, 43. Baumeister RF, Leary MR. The need to belong: Desire for
Maywood IL. Functional outcome following traumatic upper interpersonal attachment as a fundamental human motivation.
limb amputation and prosthetic limb fitting. J Hand Surg Psycholog Bull 1995;117:497 – 529.
1994;19A:836 – 839. 44. Williamson GM, Schulz R, Bridges MW, Behan AM. Social
24. Davidson JH. A survey of the satisfaction of upper limb and psychological factors in adjustment to limb amputation.
amputees with their prostheses, their lifestyle and their J Soc Behav Personality 1994;9:249 – 268.
abilities. J Hand Ther 2002;15:62 – 70. 45. Marinelli RP, Dell Orto AE. The psychological and social
25. Biddiss E, Chau T. The roles of predisposing characteristics, impact of physical disability. New York: Springer; 1984.
established need, and enabling resources on upper extremity 46. Robinson E, Rumsey N, Partridge J. An evaluation of the
prosthesis use and abandonment. Disabil Rehabil: Assist impact of social interaction skills training for facially disfigured
Technol 2007;2:71 – 84. people. Br J Plast Surg 1996;49:281 – 289.
bs_bs_banner

Geriatr Gerontol Int 2013; 13: 264–273

REVIEW ARTICLE

Rehabilitation of the older vascular amputee: A review of

the literature
Aisling M Fleury,1 Salih A Salih2 and Nancye M Peel2
1
Geriatric and Rehabilitation Unit, Princess Alexandra Hospital and 2Centre for Research in Geriatric Medicine, The University of
Queensland, Brisbane, Queensland, Australia

The aim was to review the literature on factors affecting prosthetic rehabilitation of older vascular lower extremity
amputees. A search of the literature was carried out using the MEDLINE, EMBASE and CINAHL databases up to
September 2010. Only original English language articles were accepted. Relevant studies were reviewed and selected
based on prespecified inclusion criteria namely: age more than 65 years; vascular cause for the amputation; ampu-
tation level of below knee, through-the-knee or above knee; and discussed factors affecting rehabilitation. Age alone
is not an absolute contraindication to prosthetic limb prescription; however, it does influence the potential success of
gait retraining. Other factors influencing prosthetic fitting and use include comorbidities, premorbid function, level
of amputation, status of the remaining limb and patient motivation. Prosthetic gait retraining is not possible in every
older dysvascular amputee; however, almost all amputees will benefit from a rehabilitation program to increase
independence in transfers and learn wheelchair skills. The MESH key words were: “aged” “diabetes complications”
“peripheral vascular diseases” “amputees” “amputation” “rehabilitation” and “artificial limbs”. Geriatr Gerontol Int
2013; 13: 264–273.

Keywords: older person, prosthesis, rehabilitation, successful user, vascular amputee.

Introduction successful prosthetic users.1,2,12–16 Therefore, for older

amputees, an alternative to prosthetic mobility needs to
A total of 75% of all lower extremity amputations occur be considered.
in persons over the age of 65 years, with vascular disease The aim of the present study was to review the litera-
being a significant cause for 90% of such amputations ture on rehabilitation of older dysvascular amputees to
in older adults.1 The proportion of amputees aged 85 understand the factors in this cohort affecting suitability
years or older is projected to increase from 20% to 35% for prosthetic rehabilitation. In the light of these find-
in the next 40 years.2 A Swiss study calculated an inci- ings, the management of mobility for those assessed as
dence of 11.41 amputations/year/10 000 population in unsuited for a prosthetic limb was discussed.
males aged 85 years.3 As a consequence, the proportion
of amputees utilizing inpatient rehabilitation or skilled Methods
nursing facilities (where rehabilitation is provided) for
post-acute care is increasing,4 although the majority of A search of the MEDLINE database was undertaken
older amputees do return home post-rehabilitation.5–10 using PubMed (September 2010; see Appendix I for
Amputation represents a major challenge for both search strategies). The PubMed search was modified for
the patient and the rehabilitation team. It not only has use with Emtree subject headings and used to search
consequences for mobility and physical independence, the EMBASE database (September 2010; see Appen-
but it is also associated with adverse psychosocial dix II for search strategy). It was then modified again for
sequelae.11 For older dysvascular amputees with above- use with CINAHL headings and used to search the
knee amputation, less than one-third will become CINAHL database (September 2010; see Appendix III
for search strategy). The reference lists of studies
Accepted for publication 28 October 2012. retrieved and relevant review articles were searched for
additional studies (just two more studies were retrieved).
Correspondence: Dr Salih A Salih MBChB FRACP MPH,
Results from the searches were combined, and after
Centre for Research in Geriatric Medicine, The University of
Queensland, Ground Floor, Building 7 (GARU), Princess removal of duplication, 308 articles were found for pos-
Alexandra Hospital, Ipswich Rd., Woolloongabba. Brisbane, sible inclusion. The articles were then assessed using
QLD 4102, Australia. Email: [Link]@[Link] the inclusion criteria below.

264 兩 doi: 10.1111/ggi.12016 © 2012 Japan Geriatrics Society

 Rehabilitation of older vascular amputee

Inclusion criteria e Excluded were:

i amputations below the level of BKA, including
1 Original journal article
Syme’s or distal tibial amputations
a Reviews and meta-analyses excluded
ii above the level of an AKA, including hip
2 English language articles only
disarticulation
3 Mean or median age of the study sample 65 years or
6 Examining rehabilitation including predictors, influ-
older OR account for this older group separately
encing factors, outcomes and success rates OR have
a An exception to this was made for two articles that
rehabilitation as a significant theme of the article.
dealt with the geriatric issues of stroke and cogni-
The flow diagram of article selection is shown in
tive impairment
Figure 1.
4 Dysvascular cause for the amputation OR dysvascular
Study selection and data extraction were undertaken
causes account for >75% of the study sample
by the first author in consultation with co-authors.
5 Amputation levels of:
a Below-knee amputation (BKA) or transtibial ampu-
tation (TTA)
b Through-the-knee (TKA) or transgenicular (TGA) Results
or Gritti Stokes amputation
c Above-knee amputation (AKA) or transfemoral Table 1 summarizes study descriptors and outcomes of
amputation (TFA) selected studies quantifying the association between
d Bilateral amputees of above groups in all predictive factors and successful prosthetic fitting
combinations and/or use.

308 articles identified by search strategies

Excluded 61 non-English articles

247 articles

Excluded 187 articles not

meeting inclusion criteria

60 articles included in the review

Data extracted from 15 articles

quantifying the association between predictive
factors and successful prosthetic fitting / use
Figure 1 The flow diagram of article
selection.

© 2012 Japan Geriatrics Society 兩 265

266 兩
Table 1 Factors affecting successful prosthetic rehabilitation of the older vascular amputee

Study author, year Study design Characteristics of Outcome measure Predictors of successful
study population prosthetic use/fitting

Fletcher, 20022 Retrospective • n = 292 Successful prosthetic fitting defined as • Younger age
cohort • Age >65 years dismissal from physical therapy • Lower level of amputation
(inpatient or outpatient) with a definitive
prosthesis
Schoppen, 20036 Prospective • n = 46 Graded classification of prosthetic use • Better one leg balance on the unaffected
cohort • Mean age at amputation 73.9 years dichotomised into functional use limb
(SD 7.9) (score = I-IV) and non-functional use
• 70% male (score = V-VIII)
Callaghan 200817 Prospective • n = 166 1 and 6 months FU with: FMA
cohort • Mean age 66.73 years (SD 10.3) FMA (indoor use; outdoor use; h/day; • Believes treatment effective
• 69.3% male days/week) • Perceives symptoms as not fluctuating
LCI LCI
• Lack of distressing thoughts
Johannesson 201018 Prospective • n = 217 Functional status at 1 year follow up after • No significant predictors for good
cohort • Transtibial amputation amputation: function at 1 year follow up
• Mean age = 77 years Good: Patient received prosthesis, wore it
• 51.6% male daily and walks alone indoor, walks with
or without assistance outdoor.
Poor: Patient did not receive a prosthesis,
or received it but did not wear it daily,
AM Fleury et al.

unable to walk indoor without

assistance; mostly used wheelchair
O’Neill 200919 Prospective • n = 34 LCI • Better visual memory (figure recall)
cohort • Mean age 60.69 years (SD 13.98) SIGAM mobility grades • Immediate verbal memory
• 82.4% male • Younger age
• Lower level of amputation
• Absence of pain
Self reported hours of use • Better verbal fluency
Erjavec 200820 Prospective • n = 63 Whether prosthesis fitted or not based on • Better results on exercise stress test
cohort • Age 72.5 years (SD 9.1) clinical decision • Better results for 6-min walk time
• Transfemoral amputation • Higher FIM score
• Younger age
Traballesi 200721 Retrospective • n = 30 Barthel Index • Younger age
cohort • Age 65 years (SD 10) LCI • Good stump quality
• Bilateral above-knee amputees

© 2012 Japan Geriatrics Society

 Traballesi 199822 Prospective • n = 144 Barthel Index • Younger age
cohort • Mean age 68.7 years (SD 10.2) Rivermead Mobility Index • Absence of vascular impairment of
• 66% males residual limb
• Above-knee amputees
Kurichi 200715 Retrospective n = 2375 Prosthetic prescription within 1 year of • Younger age (<75)
cohort amputation • Higher cognitive and motor FIM score

© 2012 Japan Geriatrics Society

• Lower level of amputation
Larner 200323 Prospective • n = 43 Use of prosthesis during rehabilitation • Lower level of amputation
design • Mean age 66.35 years (SD 14.99) • Higher score on test of learning ability
study • Transfemoral/transtibial amputation
Chin 200224 Prospective n = 17 Ability to walk 100 m with/without cane • Higher %VO2 max
cohort • Fewer comorbidities
• Good ability to stand on remaining leg
• Strong will to walk
Fletcher 20011 A • n = 199 Successful prosthetic fitting defined as • Younger age
retrospective • Mean age at amputation 79.7 years dismissal from physical therapy • Absence of dementia
cohort (inpatient or outpatient) with a definitive • Lower level of amputation
prosthesis • Absence of cerebrovascular disease
Hermodsson Prospective • n = 112 Being prescribed a prosthesis • Younger age
199825 cohort • 51% male • Independent mobility outdoor prior
• Age 76.6 years (SD 10.3) amputation
• Unilateral transtibial amputation • No using wheelchair before amputation
• Left leg amputation
Functional use of prosthesis at 6 months • Male
FU • Independent mobility outdoor prior
Rehabilitation of older vascular amputee

amputation
Siriwardena 199126 Prospective • n = 598 Walking ability index with prosthesis at 6, • Younger age
cohort • Age >50 years 9 and 12 months follow up • Absence of IHD at 6 months
• Absence of hemiplegia at 12 months
• Absence of bronchitis at 12 months
• Single amputee at 12 months
OConnell 198927 Cohort • n = 46 Independent ambulation with prosthesis • Mild hemiparesis
study • Dysvascular amputee and • Lower level of amputation
hemiplegiaMean age 63 years • Prior ability to ambulate before second
disability
%VO2, percentage of volume oxygen maximum; FIM, Functional Independence Measure; FMA, functional measure for amputees; FU, follow up; IHD, ischemic heart disease; LCI, Locomotor Capability
Index; SIGAM, special interest group in amputation medicine.

兩 267
 AM Fleury et al.

Age Independence in ADL is highly correlated with the

severity of the hemiparesis. A total of 57% of amputees
Age alone is not an absolute contraindication to gait
with mild hemiparesis with residual hand function
retraining, with studies reporting success in patients
became independent in ADL compared with none of
aged 90 years and older.7,28–31 However, increasing
those with severe hemiplegia.27 Continence of bowel
age does reduce the likelihood of being fitted with a
and bladder, and age less than 60 years also predicted
prosthetic limb, of successfully mastering prosthesis
independence with ADL;27 however, the cohort were
use and returning to independent living post-
relatively young, with a mean age of 63 years and in an
amputation.2,5,6,12,15,22,30,32 Traballesi et al. concluded that
older frailer population, it is likely that this success rate
age was the most powerful prognostic factor in gait
would drop.
retraining, showing that the probability of successful
End-stage renal failure (ESRF) is a particular chal-
prosthesis use in an amputee aged less than 65 years was
lenge in the rehabilitation of elderly patients with
2.92-fold greater than of amputees aged 65 years and
lower limb amputation, frequently leading to local
older.22 Age also increases the likelihood of a higher
complications of prosthetic fitting as a result of con-
level of amputation.3,5
siderable changes in limb volume, as well additional
comorbidities, which are the underlying cause for
ESRF.42 Patients on hemodialysis require frequent
Comorbidities
absences from the rehabilitation ward, which, together
Comorbidities are common in this group of older with fatigue after dialysis, can impact on therapy.
patients. Bhangu et al. reported an average of 6.5 comor- Dialysis and rehabilitation units can usually negotiate
bidities per patient in their study sample.33 The number mutually agreeable times for hemodialysis to minimize
of comorbidities can affect the likelihood of successful impact on an amputee’s therapy sessions. One study
prosthetic rehabilitation.24,32,34 reported equal success in prosthetic rehabilitation of
Up to 75% of dysvascular amputees have concomi- amputees with ESRF and amputees without ESRF.42
tant cardiac disease.20,35,36 Many amputees do not show However, there were more TTA in the ESRF group
ischemic heart disease symptoms before amputation; than the control group, which might have biased the
however, as a result of the increased energy expenditure results. Most studies conclude that ESRF does
required for walking with a prosthesis, cardiac symp- adversely affect prosthetic rehabilitation.15,43
toms might become more apparent and exercise Cognitive impairment is a relative contraindication to
stress testing has been found to predict suitability prosthetic rehabilitation.37,39,44 Many studies on pros-
of candidates for prosthetic prescription.20 Cardiac thetic rehabilitation exclude amputees with dementia,
disease is a frequent reason for failure of prosthetic although criteria for the diagnosis of dementia are often
rehabilitation.15,30,35,37 unclear. Dementia, as defined by a Mini-Mental State
Underlying severe respiratory disease will also affect Examination score of less than or equal to 23 out of a
the ability of an amputee to successfully gait retrain.26,38 maximum of 30, was found in 18.4% of dysvascular
Chin et al. examined volume oxygen maximum (VO2 amputees in an Australian study.12 Harris et al. con-
max) of dysvascular transfemoral amputees to investi- cluded that the ability to learn is probably the most
gate if the percentage of VO2 max, as a measure of important determining factor for gait retraining.7 Larner
physical fitness, is a useful predictor of successful pros- and et al. found that higher scores on the Kendrick
thetic rehabilitation outcome (defined as the ability to Learning Objective Tool, together with lower level of
walk 100 m or more independently with/without a amputation, correctly predicted 81% of a patient’s
cane). The successful group were capable of strenuous potential ability to master prosthesis use.23 One study
exercise, reaching an intensity of 50% VO2 max or directly examined the effect of memory and executive
more, compared with the non-successful group who function on mobility in amputees, and found figure
failed to reach the intensity of 50% VO2 max.24 In addi- recall accounted for 25% of the variance in the locomo-
tion, many amputees are current or former smokers; tor capability index.19 However, that study had a small
therefore chronic obstructive pulmonary disease sample size and a young age group with a mean age of
(COPD) is a common comorbidity, with a prevalence of 60.69 years. Cognitive impairment increases with
25% in a Western Australian series.12 increasing age; therefore, the study outcomes in an
Stroke in dysvascular amputees has a prevalence of up older population might have given different results.
to 10%.20,27 This dual disability presents a major chal- Although cognitive impairment influences the ability to
lenge in rehabilitation of the older amputee.7,39–41 Stroke gait retrain, it also affects the ability of the amputee to
affects not only mobility,26 but also activities of daily monitor their stump and the remaining limb for signs
living (ADL) independence. OConnell and Gnatz of ulceration or infection, putting them at increased risk
reported a success rate of 39% for ADL independence of re-amputation at a higher level or contralateral
post-rehabilitation for amputees with dual disability.27 amputation.

268 兩 © 2012 Japan Geriatrics Society

 Rehabilitation of older vascular amputee

Older dysvascular amputees have falls risk factor pro- with premorbid function in the geriatric population is
files similar to those of older non-amputees, including social support. The presence of a supportive family
neuromuscular weakness, medication side-effects, pos- member or caregiver can be the difference between an
tural blood pressure changes, cognitive disorders, loss older amputee requiring residential care and continuing
of vision, and changes to sensation and propriocep- to live in the community.8,48 Amputees often require
tion.45 Diabetic retinopathy is a common complication assistance in ADL post-rehabilitation, although the
of poorly controlled diabetes and is frequently seen in proportion requiring assistance varies considerably
the amputee population.39 Amputees have, in addition between studies.8,48,49
to the aforementioned risk factors, significant changes
in sensation as a result of the absence of part of a limb.45
Level of amputation
Amputees frequently have phantom sensations postop-
eratively, and believe their amputated limb is still The level of amputation is generally selected by the
attached. When the amputee attempts to stand with this treating surgeon at the time of operation.41,50 Selection
sensation, they can fall as a result of overbalancing. of the correct level of amputation is problematic;
Proprioceptive input is also altered; for example, an however, in the elderly population all efforts should
above-knee amputee has proprioceptive input only from be made to preserve the knee joint if gait retraining is
the hip joint. Ulger et al. studied 25 prosthetic-wearing being considered.14,30,50,51 This is not only because of the
amputees aged over 65 years and found that 64% of increased energy requirements of prosthetic walking the
them had more than one fall in a year. They also found higher the level of amputation,35 but also the physical
that bilateral amputees fall more frequently than those challenge of getting from sitting to standing for the
with unilateral amputation.45 above-knee amputee. Vascular above-knee amputees
The influence of weight on amputee mobility has walking with a prosthesis have 120% higher O2 con-
been infrequently studied. Kalbaugh et al. examined the sumption than normal, whereas this is 62% above
role of body mass index (BMI) on outcome of prosthetic normal for a below-knee amputee.30 For bilateral AKA,
rehabilitation. They found no link between obesity and this has been estimated at 280% higher O2 consump-
poorer outcomes. However, being underweight (BMI tion than normal.21,31 Amputees reduce their gait speed
<18.5) was not only associated with poorer prosthetic to attempt to compensate for the increased energy
outcomes, but also decreased survival and reduced requirements.20,34,36 It has been shown consistently that
maintenance of ambulation, and led to reduction in patients with a TTA are more likely to walk with a
independent living.46 prosthesis than those with a higher amputation
level.12,14,15,18,26,30,32,40,41,50–54 Some authors advocate better
success in rehabilitating transgenicular amputees than
Premorbid function
TFA.50,52,55 More proximal amputation is also associated
An amputee’s premorbid function and mobility is one with increased mortality.3,18,41 However, the improved
of the most important factors in helping plan realistic mobility and survival is almost certainly as a result of
goals of rehabilitation.8,15,18,25,32,44 An American Veteran’s selection bias, because patients selected for a TTA are,
study43 showed that preoperative ambulatory status was in general, fitter. TFA might be selected for those with a
associated with the level of amputation. Patients who poor baseline function and not expected to undergo
were non-ambulatory pre-amputation, were twofold prosthetic fitting, as well as those requiring a palliative
more likely to have an AKA than a BKA. This was procedure.
possibly because they might have been in a worse overall
medical condition at the time of initial presentation.43
Remaining limb
Wong reported 66.1% of their study population were
community ambulators pre-amputation, but just 29.9% Progressive ischemia of the remaining limb can be a
were at 1 year post-amputation.32 When assessing an barrier to prosthesis prescription.39 The risk of con-
amputee, an understanding is required not only of the tralateral amputation is 10% per annum or 33–50% risk
previous mobility level, but also when this was last per- at 5 years.5,8 Therefore, amputees need to be taught how
formed, as many older adults might have been sedentary to monitor their remaining limb for any signs of ulcer-
for long periods.13 In particular, for diabetic patients ation or infection and act on these promptly. Bilateral
with occlusive arteriosclerotic disease, immobility as a amputees have much poorer functional outcomes than
result of claudication might have already led to decon- unilateral amputees; therefore every effort must be
ditioning before the amputation.13 Stump healing can be made to preserve the remaining limb.1,8,29 It is more
slow post-amputation, which can lead to further decon- appropriate to aim for wheelchair mobility than attempt
ditioning. It is clear that older patients require early gait retraining in the older dysvascular bilateral ampu-
post-amputation rehabilitation to maximize their func- tee.29,35,41,56 Several factors contribute to this including
tion and prevent further decline.14,47 Intrinsically linked the energy required, proprioceptive input with balance

© 2012 Japan Geriatrics Society 兩 269

 AM Fleury et al.

issues and frailty with comorbid diseases.21 The ability reduce the risk of functional decline and determine
to stand unsupported on the remaining limb is a key suitability for permanent prosthesis prescription.38,53,61
factor in predicting successful prosthesis use.6,24 However, for many older dysvascular amputees, wheel-
chair mobility might be a more appropriate goal. These
patients still require a period of rehabilitation to learn
Psychosocial factors transfers, ADL and become wheelchair independent.9,13
A patient’s motivation would seem crucial to any reha- They also require home assessment with an occupa-
bilitation program; however, there is mixed evidence to tional therapist and, if required, home modifica-
support this.23,39 Depression needs to be assessed and tion.11,44,49 However, local building styles and geography
managed appropriately in this group of older patients.44 can make home modifications, to enable community
Up to 60% of this group might have a depressive illness; access in a wheelchair, challenging.
however, interestingly, one study showed a higher In the present review, no study was identified that
prevalence of depression in those with a prosthetic limb specifically examined those unsuited to or who failed
than those without.57 Grief reactions are also noted due gait retraining. Studies of older amputees concluded
to the loss of a limb, and the perceived loss of indepen- that independence in ADL and wheelchair mobility are
dence and body image that goes with this. Many appropriate rehabilitation goals for non-prosthetic can-
amputee rehabilitation units have an amputee support didates.7,13,25,29 However, for some patients, indepen-
group. Weekend leave is encouraged once wheelchair dence post-amputation might be impossible and
independence is attained and the home environment residential aged care might be required.44 As yet, we do
is suitable. Management of pain can also improve not know what factors are most important in deciding
motivation. Approximately 70% of amputees suffer to pursue wheelchair independence with the aim of
from phantom pain, although Traballesi et al. did not returning home, compared with those requiring resi-
show stump pain as a factor influencing outcome of dential aged care. Further studies focusing on those
bilateral amputees.21 Callaghan et al. in a small study unsuitable for prosthetic rehabilitation are required.
using the common sense self-regulation model showed The role of the geriatrician in the management of this
that those who perceived symptoms as fluctuating, older group of patients is not well defined, with reha-
especially pain, were more at risk of not rehabilitating bilitation specialists continuing to manage the majority
successfully with a prosthesis.17 of amputees. Given the predicted rise in the numbers of
older people with amputations,2 especially the very old,
it is likely that geriatricians will need to become involved
Discussion in the care of this amputee population. It might be that
rehabilitation specialists continue to manage prosthetic
In summary, there are many factors influencing suc- rehabilitation candidates while geriatricians care for the
cessful prosthetic use in older dysvascular amputees, frailer, more medically unwell patient unlikely to benefit
and identifying those that might become successful from prosthetic rehabilitation, but who needs to master
prosthesis users is difficult.12,38,39 This patient group has wheelchair mobility and transfers or who ultimately will
a poor survival rate, with a 2-year mortality rate of require residential aged care.
16%40–56%14 and a 5-year survival of just 22.6%3–
45%.33 The survival of this group of patients has not
changed in 40 years.2 Also, a significant number of Conclusions
amputees either discard their limbs or do not use
them functionally at follow up, as the physical effort Identifying patients who will undergo successful pros-
of using the prosthesis might exceed their thetic rehabilitation is challenging. Multidisciplinary
reserves.6,13,24,31,33,34,52,54,58 Yusuf et al. found that just teams are good at predicting successful prosthesis users,
17.5% of their total population achieved Stanmore Gra- but less so at predicting non-users.39 It is also difficult to
de III mobility (indoor walking with aids), and this predict who will continue to wear their prosthetic limb
decreased to 12.4% at a mean of 23 months of follow and to what extent. Identifying patients likely to benefit,
up.52 Therefore, a balance needs to be achieved between or not, from gait retraining is challenging because of the
the potential success of prosthetic rehabilitation and its heterogeneity within the amputee literature. Studies use
associated costs, both in financial and quality of life different outcome assessment tools and the definition of
terms.59 Ultimately, the duration required for gait successful prosthetic use also varies. Formulation of a
retraining must be balanced with the relatively poor single amputee-specific mobility assessment tool and
survival of this cohort.59 agreement on the definition of successful gait retraining
The aim of rehabilitation in older dysvascular ampu- would allow comparison between studies, and therefore
tees is to restore walking ability.60 Many rehabilitation allow more accurate patient selection. Given the
units use an early walking aid or interim prosthesis to challenges in selection of successful prosthetic

270 兩 © 2012 Japan Geriatrics Society

 Rehabilitation of older vascular amputee

rehabilitation candidates, further research into this area 17 Callaghan B, Condie E, Johnston M. Using the common
is required. sense self-regulation model to determine psychological
predictors of prosthetic use and activity limitations in
lower limb amputees. Prosthet Orthot Int 2008; 32 (3): 324–
Disclosure statement 336.
18 Johannesson A, Larsson G, Ramstrand N, Lauge-Pedersen
H, Wagner P, Atroshi I. Outcomes of a standardized sur-
The authors declare that there was no financial assis- gical and rehabilitation program in transtibial amputation
tance or relationship that imposed a conflict of interest for peripheral vascular disease: a prospective cohort study.
in carrying out this project. Am J Phys Med Rehabil 2010; 89 (4): 293–303.
19 O’Neill BF, Evans JJ. Memory and executive function
predict mobility rehabilitation outcome after lower-limb
References amputation. Disabil Rehabil 2009; 31 (13): 1083–1091.
20 Erjavec T, Presern-Strukelj M, Burger H. The diagnostic
1 Fletcher DD, Andrews KL, Butters MA, Jacobsen SJ, importance of exercise testing in developing appropriate
Rowland CM, Hallett JW Jr. Rehabilitation of the geriatric rehabilitation programmes for patients following trans-
vascular amputee patient: a population-based study. Arch femoral amputation. Eur J Phys Rehabil Med 2008; 44 (2):
Phys Med Rehabil 2001; 82 (6): 776–779. 133–139.
2 Fletcher DD, Andrews KL, Hallett JW Jr, Butters MA, 21 Traballesi M, Porcacchia P, Averna T et al. Prognostic
Rowland CM, Jacobsen SJ. Trends in rehabilitation after factors in prosthetic rehabilitation of bilateral dysvascular
amputation for geriatric patients with vascular disease: above-knee amputee: is the stump condition an influenc-
implications for future health resource allocation. Arch ing factor? Eura Medicophys 2007; 43 (1): 1–6.
Phys Med Rehabil 2002; 83 (10): 1389–1393. 22 Traballesi M, Brunelli S, Pratesi L, Pulcini M, Angioni C,
3 Carmona GA, Hoffmeyer P, Herrmann FR et al. Major Paolucci S. Prognostic factors in rehabilitation of above
lower limb amputations in the elderly observed over ten knee amputees for vascular disease. Disabil Rehabil 1998; 20
years: the role of diabetes and peripheral arterial disease. (10): 380–384.
Diabetes Metab 2005; 31 (5): 449–454. 23 Larner S, van Ross E, Hale C. Do psychological measures
4 Dillingham TR, Pezzin LE, Mackenzie EJ. Discharge des- predict the ability of lower limb amputees to learn to use a
tination after dysvascular lower-limb amputations. Arch prosthesis? Clin Rehabil 2003; 17 (5): 493–498.
Phys Med Rehabil 2003; 84 (11): 1662–1668. 24 Chin T, Sawamura S, Fujita H et al. %VO2max as an
5 Kerstein MD, Zimmer H, Dugdale FE, Lerner E. What indicator of prosthetic rehabilitation outcome after dysvas-
influence does age have on rehabilitation of amputees? cular amputation. Prosthet Orthot Int 2002; 26 (1): 44–49.
Geriatrics 1975; 30 (12): 67–71. 25 Hermodsson Y, Ekdahl C, Persson BM. Outcome after
6 Schoppen T, Boonstra A, Groothoff JW, de Vries J, trans-tibial amputation for vascular disease. A follow-up
Goeken LN, Eisma WH. Physical, mental, and social pre- after eight years. Scand J Caring Sci 1998; 12 (2): 73–80.
dictors of functional outcome in unilateral lower-limb 26 Siriwardena GJA, Bertrand PV. Factors influencing reha-
amputees. Arch Phys Med Rehabil 2003; 84 (6): 803–811. bilitation of arteriosclerotic lower limb amputees. J Rehabil
7 Harris KA, Van Schie L, Carroll SE et al. Rehabilitation Res Dev 1991; 28 (3): 35–44.
potential of elderly patients with major amputations. J Car- 27 OConnell PG, Gnatz S. Hemiplegia and amputation: reha-
diovasc Surg (Torino) 1991; 32 (4): 463–467. bilitation in the dual disability. Arch Phys Med Rehabil 1989;
8 Dawson I. Late outcomes of limb loss after failed infrain- 70 (6): 451–454.
guinal bypass. J Vasc Surg 1995; 21 (4): 613–622. 28 Graham LA, Fyfe NCM. Prosthetic rehabilitation of ampu-
9 Nitz JC. Rehabilitation outcomes after bilateral lower limb tees aged over 90 is usually successful. Disabil Rehabil 2002;
amputation for vascular disease. Physiother Theory Pract 24 (13): 700–701.
1993; 9 (3): 165–170. 29 Collin C, Wade DT, Cochrane GM. Functional outcome
10 Harris PL, Read F, Eardley A. The fate of elderly amputees. of lower limb amputees with peripheral vascular disease.
Br J Surg 1974; 61 (8): 665–668. Clin Rehabil 1992; 6 (1): 13–21.
11 Pell JP, Donnan PT, Fowkes FG, Ruckley CV. Quality of 30 Steinberg F, Sunwoo I, Roettger R. Prosthetic rehabilita-
life following lower limb amputation for peripheral arterial tion of geriatric amputee patients: a follow up study. Arch
disease. Eur J Vasc Surg 1993; 7 (4): 448–451. Phys Med Rehabil 1985; 66 (11): 742–745.
12 Lim TS, Finlayson A, Thorpe JM et al. Outcomes of a 31 Wolf E, Lilling M, Ferber I, Marcus J. Prosthetic rehabili-
contemporary amputation series. ANZ J Surg 2006; 76 (5): tation of elderly bilateral amputees. Int J Rehabil Res 1989;
300–305. 12 (3): 271–278.
13 Davis WC, Blanchard RS, Jackson FC. Rehabilitation of 32 Wong MWN. Changing dynamics in lower-extremity
the geriatric amputee: a plea for moderation. Arch Phys Med amputation in China. Arch Phys Med Rehabil 2005; 86 (9):
Rehabil 1967; 48 (1): 31–36. 1778–1781.
14 Pohjolainen T, Alaranta H, Wikstrom J. Primary survival 33 Bhangu S, Devlin M, Pauley T. Outcomes of individuals
and prosthetic fitting of lower limb amputees. Prosthet with transfemoral and contralateral transtibial amputation
Orthot Int 1989; 13 (2): 63–69. due to dysvascular etiologies. Prosthet Orthot Int 2009; 33
15 Kurichi JE, Kwong PL, Reker DM, Bates BE, Marshall CR, (1): 33–40.
Stineman MG. Clinical factors associated with prescription 34 Beekman CE, Axtell LA. Prosthetic use in elderly patients
of a prosthetic limb in elderly veterans. J Am Geriatr Soc with dysvascular above-knee and through-knee amputa-
2007; 55 (6): 900–906. tions. Phys Ther 1987; 67 (10): 1510–1516.
16 Buzato MA, Tribulatto EC, Costa SM, Zorn WG, 35 Moore TJ, Barron J, Hutchinson F 3rd, Golden C, Ellis C,
van Bellen B. Major amputations of the lower leg. The Humphries D. Prosthetic usage following major lower
patients two years later. Acta Chir Belg 2002; 102 (4): 248– extremity amputation. Clin Orthop Relat Res 1989; 238: 219–
252. 224.

© 2012 Japan Geriatrics Society 兩 271

 AM Fleury et al.

36 Cruts HEP, de Vries J, Zilvold G, Huisman K, van Alste JA, 54 McWhinnie DL, Gordon AC, Collin J, Gray DW, Morri-
Boom HBK. Lower extremity amputees with peripheral son JD. Rehabilitation outcome 5 years after 100 lower-
vascular disease: graded exercise testing and results of limb amputations. Br J Surg 1994; 81 (11): 1596–1599.
prosthetic training. Arch Phys Med Rehabil 1987; 68 (1): 55 Steen Jensen J, Mandrup Poulsen T. Success rate of pros-
14–19. thetic fitting after major amputations of the lower limb.
37 Met R, Janssen LI, Wille J et al. Functional results after Prosthet Orthot Int 1983; 7 (2): 119–121.
through-knee and above-knee amputations: does more 56 Volpicelli LJ, Chambers RB, Wagner FW Jr. Ambulation
length mean better outcome? Vasc Endovascular Surg 2008; levels of bilateral lower-extremity amputees. Analysis of
42 (5): 456–461. one hundred and three cases. J Bone Joint Surg Am 1983; 65
38 Marzoug EA, Landham TL, Dance C, Bamji AN. Better (5): 599–605.
practical evaluation for lower limb amputees. Disabil 57 Langer KG. Depression in disabling illness: severity and
Rehabil 2003; 25 (18): 1071–1074. patterns of self-reported symptoms in three groups. J
39 Campbell WB, Ridler BMF. Predicting the use of prosthe- Geriatr Psychiatry Neurol 1994; 7 (2): 121–128.
ses by vascular amputees. Eur J Vasc Endovasc Surg 1996; 12 58 De Fretes A, Boonstra AM, Vos LD. Functional outcome
(3): 342–345. of rehabilitated bilateral lower limb amputees. Prosthet
40 Gugulakis AG, Lazaris AM, Vasdekis SN et al. Rehabilita- Orthot Int 1994; 18 (1): 18–24.
tion outcome in patients with lower limb amputations 59 Hubbard WA. Rehabilitation outcomes for elderly lower
because of arterial occlusive disease: is it worth trying for limb amputees. Aust J Physiother 1989; 35 (4): 219–224.
the lowest possible amputation level? A prospective study. 60 Sansam K, Neumann V, O’Connor R, Bhakta B. Predicting
Int J Rehabil Health 1999; 5 (1): 65–70. walking ability following lower limb amputation: a system-
41 Robinson KP. Long posterior flap amputation in geriatric atic review of the literature. J Rehabil Med 2009; 41 (8):
patients with ischaemic disease. Ann R Coll Surg Engl 1976; 593–603.
58 (6): 440–451. 61 Monga TN, Symington DC. The airsplint as a pneumatic
42 Lucke C, Beindorff N, Roy T, Hoy L. Rehabilitation of prosthesis in management of the elderly amputee. Physio-
lower extremity amputation due to peripheral arterial ther Can 1984; 36 (2): 61–65.
occlusive disease in patients with end-stage renal failure.
Vasc Surg 1999; 33 (1): 33–41.
43 Toursarkissian B, Shireman PK, Harrison A, D’Ayala M, Appendix I: PubMed Search Terms
Schoolfield J, Sykes MT. Major lower-extremity amputa-
tion: contemporary experience in a single Veterans Affairs
“Aged”[Mesh] OR Aged OR Elderly OR Old OR Frail
institution. Am Surg 2002; 68 (7): 606–610.
44 Rowe J. Generic geriatric skills in prosthetic rehabilitation OR Geriatric OR Geriatrics
following lower limb amputation. CME J Geriatr Med 2006; AND
8 (1): 40–43. “Diabetes Complications”[Mesh] OR “Peripheral Vas-
45 Ulger O, Topuz S, Bayramlar K, Erbahceci F, Sener G. cular Diseases”[Mesh] OR “Peripheral Vascular Dis-
Risk factors, frequency, and causes of falling in geriatric
eases” OR “Peripheral arterial Diseases” OR “Diabetes
persons who has had a limb removed by amputation. Top
Geriatr Rehabil 2010; 26 (2): 156–163. Complications” OR “Diabetic foot” OR “Diabetic neu-
46 Kalbaugh CA, Taylor SM, Kalbaugh BA et al. Does obesity ropathy” OR “Diabetic Complications” OR
predict functional outcome in the dysvascular amputee? “Diabetic Complication” OR “Diabetic Feet” OR
Am Surg 2006; 72 (8): 707–712. Dysvascular
47 Cutson TM, Bongiorni D, Michael JW, Kochersberger G.
AND
Early management of elderly dysvascular below-knee
amputees. J Prosthet Orthot 1994; 6 (3): 62–66. “Amputees”[Mesh] OR “Amputation”[Mesh] OR
48 Mac Neill HL, Devlin M, Pauley T, Yudin A. Long-term Amputees OR Amputee OR Amputation
outcomes and survival of patients with bilateral transtibial AND
amputations after rehabilitation. Am J Phys Med Rehabil “Artificial Limbs”[Mesh] OR Prosthesis OR Prostheses
2008; 87 (3): 189–196.
OR Prosthetic OR Prosthetics OR “Artificial Leg” OR
49 Jones L, Hall M, Schuld W. Ability or disability? A study of
the functional outcome of 65 consecutive lower limb “Artificial Limb” OR “Artificial Limbs”
amputees treated at the Royal South Sydney Hospital in AND
1988–1989. Disabil Rehabil 1993; 15 (4): 184–188. “Rehabilitation”[Mesh] OR Rehabilitation OR Rehab
50 Stirnemann P, Mlinaric Z, Oesch A, Kirchhof B, Althaus OR “Restorative care” OR Restoration
U. Major lower extremity amputation in patients with
peripheral arterial insufficiency with special reference to
the transgenicular amputation. J Cardiovasc Surg (Torino) Appendix II: EMBASE Search Terms
1987; 28 (2): 152–158.
51 Hierton T, James U. Lower extremity amputation in
Uppsala county 1947–1969. Incidence and prosthetic “aged”/exp OR aged OR elderly OR old OR frail OR
rehabilitation. Acta Orthop Scand 1973; 44 (4): 573–582. geriatric OR geriatrics
52 Yusuf SW, Baker DM, Wenham PW, Makin GS, Hopkin- AND
son BR. Role of Gritti-Stokes amputation in peripheral “diabetic angiopathy”/exp OR “diabetic neuropathy”/
vascular disease. Ann R Coll Surg Engl 1997; 79 (2): 102–
exp OR “diabetic foot”/exp OR “peripheral vascular
104.
53 Condie E, Treweek S, Jones D, Scott H. A one-year disease”/exp OR “peripheral vascular diseases” OR
national survey of patients having a lower limb amputa- “peripheral arterial diseases” OR “diabetes complica-
tion. Physiotherapy 1996; 82 (1): 14–20. tions” OR “diabetic foot” OR “diabetic neuropathy” OR

272 兩 © 2012 Japan Geriatrics Society

 Rehabilitation of older vascular amputee

“diabetic complications” OR “diabetic complication” (MH “Diabetic Angiopathies+”) OR (MH “Diabetic

OR “diabetic feet” OR dysvascular Neuropathies+”) OR (MH “Peripheral Vascular Dis-
AND eases+”) OR “Peripheral Vascular Diseases” OR
“disabled person”/exp OR “amputation”/exp OR ampu- “Peripheral arterial Diseases” OR “Diabetes Complica-
tees OR amputee OR amputation tions” OR “Diabetic foot” OR “Diabetic neuropathy”
AND OR “Diabetic Complications” OR “Diabetic Complica-
“limb prosthesis”/exp OR prosthesis OR prostheses OR tion” OR “Diabetic Feet” OR Dysvascular
prosthetic OR prosthetics OR “artificial leg” OR “artifi- AND
cial limb” OR “artificial limbs” (MH “Amputees”) OR (MH “Amputation+”) OR
AND Amputees OR Amputee OR Amputation
“rehabilitation”/exp OR rehabilitation OR rehab OR AND
“restorative care” OR restoration (MH “Limb Prosthesis”) OR Prosthesis OR Prostheses
OR Prosthetic OR Prosthetics OR “Artificial Leg” OR
Appendix III: CINAHL Search Terms “Artificial Limb” OR “Artificial Limbs”
AND
(MH “Aged+”) OR Aged OR Elderly OR Old OR Frail (MH “Rehabilitation+”) OR Rehabilitation OR Rehab
OR Geriatric OR Geriatrics OR “Restorative care” OR Restoration
AND

© 2012 Japan Geriatrics Society 兩 273

Care of elderly lower limb amputees, as described in medical
and nursing records

Siv Bäck-Pettersson BScN, RN and Cecilia Björkelund MD, PhD

Department of Primary Health Care, Faculty of Medicine, The Sahlgrenska Academy at Göteborg University, Göteborg, Sweden

Scand J Caring Sci; 2005; 19; 337–343 most common amputation level was below the knee. LLA
Care of elderly lower limb amputees, as described patients were reported to have major problems main-
in medical and nursing records taining physical and mental functions, markedly deteri-
orated general health status and severe pain problems.
The aim of this study was to characterize elderly lower
LLA patients require well functioning and qualified care
limb amputees and explore problems/requirements
and rehabilitation. The patients surviving after 6 months
inherent in their care. A retrospective study of medical
had permanent problems in the area of nutrition, elimin-
and nursing records of patients who had undergone lower
ation, skin ulceration, sleep, pain and pain alleviation. The
limb amputation (LLA) at Uddevalla General Hospital in
patients who died during the hospital stay had problems in
1997 was conducted. Demographic data were compared
all these areas. With increasing shorter stays in hospital
with those from a comparable regional health care dis-
and decreasing resources in primary and municipal care,
trict. Hospitalization, rehabilitation and nursing-related
there is a risk that these patients’ needs and requirements
data related to subjects alive after 6 months were com-
for professional care might be underestimated and thus
pared with data concerning those deceased during hos-
remain unfulfilled.
pital stay and within 6 months after amputation. During
the defined period, the study population consisted of 45
Keywords: lower limb amputation, diabetes,
patients aged 60 and above. Fifty-six percent were men,
hospitalization, mortality, medical records, nursing
with a mean age of 81.6. Eight patients died in the hos-
records.
pital post-amputation and five died within 6 months of
surgery. The aetiology of the diagnosis leading to the LLA
Submitted 10 May 2004, Accepted 22 July 2005
was cardiovascular disease in the majority of cases. The

During recent years, international studies have reported

Introduction
a decrease in the incidence of major amputation in both
According to the Swedish National Board of Health and diabetic and nondiabetic patients, irrespective of sex (5).
Welfare (SNBHW), more than 3000 patients undergo However, current national studies report that vascular
amputation annually in Sweden. The amputation rate surgery has not significantly improved morbidity or
increases with age, and most amputations are performed amputation rate in this group of patients (6).
on patients over 60, with a mean age of 72 (1). A little Elderly patients who undergo lower limb amputation
more than half of amputees are men. The most common (LLA) are large consumers of medical and nursing care,
primary diagnosis is arteriosclerosis, and diabetes with with a mean of 1.25 care episodes per amputee (1). It has
peripheral circulatory disorder is the most common form of been reported that the disability following LLA is perma-
diabetes in connection with amputation (1). International nent in elderly patients due to diabetes and other pathol-
figures are similar to Swedish figures in this context. The ogy in many organs (i.e. heart, brain, eye, kidney and
overall aging of the population is expected to increase the nervous system), and that these patients therefore will be
prevalence of amputation in the elderly (2–4). continuously dependent on health care services (7).
The risk that the health care needs of individuals with
chronic disabilities will remain unfulfilled in the current
health care organization has been emphasized. The
Correspondence to: increasing shorter stays in hospital are a particular problem
Siv Bäck-Pettersson, Citronfjärilvägen 1, SE-444 46 Stenungsund, for the elderly. In addition, the ongoing decrease in the
Sweden. number of hospital beds leads to requirements for more
E-mail: [Link]-pettersson@[Link] outpatient care and municipal services, placing an

Ó 2005 The Authors, Scand J Caring Sci; 2005; 19, 337–343 337
338 S. Bäck-Pettersson, C. Björkelund

increased responsibility on the municipalities and primary The information obtained from the medical records was
health care centres for providing continuity in the care of entered into computer files, verified and analysed by des-
elderly, chronically ill patients (8). criptive analysis using SAS (12). The chi-square test was
Postoperative care and rehabilitation of older amputees used to compare frequencies. When comparing means, the
is complex due to the various medical comorbidities. Student’s t-test was used and a logistic regression model
Although the medical treatment of these patients may be was used to compare amputation frequencies between
completed a short time after the operation, they never- populations.
theless have a residual need of treatment, extensive The information obtained from the nursing records
rehabilitation and nursing care (9–11). As most available (n ¼ 45) was subjected to a method of qualitative content
research has been carried out in the US and Great Brit- analysis as in which both manifest and latent meanings
ain, we thought it as important to acquire knowledge of were sought (13, 14). The analysis was conducted by data
the current situation for aged amputees in a Swedish reduction. The Swedish documentation model, based on
context. Wellbeing, Integrity, Prevention and Security (VIPS) with
Uddevalla General Hospital (UGH) was regarded as keywords on two levels corresponding to the nursing
suitable for this type of study, due to the representative- process and relevant concepts for patient care, was used as
ness of the surrounding population. The catchment area of a scheme for categorizing during the analysis process (15).
the hospital covers a little less than 300 000 of the popu- Frequencies of descriptive statements were calculated
lation in the northern part of the Västra Götaland Region. and qualitative aspects of statements were organized into
The aim of this study was to describe the characteristics of categories by themes, indicators and tables. The data were
elderly lower limb amputees and problems/requirements re-examined and re-assessed several times by the two
inherent in their care, as described in medical and nursing authors to obtain inter-rater reliability. The analysis and
records. interpretation process involved a rigorous recontextuali-
zation of the information in the hospital records to obtain
external validity (16).
Material and methods
The Ethics Committee of Göteborg University approved
A retrospective cohort study was conducted to characterize the study (reference number L 435–97).
elderly LLA. The total number of LLA in the UGH district in
1997 was 90, of which 48 patients were >60 years of age.
Results
The medical records of these patients were retrospectively
scrutinized. As the records of three patients were missing,
Demographic data
the study population consisted of 45 patients.
Demographic characteristics, hospitalization, rehabilit- The total number of LLA in the UGH district was 90 in
ation and nursing-related characteristics were collected 1997. Mean age at amputation was 80.2 for women and
and compared for patients alive after and deceased 75.9 for men (Table 1).
respectively within 6 months after amputation. Data from LLA appeared to be twice as common in the UGH catch-
the UGH catchment area were cross-checked with SNBHW ment area than in another, comparable, catchment area in
data to ensure that all lower limb amputees in the catch- Sweden (Table 1). The difference was statistically significant
ment area were included. Data from a demographically for men (p < 0.05). There was also a significant difference
corresponding region the Kristianstad Health Care District, between the two areas concerning mean age at the time of
were obtained from the SNBHW to enable comparison amputation (p ¼ 0.013) which was higher in Kristianstad
between the two districts. (mean age men 76.2, women 82.0). Both men and women

Table 1 Size as well as mean age and SD of the population, number of lower limb amputations (LLA), mean age of amputees (aged > 60) in the
catchment areas of Uddevalla General Hospital and Kristianstad Health Care District (KHCD) respectively

Uddevalla Kristianstad

Men Women Total Men Women Total

Population 136 068 135 591 271 659 134 629 137 746 272 375
Mean age (SD) 39.09 (23.42) 41.40 (24.65) 40.25 (24.06) 38.90 (23.35) 41.48 (24.70) 40.20 (24.08)
LLA (n) 53 37 90 22 27 49
Mean age (SD) 75.85 (12.87) 80.20 (11.37) 77.63 (12.39) 76.18 (10.72) 82.00 (8.65) 79.39 (9.97)
LLA per 100 000 10* 7 8 4 5 4

*p < 0.05, compared to KHCD men.

Ó 2005 The Authors, Scand J Caring Sci; 2005; 19, 337–343

 Care of elderly lower limb amputees 339

were younger when undergoing amputation at UGH (mean majority had more than one diagnosis (e.g. CVD, diabetes).
age men 75.9, women 80.2). Moreover, men were younger There was a tendency towards difference between men
than women at the time of surgery in both areas. and women regarding marital status. Twice as many men
were married or cohabiting. The men lived with relatives
more frequently than the women, while the women more
Hospitalization, morbidity and mortality data
often lived in service residences. All men and women older
The study population consisted of 25 men and 20 women than 87 years lived in nursing homes. A majority of those
(Table 2). No statistically significant age difference was who had their own residences were men. Smoking habits
found between men and women. Forty-two percent of the were incompletely documented, mentioned in 31% of the
patients had diabetes and all the men and 85% of the medical records.
women had cardiovascular disease (CVD). There were no Nearly one-third of the study population died within
significant differences between men and women. A 6 months (Table 3), with no significant difference

Table 2 Demographic characteristics of the

study population of LLA at UGH Men, age Women, age Total, age
(years) (years) (years)

n Mean (SD) n Mean (SD) n Mean (SD) p

Total 25 81.8 (7.1) 20 81.4 (6.3) 45 81.6 (6.7)

Diabetes 9 81.9 (7.4) 10 82.0 (5.7) 19 81.9 (6.4) NS
Cardiovascular disease 25 81.8 (7.1) 17 81.1 (6.4) 42 81.5 (6.8) NS
Number of diagnosis
1 3 85.0 (8.7) 4 84.0 (5.9) 7 84.2 (6.6)
2 4 86.8 (4.9) 3 86.0 (4.6) 7 86.4 (4.4)
3 7 75.6 (6.6) 4 78.5 (7.9) 11 76.3 (6.9)
>3 11 83.0 (5.3) 9 80.3 (6.0) 20 81.1 (5.9) NS
Marital status
Married/cohabiting 15 80.1 (7.3) 7 78.7 (6.8) 22 79.7 (7.0)
Single 10 84.3 (6.3) 13 82.8 (5.8) 23 83.5 (6.0) NS
Housing
Own 17 79.6 (6.4) 9 79.2 (6.2) 26 79.5 (6.3)
With relatives 3 85.0 (3.6) 1 76.0 (0.0) 4 82.8 (5.4)
In service residence 1 75.0 (0.0) 5 80.2 (4.9) 6 79.3 (4.8)
In a nursing home 4 90.5 (4.8) 5 87.6 (4.4) 9 88.9 (4.6) NS
Smoking habits
Current smoker 5 83.0 (10.3) 2 77.5 (4.9) 7 81.4 (9.0)
Ex-smoker 3 80.0 (9.5) 1 72.0 (0.0) 4 78.0 (8.8)
Nonsmoker 2 91.5 (4.9) 1 77.0 (0.0) 3 86.7 (9.0)
Unknown 15 80.5 (5.0) 16 82.3 (6.2) 31 81.6 (5.7) NS

NS, no statistical significance; LLA, lower limb amputation; UGH, Uddevalla General Hospital.

Table 3 Mean duration of hospitalization (days), amputation frequency and mortality within 6 months in relation to sex and age in patients alive after
6 months and deceased within 6 months respectively

Men Women Total Diabetes CVD* 67–84 >85

Alive after 6 months

Number 19 13 32 15 29 20 12
Mean duration of hospitalization (SD) 22.0 (22.3) 17.5 (11.3) 20.2 (18.5) 18.5 (16.4) 21.3 (19.1) 20.1 (14.1) 20.4 (24.9)
Prior amputation 7 8 15 9 13 9 6
First amputation 12 5 17 6 16 11 6
Deceased within 6 months
Number 6 7 13 4 13 6 7
Mean duration of hospitalization (SD) 14.3 (9.5) 12.1 (7.9) 13.6 (8.4) 12.7 (9.6) 13.6 (8.4) 17.8 (9.7) 9.1 (4.5)
Prior amputation 4 1 5 1 5 2 3
First amputation 2 6 8 3 8 8 4

CVD, cardiovascular disease.

Ó 2005 The Authors, Scand J Caring Sci; 2005; 19, 337–343

340 S. Bäck-Pettersson, C. Björkelund

between men and women. The mean duration of hospi- The total number of lower limb amputations was 56 in 45
talization was lower for those deceased within 6 months, patients. No relationship was found between the number
compared to survivors, with no statistically significant of amputations and age and between the number of
difference between men and women. In this latter group, operations and mortality.
the patients over 85 years had the shortest mean hospi- A majority of the deceased died in connection with
talization stays, this being due to high mortality. Corres- the first operation. The majority of the 13 deceased
ponding differences were not observed among those alive patients died within 18 days. Among the deceased, the
after 6 months. A majority in both groups had undergone mean hospitalization duration decreased with age. Most
their first amputation. There was no significant difference deceased patients died during hospitalization, either in
in mean hospitalization duration between men and direct relation to the amputation or from postoperative
women (Table 3). complications. All deaths except one during the study
There was no significant difference in mortality between period occurred during the first 6 months after the
men and women (Table 4), nor any significant difference amputation. Hence, 6 months was regarded as the post-
in age at the time of amputation, but the deceased were operative period. A majority of the amputees were dis-
older. No significant difference was found in the occur- charged either to nursing homes or rehabilitation units.
rence of diabetes between men and women in the two Wound care and walking exercises were prescribed most
groups, but there was a significant difference in mean age frequently to the surviving group. An equal number of
(i.e. 5 years) between patients with and without diabetes. patients in both groups were given no orders at all by
their doctors.
Table 4 Mean age, gender and mean days of hospitalization, diagno-
sis, number and type of amputations, discharge destination and Nursing data
doctor’s orders for those alive after 6 months and for those deceased
within six months after amputation respectively The following problems have been recorded in nursing
records: breathing, circulation, nutrition, elimination,
skin, sleep, activity, mobilization, pain, pain alleviation,
Variable studied Alive Deceased Total
communication, mental condition, well being and psy-
Age (SD) 80.6 (6.7) 84.1 (6.4) 81.6 (6.7) chosocial status. Administrative data were also recorded
Sex (Table 5).
Men 19 (76) 6 (24) 25 (56) Nursing records revealed that the most frequently
Women 13 (65) 7 (35) 20 (44)
reported problems pre- and postamputation were related
Total 32 (71) 13 (29) 45 (100)
to pain/pain alleviation, nutrition, elimination, circulation,
Mean duration of 20.2 (18.5) 13.1 (8.4) 18.1 (16.5)
ulceration and sleep. All patients were reported as suffer-
hospitalization (SD)
Diagnosis
ing severe pain and/or having problems with pain allevi-
Diabetes-related diseases 15 (49) 4 (31) 19 (42) ation. The analysis indicates three specific categories of
Cardiovascular diseases 29 (91) 13 (100) 42 (93) pain: pain related to underlying illness, pain as a compli-
Other diseases 11 (34) 5 (38) 16 (36) cation of amputation or pain without cause. In addition to
Number of amputations severe pain and problems with pain alleviation, all
1 21 (66) 12 (92) 33 (73) amputees who died during hospitalization had problems
>1 11 (34) 1 (8) 12 (27) with nutrition, elimination, wound healing, sleep, com-
Type of amputation munication, well being and mental condition.
Above knee 4 (13) 4 (31) 8 (18)
Problems with breathing, activity, personal care, mobil-
At knee 5 (16) 1 (8) 6 (13)
ization and communication were reported less frequently
Below knee 20 (63) 7 (54) 27 (60)
in the surviving group than in those deceased during
Other 3 (9) 1 (8) 4 (9)
Discharge destination
hospital stay.
Home 6 (19) 1 (8) 7 (16) Several notations relating to the same problem were
Rehabilitation unit 10 (31) 1 (8) 11 (24) found. The most varied notations were used when for-
Hospital care 3 (9) 1 (8) 4 (9) mulating the patients’ problems related to the elimination,
Nursing home 13 (41) 2 (15) 15 (33) pain, skin and well being keywords. The well being,
Deceased in hospital 0 8 (62) 8 (18) mental condition and communication keywords were used
Doctor’s orders synonymously in several of the nursing records. The
Walking exercises 13 (93) 1 (8) 14 (31) Swedish word ‘ont’ (pain) was used in several records
Strengthening exercises 9 (75) 3 (23) 12 (27)
without explanation, characterization or specific localiza-
Wound care 18 (82) 4 (31) 22 (49)
tion. Patient wishes, values, subjective experiences and
No doctor’s orders 7 (50) 7 (54) 14 (31)
participation were only noted occasionally. Smoking
Values in parentheses are expressed in percentage. behaviour was noted in the nursing records.

Ó 2005 The Authors, Scand J Caring Sci; 2005; 19, 337–343

 Care of elderly lower limb amputees 341

Table 5 Nursing problems/patient needs as documented in nursing records of LLA patients at UGH (n ¼ 45)

Alive during hospital stay Deceased during hospital stay

Men Women Men + women Men Women Men + women Total

Criteria in patient recordsa Notations (n ¼ 21) (n ¼ 16) (n ¼ 37) (n ¼ 4) (n ¼ 4) (n ¼ 8) (n ¼ 45)

Circulation (n ¼ 31)
No circulatory problemsb 2 2 1 3 0 0 0 3
Circulatory problems 29 17 16 33 4 4 8 41
Nutrition (n ¼ 30)
No drinking/eating problemsb 8 22 15 37 2 0 2 39
Drinking/eating problems 22 19 16 35 4 4 8 43
Elimination (n ¼ 79)
No urination problemsb 10 4 8 12 0 0 0 12
No defecation problemsb 9 9 8 17 0 0 0 17
Gastrointestinal problems 3 9 7 16 4 4 8 16
Urination problems 34 18 15 33 4 4 8 41
Defecation problems 23 17 15 32 3 3 6 38
Skin (n ¼ 53)
No ulceration problemsb 5 4 2 6 0 0 0 6
Ulceration problems 48 14 14 28 4 4 8 36
Sleep (n ¼ 37)
No sleeping problemsb 12 2 5 7 0 0 0 7
Sleeping problems 25 16 17 33 4 4 8 41
Pain (n ¼ 63)
No pain problemsb 7 5 7 12 0 0 0 12
Pain problems r/t to amputation, 56 21 16 37 4 4 8 45
complications, other
Phantom pain documented 4 5 9 0 0 0 9
Pain alleviation (n ¼ 18)
No pain alleviation problemsb 5 4 5 9 0 0 0 9
Pain alleviation problems 13 21 16 37 4 4 8 45
Communication (n ¼ 27)
No communication problemsb 5 5 3 8 0 0 0 8
Communication problems 22 9 5 14 4 4 8 22
Well being (n ¼ 50)
Signs of well beingb 21 10 7 17 1 0 1 18
Signs of discomfort 29 9 8 17 4 4 8 25
Mental condition (n ¼ 42)
No mental problemsb 2 2 0 2 0 0 0 2
Mental problems 40 11 9 20 4 4 8 28

LLA, lower limb amputation; UGH, Uddevalla General Hospital; r/t, related to.
a
The same patient, both with and without recorded problems, can be the object of repeated observations during the hospital stay.
b
According to present physical/mental/functional status.

regarded as the explanation for this geographical variation

Discussion
(18). Internationally, the LLA rate is reported to vary sig-
nificantly between centres and different countries due to
Demographic data
differences in inclusions and exclusions of patients with
This study presents characteristics of LLA at UGH. When diabetes in the overall rates, the inclusion of first, ever or all
comparing these data with data from a comparable health amputations, the level of amputation and the use of dif-
care district, there was a considerable difference in ampu- ferent data sources to identify amputations (19).
tation rate, although both areas had lower rates than in
other comparable studies. A national study reports similar
Hospitalization, morbidity and mortality data
differences. In the western counties of Sweden, the annual
amputation rate varied between 12 and 71 per 100 000 Diabetes is a common cause of LLA. In Sweden,
inhabitants (17). Variation in clinical decision routines is the current prevalence of diabetes is 4% in the total

Ó 2005 The Authors, Scand J Caring Sci; 2005; 19, 337–343

342 S. Bäck-Pettersson, C. Björkelund

population. The number of patients with diabetes is pain was only recorded in nine of the nursing records.
expected to increase 10 times during the next 10 years Although general principles of pain management are
in the over-sixties population (20). The diabetes preval- applied at UGH, this did not seem to influence the
ence is reported to be dependent on the age-specific recording of pertinent patient data in the clinical nursing
incidence, the number of individuals at risk in the record. The patients surviving after 6 months had per-
background population and the age-specific mortality manent problems in the area of nutrition, elimination,
among patients with diabetes (21). The incidence of ulceration, sleep and pain alleviation. Patients deceased
major amputations has been reported to be more or less during the hospital stay had problems in all studied
constant over a 10-year period (22). However, the areas.
amputation incidence among diabetics tends to decrease, Nursing records revealed that LLA patients, in addition
irrespective of sex and age, in contrast to patients with to severe pain and markedly deteriorated general health
other smoking-related disease (23). Our data shows a status, had major problems in regaining physical and
significant difference in mean age between patients with mental functions and were thus in great need of profes-
and without diabetes. A similar difference in mean age sional nursing care. The majority experienced the pain as
has been reported by the SNBHW. unendurable and was in great pain during hospitalization
Smoking habits are presumed to influence mean age at despite high analgesic doses.
amputation in an aged population and should therefore be To be old and weak and to be subjected to a LLA seems
routinely documented in the medical records (24). In our to be associated with considerable suffering even if this was
data, the lack of information made it impossible to explore not evident or explicitly pointed out either in the medical
such associations. or nursing records. As the full extent of patients’ suffering
In-hospital mortality reached 17%. The mean hospital- was not clearly expressed in the nursing records, there is a
ization duration was 18 days for the studied population, risk that this patient group’s need of professional care will
and less for those who died within 6 months. This might be be inadequately prioritized. In times of increasing shorter
explained by a high-mortality rate within 10 days. The stays in hospitals, and decreasing resources in primary and
above-knee amputation rate was lower, compared to other municipal care, there is a risk of underestimating these
studies of aged amputees (2, 4). We did not find significant patients’ needs, rendering them especially exposed and
differences concerning above-knee amputations when vulnerable.
studying a comparable catchment area, but this could be Studying medical and nursing records provides valuable
due to a small sample size. However, our results are in information, provided that data is thoroughly documen-
accordance with other international studies. Medical ted. To further increase knowledge and understanding of
records showed that these patients had a long history of elderly amputees’ life situations, in-depth interviews might
severe chronic disease with a high frequency of hospital provide valuable data.
treatment and care episodes, and were thus major
consumers of care.
Acknowledgements
The authors would like to express their gratitude to Valter
Nursing data
Sundh for statistical assistance, to Lottis Eriksson and Claes
Nursing records revealed that LLA patients had several Rothelius for valuable clinical assistance, to the Swedish
severe pre- and postamputation problems that can be Centre of Epidemiology at the Swedish National Board of
assumed to cause great suffering. Health and Welfare for the provision of demographic data
Objective symptoms and problems were relatively and to Evelyn Hermansson for her valuable comments
clearly described, although nonsystematically presented. during the revision process.
Standard keywords were not used consistently through-
out the nursing records. According to the VIPS model,
Author contribution
the very idea of using the keyword ‘pain’ as the heading
is to facilitate documentation, i.e. the location of the Siv Bäck-Pettersson and Cecilia Björkelund designed the
pain, its character, pattern and intensity and the results study. Siv Bäck-Pettersson collected data from medical and
of pain assessments. In addition, the patient’s own nursing records. Siv Bäck-Pettersson and Cecilia Björkel-
comments about the pain and his/her ability to und analysed the data and wrote the paper.
manage and cope with it should be noted (25–27).
Phantom pain is described as a common phenomenon
Funding/sponsorship
among amputees during the immediate postoperative
period and persisting for up to 2 years. Phantom pain is This research was funded in part by a grant from the
reported as difficult to treat and is often unresponsive to Research and Development Council of Göteborg and
analgesia (28, 29). The patients’ suffering from phantom Southern Bohuslän.

Ó 2005 The Authors, Scand J Caring Sci; 2005; 19, 337–343

 Care of elderly lower limb amputees 343

16 Downe-Wamboldt B. Content analysis: method, applications

References
and issues. Health Care Women Int 1992; 13: 313–21.
1 Socialstyrelsen. Amputationer i Sverige 1987–1995 (The 17 The Westcoast Vascular Surgeons (WVS) Study Group. Var-
National Board of Health and Welfare. Amputations in Sweden iations of rates of vascular surgical procedures for chronic
1987–1995). The Swedish Centre for Epidemiology. Stock- critical limb ischaemia and lower limb amputation rates in
holm, 1997. western Swedish counties. Eur J Vasc Endovasc Surg 1997; 14,
2 Fletcher DD, Andrews KL, Butters MA, Jacobsen SJ, Row- 310–4.
land CM, Hallett JW Jr. Rehabilitation of the geriatric vas- 18 Connely J, Airey M, Chell S. Variation in clinical decision
cular amputee patient: a population-based study. Arch Phys making is a partial explanation for geographical variation in
Med Rehabil 2001; 82: 776–9. lower extremity amputation rates. Br J Surg 2001; 88: 529–
3 Pohjolainen T, Alaranta H. Lower limb amputations in 35.
Southern Finland 1984–1985. Prosthet Orthot Int 1988; 12: 19 The Global Lower Extremity Amputation Study Group. Epi-
9–18. demiology of lower extremity amputation in centers in
4 Adunsky A, Wershawski M, Arad M, Heruti R, Siev-Ner I, Europe, North America and East Asia. Br J Surg 2000; 87:
Heim M. Rehabilitation in practice. Non-traumatic lower 328–37.
limb older amputees: a database survey from geriatric centre. 20 Luft R. Diabetes. Underlag till vårdprogram för primärvården
Disabil Rehabil 2001; 23: 80–84. (Care Program for Primary Health Care) Socialstyrelsen,
5 Ebskov B, Ebskov L. Major lower limb amputation in diabetic Vårdprogramsnämnden. 1982, Utbildningsproduktion AB,
patients: development during 1982 to 1993. Diabetologia Malmö.
1996; 39: 1607–10. 21 The National Board of Health and Welfare. Medical Care in
6 Socialstyrelsen. SWEDVASC Svenska Karlregistret prel. årsrap- Sweden 1998. Stockholm, 1998.
port 2000 (The National Board of Health and Welfare. Vascular 22 Berger B, Stenström G, Sundqvist G. Incidence, prevalence,
registry in Sweden 2000). The Swedish Centre for Epidemiol- and mortality of diabetes in a large population: a report
ogy. Stockholm, 2001. from the Skaraborg Diabetes Registry. Diabetes Care 1999; 22:
7 Pohjolainen T, Alaranta H. Ten-year survival of Finnish 773–8.
lower limb amputees. Prosthet Orthot Int 1998; 22: 10–16. 23 Holstein P, Ellitsgaard N, Bornefeldt Olsen B, Ellitsgaard V.
8 Socialstyrelsen. Äldreuppdraget – slutrapport 2000 (The Decreasing incidence of major amputations in people with
National Board of Health and Welfare. Elder Care in Sweden). diabetes. Diabetologia 2000; 43: 0844–7.
Stockholm, 2000. 24 Liedberg E, Persson BM. Age, diabetes and smoking in lower
9 Buthler DJ, Turkal NW, Seidl JJ. Amputation: preoperative limb amputation for arterial occlusive disease. Acta Orthop
psychological preparation. J Am Fam Pract 1992; 5: 69–73. Scand 1983; 54: 383–8.
10 Drench ME. Changes in body image secondary to disease and 25 Ehnfors M, Thorell-Ekstrand I, Ehrenberg A. Towards basic
injury. Rehabil Nurs 1994; 19: 31–36. nursing information in patients’ records. Vård i Norden 1991;
11 Rudolphi D. Limb loss in elderly peripheral vascular disease 21: 12–31.
patient. J Vasc Nurs 1992; 10: 8–13. 26 Ehnfors M. Nursing care as documented in patients’ records.
12 SAS Institute Inc. SAS User’s Guide. Statistics, Version 5 Edition. Scand J Caring Sci 1996; 4: 209–20.
1985, SAS Institute Inc., Cary, NC. 27 Nordström G, Gardulf A. Nursing documentation in patients’
13 Krippendorrf K. Content Analysis: An Introduction to Its Meth- record. Scand J Caring Sci 1996; 1: 27–33.
odology, 2nd edn. 2004, Sage, Beverly Hills, CA. 28 Ellis K. A review of amputation, phantom pain and nursing
14 Weber RP. Basic Content Analysis. 1985, Sage, Beverly Hills, responsibilities. Br J Nurs 2002; 14: 155–63.
CA. 29 Gallagher P, Allen D, Maclachlan M. Phantom limb pain and
15 Ehrenberg A, Ehnfors M, Thorell-Ekstrand I. Nursing docu- residual limb pain following lower limb amputation: a des-
mentation in patient records: experiences of the use of the criptive analysis. Disabil Rehabil 2001; 23: 522–30.
VIPS model. J Adv Nurs 1996; 24: 853–67.

Ó 2005 The Authors, Scand J Caring Sci; 2005; 19, 337–343

 medical journal armed forces india xxx (xxxx) xxx

Available online at [Link]

ScienceDirect

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / m j a fi

Original Article

A study on quality of life among lower limb

amputees at a tertiary prosthetic rehabilitation
center

Pooja Shankar a, V.S. Grewal b,*, Sunil Agrawal c, Sreeni V. Nair d

a
ADH, HQ MB Area, Jabalpur Cantonment, 482 001, India
b
Associate Professor, Department of Community Medicine, Armed Forces Medical College, Pune, 411 040, India
c
Director (Health), O/o DGMS (Air), Air HQ, R K Puram, New Delhi, India
d
Resident, Department of Community Medicine, Armed Forces Medical College, Pune, 411 040, India

article info abstract

Article history: Background: Limb loss is one of the most physically and psychologically devastating events
Received 6 September 2018 causing major disfigurement, rendering people less mobile, and at risk for loss of inde-
Accepted 18 February 2019 pendence. This research aims to study the quality of life (QoL) among lower limb amputees
Available online xxx (LLAs) at a tertiary prosthetic rehabilitation center.
Methods: The present study is based on the self-perceived QoL of new LLAs reporting to a
Keywords: tertiary prosthetic rehabilitation center for prosthesis using World Health Organization
Lower limb amputees Quality of Life-BREF scale.
Quality of life Results: 35% of this study population were currently serving personnel of Indian Armed
Rehabilitation Forces, while 17.3% were dependents. 17.3% of the study participants were retired Indian
Diabetic complications Armed Forces personnel. 51.5% of study participants had undergone a lower limb ampu-
Psychological health tation following trauma; 22.7% of the traumatic amputations had resulted from road traffic
accidents, 16.7% from trauma sustained following railway accidents, and combat-related
injuries accounted for 11.4% of traumatic lower limb amputations. Diabetic foot/gangrene
was responsible for 32% of lower limb amputations in the current study. Overall QoL score
of 78.76 (standard deviation [SD] ¼ 15.03) was seen revealing highest mean score in envi-
ronmental domain 26.59 (SD ¼ 5.38), followed by physical domain, psychological domain,
and lowest score being in social domain.
Conclusion: When judging the success or failure of lower limb amputation, the assessment
of QoL outcome is paramount. A number of factors need to be taken into consideration to
ensure holistic reintegration of the amputees back into the society.
© 2019, Armed Forces Medical Services (AFMS). All rights reserved.

* Corresponding author.
E-mail address: drvikramgrewal@[Link] (V.S. Grewal).
[Link]
0377-1237/© 2019, Armed Forces Medical Services (AFMS). All rights reserved.

Please cite this article as: Shankar P et al., A study on quality of life among lower limb amputees at a tertiary prosthetic rehabilitation
center, Medical Journal Armed Forces India, [Link]
2 medical journal armed forces india xxx (xxxx) xxx

rehabilitation center for refitting or change of old prosthesis

Introduction and individuals with congenital amputation or amputation of
digits of the lower limb only were excluded from the study.
The history of amputation may be traced back to the Neolithic To estimate impact of amputation on QoL among LLAs
era. In modern medicine, amputation is resorted to as the last reporting to the tertiary prosthetic rehabilitation center, using
measure when saving a limb becomes a near-impossible WHOQOL-BREF scale, under various domains, with 95% con-
scenario.1 Limb loss, in addition to causing disfigurement, fidence interval and assumed error of margin with standard
renders people less mobile and at risk of loss of independence. deviation (SD) from previous literature enlisted in Table 1,7 the
Loss of a limb has major economic, social, and psychological minimum required sample size was calculated as 141; how-
effects, both on the individual as well on his/her social ever, to achieve the objectives of the study, 150 participants
domain.2 were taken into consideration for this project.
In developing countries like ours, causes of amputation is The WHOQOL-BREF (field trial version) self-administered
quite often influenced by degree of industrialization, trans- questionnaire was used as the data collection tool. The
portation system, and safety and access to medical care. A WHO has developed the WHOQOL-BREF scale, which captures
study conducted in 2009 reported trauma as the cause for many subjective aspects of QoL. WHOQOL-BREF is one of the
majority of amputations in India. Various other studies, con- best known instruments that has been developed for cross-
ducted at Chennai and Kolkata, have also cited road traffic cultural comparisons of QoL and is available in more than 40
accident (RTA) as the prime cause of amputation in the Indian languages, including Hindi.8
subcontinent.3 The study participants were administered the question-
Previous literature indicate that lower extremity injuries naire in their language of preference (English/Hindi), with the
account for almost 50% of combat wounds among US military interviewer thoroughly explaining to them the purpose of the
personnel deployed in Iraq and Afghanistan. In conflict sce- study and approach to selecting the response they felt was
nario, amputation often ends up being the only treatment most applicable to them. Written informed consent was ob-
option available.4 Mine blast injuries leads the way as the tained from each participant, and ethical clearance was ob-
predominant cause among cases due to trauma. Other tained from the institutional ethical committee prior to the
important causes of battle field injuries are those resulting conduct of the study.
from gunshots and shrapnel. RTA is the second most common The field-trial version of the WHOQOL-BREF contains 26
cause of disability. Railway accidents also constitute signifi- items covering four different domains of QoL
cant number of cases. Cold injuries also contributed for a large
number of cases, because of stationing of troops at high alti-  Physical health (7 items)
tude and snow covered areas.5  Psychological health (6 items)
The World Health Organization Quality of Life (WHOQoL)  Social relationships (3 items)
Group defines quality of life (QoL) as an individual's perception  Environmental health (8 items)
of his/her position in circumstance of the culture and values
in which he or she lives and with respect to his/her goals, Out of 26 items, 02 items are examined separately, one
expectations, principles, and concerns.6 Despite growing addressing an individual's overall perception of his/her QoL
popularity of QoL and amputee studies worldwide, it remains and the other regarding perception of his/her own health. The
a relatively poorly studied concept in developing countries domain scores are scaled in a positive direction, and the mean
such as India. The aim of this study was to assess the score of items within each domain is used to calculate the
QoL among lower limb amputees (LLAs) at a tertiary prosthetic domain score. The raw scores are converted into transformed
rehabilitation center. The key objectives of the study were to scores using an SPSS syntax file, the latter having being ob-
study the QoL among LLAs under various domains of the tained by the authors from WHO designated researcher. Data
WHOQOL-BREF scale to determine sociodemographic factors on sociodemographic factors were collected using a pilot-
associated with QoL among LLAs and to suggest suitable rec- tested validated questionnaire. The data were analyzed using
ommendations for improving QoL in LLAs found in the study. the SPSS (version 20). Normality test (Kolmogorov-Smirnov
Test) was applied to test domains of QoL and the subgroups
wherever required. Homogeneity test was applied to test
Materials and methods

This was a descriptive, tertiary-care centerebased cross- Table 1 e Sample size calculation.
sectional study carried out at a prosthetic rehabilitation cen- [Link]. Domains Error of Standard Sample
ter located at Pune. All fresh/new LLAs, both men and women, margin/ deviation (s) size
18 years of age and above, irrespective of the level of ampu- precision
tation, and reporting to the tertiary prosthetic rehabilitation 1. Q1-QOL 15% 0.85 123
center, were included in the study. In the current study, fresh/ 2. Q2-Health 15% 0.91 141
new LLAs were defined as all individuals who had undergone a 3. Physical 5% 18.70 54
lower limb amputation (unilateral/bilateral) in the preceding 6 4. Psychological 5% 17.93 49
5. Social 5% 19.09 56
months from the date of conduct of the study and who had
6. Environmental 5% 15.19 35
never got a trial of or been fitted with a lower limb prosthesis.
Old amputees reporting to the tertiary prosthetic QOL, quality of life.

Please cite this article as: Shankar P et al., A study on quality of life among lower limb amputees at a tertiary prosthetic rehabilitation
center, Medical Journal Armed Forces India, [Link]
 medical journal armed forces india xxx (xxxx) xxx 3

whether subgroups were homogenous. Kruskal-Wallis test In the current study, 46% of the study participants had
was applied wherever the normality or homogeneity tests undergone the amputation at a civil hospital, 40.6% amputa-
failed. Analysis of variance was applied for the groups satis- tions had been performed at a military hospital, and 13.4% of
fying the normality and homogeneity tests. Students unpaired the amputations had been performed at an Ex-servicemen
t-test was applied wherever two groups were there. Contributory Health Scheme empanelled hospital.
In the current study, trauma was the most common cause
of amputation, among which RTAs were the major cause of
Results lower limb amputation among the study participants at 34
(22.7%), followed by trauma sustained because of railway ac-
The various sociodemographic characteristics of the study cidents (fall from moving train on attempting to alight from or
participants are enlisted in Table 2. get into a moving train) at 25 (16.7%). The second most com-
mon cause of lower limb amputation was diabetic gangrene/
diabetic foot, with 48 (32%) having cited the same. Seventeen
(11.4%) of the study participants had undergone a lower limb
Table 2 e Sociodemographic and clinical characteristics
amputation following injury sustained during combat opera-
of study participants.
tions (splinter injury/mine blast injury/firearm injury).
[Link]. Characteristic Frequency Percentage
Vascular causes contributed to lower limb amputation among
(N ¼ 150) (%)
12 (8%) of the study participants, and malignancy was the
1. Gender cause of amputation in 8 (5.4%) of them. Five (3.5%) of the
a. Male 110 73.3
study participants cited other reasons for having undergone a
b. Female 40 26.7
lower limb amputation. (Graph 1)
2. Age (yrs)
a. 20e39 58 38.7 Among the study participants, 83 (55.3%) had no other
b. 40e59 40 26.7 morbidities co-existing along with the lower limb amputation.
c. 60e79 48 32.0 Forty-three (28.7%) of the participants had preexisting dia-
d. 80e99 04 2.7 betes mellitus type II, while 10 (6.7%) had both diabetes mel-
3. Educational status litus type II and primary hypertension. Five (3.3%) of the
a. None at all 18 12.0
participants had preexisting primary hypertension only, while
b. Primary school 14 9.3
c. Middle school 35 23.3
four (2.7%) had more than 2 co-existing diseases, namely
d. High school 64 42.7 diabetes mellitus type II, ischemic heart disease, and primary
e. Beyond high school 19 12.7 hypertension.
4. Marital status The study participants had an overall QoL score of 78.76
a. Unmarried 33 22.0 (SD ¼ 15.03). The current study revealed that highest mean
b. Married 91 60.7
score was in environmental domain with mean QoL score of
c. Widow/widower 22 14.7
26.59 (SD ¼ 5.38), followed by physical domain with mean
d. Separated 01 0.7
e. Divorced 03 2.0 score of 19.71 (S.D ¼ 3.30), psychological domain with mean
5. Serving/retired/civilian score of 16.86 (S.D ¼ 4.55), and lowest score being in social
a. Serving 53 35.3 domain with mean score of 9.61 (S.D ¼ 3.25). In the current
b. Serving (dependents) 19 12.7 study, the transformed domain scores were found to be
c. Retired 26 17.3 highest in the social domain (mean 71.66, SD ¼ 20.50), and
d. Retired (dependents) 10 6.7
lowest score was found to be in the physical domain (mean ¼
e. Civilian 42 28.0
48.92, SD ¼ 16.45).
6. Employment status
a. Working 72 48
b. Pensioner 29 19.3
c. Unemployed 02 1.3
d. Dependent wholly on family 47 31.3
members financially
7. Socio-economic status
a. Upper class 107 71.3
b. Upper middle class 34 22.7
c. Middle class 08 5.3
d. Lower middle class 01 0.7
e. Lower class 0 0
8. Time elapsed since amputation
a. 1e3 months 85 56.7
b. 4e6 months 65 43.3
9. Side of amputation
a. Right 67 44.7
b. Left 52 34.7
c. Bilateral 31 20.6
10. Level of amputation
a. Above knee 60 40.1 Graph 1 e Distribution of study participants as per cause of
b. Below knee 90 60
lower limb amputation.

Please cite this article as: Shankar P et al., A study on quality of life among lower limb amputees at a tertiary prosthetic rehabilitation
center, Medical Journal Armed Forces India, [Link]
4 medical journal armed forces india xxx (xxxx) xxx

As far as level of education of the study participants was

Table 3 e Mean score of different QoL domains among
concerned, a statistically significant association among
study participants.
different education groups was found only under the social
Category Mean domain Transformed
domain (p value ¼ 0.0001) (Table 4).
score domain score
The overall QoL score was highest for serving personnel at
(on scale of 100)
81.08 (SD ¼ 13.96), followed by retired personnel at 76.69
Physical domain 19.71 48.92
(SD ¼ 17.64) and civilians at 76.54 (SD ¼ 14.14). No statistically
Psychological domain 16.86 53.35
Social domain 9.61 71.66
significant association was found among different employ-
Environmental domain 26.59 58.10 ment status groups under any of the QoL domains (p < 0.05).
Overall QoL score of 78.77 (S.D ¼ 16.50) was found among
QOL, quality of life.
study participants who had undergone a lower limb amputa-
tion within preceding 1e3 months time period, while the score
In the current study, the transformed domain scores were was 78.73 (S.D ¼ 12.99) for those within preceding 4e6 months
found to be highest in the social domain (mean ¼ 71.66, SD ¼ time frame. A statistically significant association between
20.50), and lowest score was found to be in the physical different time periods elapsed because amputation was found
domain (mean ¼ 48.92, SD ¼ 16.45). These findings are similar under social domain and overall QoL (p value ¼ 0.028) (Table 4).
to those of a study conducted by Adegoke et al., where LLAs
had highest score in the social domain, followed by environ-
mental domain and psychological domain and lowest in
physical domain (mean ¼ 52.38; SD ¼ 22.29).9
Discussion
Overall QoL score of 80.87 (SD ¼ 15.35) was observed among
Trauma following RTAs accounted for 22.7% of the lower limb
the male study participants, while it was 72.95 (S.D ¼ 12.54) in
amputations in the current study, followed by trauma sus-
females. Higher scores were observed among males under all
tained as a result of railway accidents (fall from moving train
domains compared with that in females. A statistically sig-
on attempting to alight from or get into a moving train) at
nificant association between both gender groups was
16.7%.The findings of the present study are consistent with
observed only under the environmental domain (p < 0.05).
other similar studies depicting that trauma is the leading
Highest QoL score was found in the age group of 40e59
cause of lower limb amputation followed by diabetes or its
years, with a significant score within the social domain (p
related complications.10e12
value ¼ 0.03) (Table 4). Higher overall QoL scores were
The current study showed a statistically significant asso-
observed among males under all domains compared with fe-
ciation between the various age groups and QoL score within
males. A statistically significant association between both
the social domain only (Table 4). The overall QoL scores as well
gender groups was observed only under the environmental
as individual domain scores was found to be better in age
domain (p value ¼ 0.03) (Table 4).

Table 4 e Association between various sociodemographic factors and QoL scores.

Sociodemographic factor QoL scores within significant domain Domain showing significant score p-value
(Mean ± SD)
Age (yrs)
20e39 10.50 ± 2.59 Social 0.03
40e59 10.82 ± 2.91
60e79 7.85 ± 3.41
80e99 5.75 ± 1.25
Gender
Male 27.38 ± 5.48 Environmental 0.03
Female 24.42 ± 4.47
Level of education
None 6.94 ± 2.07
Primary school 7.71 ± 2.92 Social 0.0001
Middle school 9.14 ± 3.97
High school 10.54 ± 2.75
Beyond high school 11.26 ± 2.18
Socio-economic class
Upper class 17.18 ± 3.70
Upper middle class 15.85 ± 3.53 Psychological 0.0001
Middle class 17.75 ± 12.66
Lower middle class 10.00
Time elapsed since amputation
1e3 months 9.44 ± 3.49
4e6 months 9.83 ± 2.92 Social 0.028

QOL, quality of life; SD, standard deviation.

Please cite this article as: Shankar P et al., A study on quality of life among lower limb amputees at a tertiary prosthetic rehabilitation
center, Medical Journal Armed Forces India, [Link]
 medical journal armed forces india xxx (xxxx) xxx 5

groups 20e39 years and 40e59 years and lower in those >60 by environmental domain (mean ¼ 60.40, SD ¼ 18.84), psy-
years of age. In a study conducted in Pakistan, individual QoL chological domain (mean ¼ 55.17, SD ¼ 15.27) and lowest in
domain and overall QoL scores were found to be higher in physical domain (mean ¼ 52.38, SD ¼ 22.29).9
groups aged 20e40 years and 41e60 years; however, no sta- Our study findings are suggestive of better adaptability of
tistically significant association was found within any of the LLAs, with education till the level of high school or beyond, to
four QoL domains and different age groups.13 According to postamputation body changes and enhanced social func-
findings of another study, lower QoL scores in older age groups tioning and fair knowledge regarding their health status and
was attributed to the possibility that these people seek com- attitude toward accessing existing healthcare services. Study
fort and maintained social standing by prioritizing personal participants with higher education might feel more socially
relationships rather than their level of physical functioning equipped and find it easier to deal with cognitive, emotional,
because of age-related musculoskeletal disorders and re- and social aspects of amputation.16,17
strictions of physical activity thereof.14 Employment status is an important determinant of QoL,
The current study found that majority of LLAs comprised of especially among the male population. In our country, it is the
males (73.3%), while females consisted of 26.7% of the study men who are primary earning members of the family and
population. The preponderance of male amputees compared their status of unemployment has a direct impact on standard
with female amputees may be because of the fact that it is of living as well as sense of self-worth.
mostly the male members of a household who venture out for Overall QoL score was highest for those study participants
purpose of work or undertake travel for the same; they are, belonging to upper socio-economic class, while score for the
hence, more at risk of road traffic and industrial accidents and lower middle class was lowest. A statistically significant as-
amputation thereof. Also, majority of the serving military sociation was found between the different socio-economic
personnel in our study belonged to the non-officer ranks-a classes within the psychological domain only. This finding
rank structure into which ladies are not eligible to be enrolled. may be attributed to the fact that greater financial stability
This might be another reason why the frequency of lower limb ensures access to high quality curative and rehabilitative ser-
amputations in males exceeded that of females amongst our vices, carefree access to essential items of daily living, both for
study population. A study conducted by Sinha et al. at a limb- self and family, and a sense of general well-beingdfactors that
fitting and rehabilitation center based in Mumbai, India, also have a direct impact on an individual's perception of QoL.
found that 88% of the respondents were males while females Studies assessing the association between lower limb ampu-
comprised of only 12% of the respondents.11 The Global Lower tations, socio-economic status, and QoL, especially at the level
Extremity Amputation Study states that the incidence of of tertiary health care service, are few.18e20
lower limb amputation is similar in both gender in some re- There are very few QoL studies that have considered time
gions and higher in females than males in other regions, since amputation in terms of months and its association with
although the overall incidence is higher in men than overall QoL score or individual domain scores. During the
women.15 In a study conducted by Adegoke et al., male par- initial few weeks immediately following amputation, patients
ticipants scored higher compared with the female partici- are admitted to the healthcare facility with medical and para-
pants in overall QoL score; highest and lowest domain scores medical staff as attendees 247. The ward or hospital they are
were found within the social and environmental domains.9 admitted to might have other patients who have undergone an
In the current study, 46% of the current study participants amputation, leaving little scope for loneliness or feeling of
had undergone the amputation at a civil hospital, 40.6% am- isolation. However, once the postamputation wound heals and
putations had been performed at a military hospital, and patients are transferred to the rehabilitation center, limb
13.4% of the amputations had been performed at an Ex- measurements are taken, and trial of prostheses is initiated.
servicemen Contributory Health Scheme empanelled hospi- The presence of other LLAs at the rehabilitation center could
tal. No studies quoting similar findings were found during have either a positive or a negative psychological impact on the
review of literature for the same. The current study had 35.3% new/fresh amputee. The amputees who have already been
of study respondents belonging to the Indian Armed Forces. using a prostheses for a considerable period of time could act as
These personnel, during period of leave, may report to nearest a source of encouragement and inspiration to the new ampu-
civil service health facility in case of medical or surgical tees; on the contrary, the new amputees, during this period,
emergencies. In case of any traumatic event, including road might feel dejected by constantly comparing their own level of
traffic/railway accidents, these individuals may have to un- mobility achieved (by use of wheelchair or crutches) with that
dergo major/minor surgeries, amputations included, at the of the older amputees who have already received a prostheses.
civil health facility. They are later transferred to an Armed This might be a possible explanation for the marginally higher
Forces hospital once their condition is stabilized. Hence, it is overall QoL score among the respondents who had undergone
not only the 28% civilian population in this study that have the amputation 1e3 months back.
undergone lower limb amputation surgeries at civil health
facilities; some of the Armed Forces personnel might have
also undergone initial treatment in a civil set up. Conclusion
The current study showed a statistically significant asso-
ciation among different levels of education within the social As has been demonstrated in our study, the overall trans-
domain of QoL. These findings are consistent with that of a formed QoL domain score was lowest in the physical domain.
study conducted by Adegoke et al., where LLAs had highest Amputees have to adapt physically, socially, and psycholog-
score in the social domain (mean ¼ 63.23, SD ¼ 23.01), followed ically to the change in structure, functioning, and appearance

Please cite this article as: Shankar P et al., A study on quality of life among lower limb amputees at a tertiary prosthetic rehabilitation
center, Medical Journal Armed Forces India, [Link]
6 medical journal armed forces india xxx (xxxx) xxx

of their body. Effective preamputation psychological coun- 4. Melcer T, Walker J, Bhatnagar V, Richard E, Sechriest II VF,
selling may prove useful in ameliorating the psychological Galarneau M. A comparison of four-year health outcomes
distress of amputees and their anxiety related to the proced- following combat amputation and limb salvage. PLoS One.
2017 Jan 25;12(1):e0170569.
ure and life thereafter.
5. Sinha R, Van Den Heuvel WJ. A systematic literature review of
Trauma was found to be the dominant cause of amputa- quality of life in lower limb amputees. Disabil Rehabil. 2011 Jan
tion in this study (51.5%), followed by diabetic complications 1;33(11):883e899.
(32%). Trauma following RTAs accounted for 22.7% of the 6. Misselbrook DW. Is for Wellbeing and the WHO definition of
lower limb amputations in the current study, followed by health. Br J Gen Pract. 2014 Nov 1;64(628):582.
trauma sustained as a result of railway accidents (fall from 7. Knez  A, Salamon T, Milankov M, Ninkovic
evic  S, Jeremic
-
Knez  M, Toma
evic -Todorovic
sevic  S. Assessment of quality of
moving train on attempting to alight from or get into a moving
life in patients after lower limb amputation. Med Pregl.
train) at 16.7%. Human factor is a significant determinant in
2015;68(3-4):103e108.
occurrence of these injuries. Serving soldiers more often meet 8. Vahedi Shahrum. World Health Organization Quality-of-Life
with road traffic and railway accidents while on leave, often Scale(WHOQOL-BREF): analyses of their item response theory
being involved in reckless driving and absolute disregard to properties based on the graded responses model. Iran J
traffic rules. Commanders at all levels, unit Regimental Med- Psychiatr. 2010;5(4):140.
ical Officers (RMOs) and junior level leaders should repeatedly 9. Adegoke BO, Kehinde AO, Akosile CO, Oyeyemi AL. Quality of
life of Nigerians with unilateral lower limb amputation.
reinforce road safety rules and rail travel safety rules during
Disability. CBR Incl Dev. 2013 Feb 5;23(4):76e89.
Sainik Sammelans, daily evening roll-calls, health lectures,
10. Maqsood M, Ali N, Bhat A, Bangroo FA, Dhanda MS, Singh R.
etc. Current trends of major lower limb amputations at a tertiary
Considering diabetic complications were the second care centre of Jammu, India. Int J Med Sci Res Pract. 2015 Jun
leading cause of lower limb amputation in this study, it be- 30;2(2):77e80.
comes imperative that LLAs with co-morbid conditions need 11. Sinha R, van den Heuvel WJ, Arokiasamy P. Factors affecting
to have a holistic rehabilitation program which should quality of life in lower limb amputees. Prosthet Orthot Int. 2011
Mar 1;35(1):90e96.
include health education with respect to dietary modifica-
12. Ministry of Statistics & Programme Implementation
tions, daily exercise program, harmful effects of addictions Government of India New Delhi. SARVEKSHANA 94th Issuevol.
(tobacco and alcohol consumption), and effects of uncon- XXVIII. New Delhi: Controller of Publications, Department of
trolled blood sugar levels and blood pressure on overall Publication, Civil Lines, Delhi; 2008:30e55, 3&4 (December 2008).
health of the individual as well as long-term outcomes of 13. Malik M, Bilal F, Ali Khan MS, Jabeen F, Fatima Dogar S,
amputated as well as normal limb. Munir N. Quality of life and its relationship with demographic
Results of this study reinforce the fact that amputation variables among physically disabled patients with artificial
limb. Rawal Med J. 2013;38(2):134e138.
continues to be associated with lower QoL scores. It is rec-
14. Deans SA, McFadyen AK, Rowe PJ. Physical activity and
ommended that the participants receive a structured reha- quality of life: a study of a lower-limb amputee population.
bilitation curriculum which is adapted to the specific needs of Prosthet Orthot Int. 2008 Jun;32(2):186e200.
people with lower limb amputation. Prospective longitudinal 15. Godlwana L, Stewart A, Musenga E. Quality of life following a
studies are recommended to study the change in QoL over major lower limb amputation in Johann esburg, South Africa.
time and to assess its determinants. Table 3). S Afr J Physiother. 2012 Nov 12;68(2):17e22.
16. Dajpratham P, Tantiniramai S, Lukkanapichonchut P. Health
related quality of life among the Thai people with unilateral
lower limb amputation. J Med Assoc Thail. 2011 Feb 9;94(2):250.
Conflicts of interest 17. Coffey L, Gallagher P, Desmond D. Goal pursuit and goal
adjustment as predictors of disability and quality of life
The authors have none to declare. among individuals with a lower limb amputation: a
prospective study. Arch Phys Med Rehabil. 2014 Feb
28;95(2):244e252.
references 18. Washington ED, Williams AE. An exploratory
phenomenological study exploring the experiences of people
with systemic disease who have undergone lower limb
amputation and its impact on their psychological well-being.
1. Ubayawansa DH, Abeysekera WY, Kumara MM. Major lower
Prosthet Orthot Int. 2016 Feb;40(1):44e50.
limb amputations: experience of a tertiary care hospital in Sri
19. Ejtahed HS, Soroush MR, Hasani-Ranjbar S, et al. Prevalence
Lanka. J Coll Physicians Surg Pak JCPSP. 2016 Jul 1;26(7):620e622.
of metabolic syndrome and health-related quality of life in
2. Kurichi JE, Ripley DC, Xie D, Kwong PL, Bates BE,
war-related bilateral lower limb amputees. J Diabetes Metab
Stineman MG. Factors associated with home discharge after
Disord. 2017 Apr 5;16(1):17.
rehabilitation among male veterans with lower extremity
20. Amtmann D, Morgan SJ, Kim J, Hafner BJ. Health-related
amputation. PM&R. 2013 May 31;5(5):408e417.
profiles of people with lower limb loss. Arch Phys Med Rehabil.
3. Pooja GD, Sangeeta L. Prevalence and aetiology of amputation
2015 Aug 31;96(8):1474e1483.
in Kolkata, India: a retrospective analysis. Hong Kong
Physiother J. 2013 Jun 1;31(1):36e40.

Please cite this article as: Shankar P et al., A study on quality of life among lower limb amputees at a tertiary prosthetic rehabilitation
center, Medical Journal Armed Forces India, [Link]
Eur J Orthop Surg Traumatol
DOI 10.1007/s00590-015-1709-z

ORIGINAL ARTICLE • GENARAL ORTHOPAEDICS - LOWER LIMB

The effect of amputation level and age on outcome: an analysis

of 135 amputees
Murat Yilmaz1 • Deniz Gulabi2 • Ibrahim Kaya1 • Erhan Bayram1 •

Gultekin Sitki Cecen2

Received: 17 August 2015 / Accepted: 22 September 2015

 Springer-Verlag France 2015

Abstract transfemoral amputation group were determined as signif-

Introduction In this retrospective study, the impact of icantly high compared to those of the transtibial group
age, amputation level and the cause of amputation were (p \ 0.05). The K index of the group aged 35 years and
examined using the Hospital Anxiety and Depression Scale below was significantly higher than that of the other groups
(HADS) and Medicare K scores of amputees with unilateral (p 0.002, p 0.001, p \ 0.01).
lower-limb amputation. Conclusion The data show higher HADS-A scores with
Materials and methods In total, 135 patients with uni- traumatic transfemoral amputation. Therefore, adequate
lateral transfemoral (TF) or (TT) transtibial amputations psychiatric evaluation and rehabilitation should be applied
were examined. All data were collected using question- to all amputees, especially in cases of young, traumatic,
naires that were either self-administered or administered transfemoral amputations.
during an interview. The HADS was developed as a self- Level of Evidence Level 3, retrospective comparative
reporting questionnaire to detect adverse anxiety and cohort study.
depressive status. K code is used to describe the functional
abilities of amputees. Keywords Amputation
HADS-A
HADS-D
K index
Results The mean age at the time of surgery was
52.79 ± 13.08 years. The mean time since amputation was
59.20 ± 24.41 months for TT, and 60.89 ± 22.09 months Introduction
for TF amputation. The HADS-A scores of the
Lower-limb amputation is performed for a variety of rea-
sons such as peripheral vascular diseases, trauma, tumor or
& Deniz Gulabi congenital anomalies [6, 10]. Amputation causes disability
dgulabi@[Link] and results in economic, psychological and social effects.
Murat Yilmaz There have been several studies on clinical depression and
muratyilmazctfhsk@[Link] functional outcomes in persons with limb loss. Cansever
Ibrahim Kaya et al. [4] reported that patients with limb loss encounter a
drikaya27@[Link] series of complex psychological responses.
Erhan Bayram There are multiple tests to analyze depression in patients
bayerhan@[Link] with limb loss. Some authors have preferred to use a clinician-
Gultekin Sitki Cecen administered structured interview, whereas others have pre-
gcecen2002@[Link] ferred standardized self-reporting measures such as the Beck
1 Depression Scale [1], the Center for Epidemiologic Study
Haseki Training and Research Hospital, Fatih,
34096 Istanbul, Turkey Depression Scale [21] or the Hospital Anxiety and Depression
2 Scale (HADS). All these scales are used for the early diagnosis
Dr. Lütfi Kırdar Kartal Training and Research Hospital,
Semsi Denizer Cad. E5. Yanyol Cevizli Kavsagi, Kartal, and treatment of psychological morbidity of patients with limb
34890 Istanbul, Turkey loss, so that long-term disabilities in an amputee can be

123
 Eur J Orthop Surg Traumatol

prevented [25]. The aim of this retrospective study was to K level 0: Does not have the ability to ambulate or transfer
evaluate the impact of age, amputation level and the cause of safely with or without assistance, and a prosthesis does not
amputation that were examined using the HADS [9] and enhance quality of life or mobility.
K scores of amputees with unilateral lower-limb amputation K level 1: Has the ability to use a prosthesis for transfers or
[27]. It was hypothesized that a higher amputation level and ambulation on level surfaces at a fixed cadence. Typical of
older age would be significant predictors for worse HADS-A, the limited and unlimited household ambulator.
HADS-D and K level scores. K level 2: Has the ability for ambulation with the ability
to transverse low-level environmental barriers such as
stairs or uneven surfaces. Typical of the limited com-
Materials and methods munity ambulator.
K level 3: Has the ability for ambulation with variable
This hospital-based study comprised 205 amputees who cadence. Typical of the community ambulator who has
underwent amputation between 2005 and 2011. Exclusions the ability to transverse most environmental barriers and
from the study were 6 amputees with knee disarticulation, 10 may have vocational, therapeutic or exercise activity that
amputees of syme or foot amputations, 11 amputees who died, demands prosthetic use beyond simple locomotion.
8 bilateral amputations, 12 amputees who did not want to K level 4: Has the ability for prosthetic ambulation that
participate in the study and 23 amputees with prior psychiatric exceeds basic ambulation skills, exhibiting high impact,
disorders and associated physical disabilities other than stress or energy levels. Typical of the prosthetic
amputation. Thus, the study was limited to a total of 135 demands of the child, active adult or athlete. K level 0
patients with unilateral transfemoral (TF) or transtibial (TT) was defined as poor, K levels 1 and 2 were defined as fair
amputations. The patients were 111 (82.2 %) males and 24 and K levels 3 and 4 were defined as good.
(17.8 %) females with a mean age at the time of surgery of
The study group was separated into two groups of TT and
52.79 ± 13.08 years (range 21–82 years). The inclusion cri-
TF amputees. Transfemoral amputation was made at mean
teria were age of 18 years or older, unilateral TT or TF
12 cm distal from the trochanter major, and transtibial
amputation from any cause, and sufficient cognitive skills to
amputation was accepted as mean 15 cm distal from the
accurately respond to the questionnaires. All data were col-
level of the knee joint. Reasons for amputation were
lected using questionnaires that were either self-administered
determined as trauma and vascular disease (DM, peripheral
or administered during an interview depending on the partici-
vascular disease). Amputations made for neoplastic or
pant’s ability to answer (e.g., no visual problems or difficulty in
congenital reasons were excluded from the study. The
understanding certain questions). Interviews were conducted
HADS and Medicare K levels were completed by a physical
by the same evaluator, a senior physical therapist with over
therapist blind to the study for amputees in both groups.
5 years of experience. The following information was extrac-
ted from the patient’s medical chart: age, gender, side, level and
Statistical evaluation
date of amputation, the cause of amputation, duration of pros-
thetic usage and associated medical conditions.
For the statistical analysis of data obtained in the study,
The Hospital Anxiety and Depression Scale (HADS)
SPSS (statistical package for social sciences) for Windows
was developed as a self-reporting questionnaire to detect
15.0 software program was used. In the evaluation of the
adverse anxiety and depressive status. It has good sensi-
study data, descriptive statistical methods (mean, standard
tivity and specificity for formal psychiatric assessment and
deviation) were used, and in the evaluation of age with
diagnosis. Participants were asked to choose one of the four
HADS-A, HADS-D and the K index, the Kruskal–Wallis
given choices. The questions relating to anxiety are marked
test was used. The Mann–Whitney test was applied in the
‘A’ and those relating to depression, ‘D’. There are 14
evaluation of amputation level with HADS-A, HADS-D
items in the scale: 7 related to anxiety and 7 to depression.
and the K index. Values of p \ 0.05 and p \ 0.01 were
Scoring is from 0 to 3, and total scores may range from 0 to
accepted as statistically significant.
21. According to the final scores at the interview, HADS-A
and HADS-D are classified into three groups. A score of
0–7 is normal, 8–10 borderline and 11–21 abnormal [8].
Medicare K code is used to describe the functional Results
abilities of persons who have undergone lower-limb
amputation. Using this system, the physical therapist Of the 205 amputees with lower-limb amputation admitted to
determines the patient’s ability to reach a ‘‘defined func- the study group, only 135 (66 %) met the eligibility criteria.
tional state within a reasonable period of time.’’ It has 5 These comprised 24 females and 111 males. The reasons for
levels defined as K levels 0–4. amputation were trauma in 58 cases (43 %) and vascular

123
Eur J Orthop Surg Traumatol

disease in 77 (57 %). In 65 cases (48.1 %), the amputation was groups (p [ 0.05). A statistically significant difference was
TT and in 70 cases (51.9 %), TF. The amputation was on the determined in the K index according to age groups
right side in 82 cases (60.7 %) and on the left in 53 cases (p \ 0.01). The result of the paired comparisons that were
(39.3 %). The mean time since amputation was made to determine in which group the difference originated
59.20 ± 24.41 months (range 24–110 months) for transtibial, was that the K index of the group aged \35 years was sig-
and 60.89 ± 22.09 months (range 24– nificantly higher than that of the 36- to 50-year age group
110 months) for transfemoral. The mean duration of prosthetic (p 0.002, p \ 0.01) and of the over-50-year age group
usage was 54.77 ± 24.23 months (range 18– (p 0.001, p \ 0.01). The K index of the 36- to 50-year age
105 months) for the TT group and 56.06 ± 21.89 months group was statistically significantly higher than that of the
(range 20–102 months) for the TF group. The demographic over-50-year age group (p 0.001, p \ 0.01).
properties of the participants are summarized in Table 1.
No statistically significant difference was determined
between the two groups in respect of age, gender, reason Discussion
for amputation, side, HADS-D and K level (p [ 0.05). The
HADS-A scores of the TF group were determined as sta- Lower-limb amputation is a permanent surgical procedure
tistically significantly high compared to those of the TT that has important functional, psychological and social
group (p \ 0.05) (Table 2). sequelae [26]. The present study was based on hospital data
No statistically significant difference was determined records and interview at the outpatient polyclinics. The
between the HADS-A and HADS-D scores according to age most important findings of the present study were that the

Table 1 Distribution and

Transtibial Transfemoral p
evaluation of general
characteristics Age 52.37 ± 13.71 (21–80) 53.17 ± 12.57 (23–82) 0.723a
Follow-up period (months) 59.20 ± 24.41 (24–110) 60.89 ± 22.09 (24–110) 0.674a
Duration of prothesis use (months) 54.77 ± 24.23 (18–105) 56.06 ± 21.89 (20–102) 0.746a
Gender
Female 10 (15.4 %) 14 (20 %) 0.63b
Male 55 (84.6 %) 56 (80 %)
Amputation
Traumatic 23 (35.4 %) 35 (50 %) 0.087c
Vascular 42 (64.6 %) 35 (50 %)
Side
Right 37 (56.9 %) 45 (64.3 %) 0.381c
Left 28 (43.1 %) 25 (35.7 %)
HADS-A 7.77 ± 3.16 (3–15) 9.09 ± 3.89 (1–17) 0.043*d
Normal 34 (52.3 %) 28 (40 %)
Borderline 13 (20 %) 15 (21.4 %)
Abnormal 18 (27.7 %) 27 (38.6 %)
HADS-D 8.57 ± 3.86 (2–17) 9.21 ± 4.01 (2–18) 0.392d
Normal 27 (41.5 %) 25 (35.7 %)
Borderline 20 (30.8 %) 20 (28.6 %)
Abnormal 18 (27.7 %) 25 (35.7 %)
K index 2.14 ± 1.13 (0–4) 1.80 ± 1.23 (0–4) 0.154d
Poor 22 (33.8 %) 26 (37.1 %)
Fair 34 (52.3 %) 39 (55.7 %)
Good 9 (13.8 %) 5 (7.1 %)
* p \ 0.05
a
Student’s t test
b
Continuity correction (Yates) test
c
Pearson’s Chi-square test
d
Mann–Whitney U test

123
 Eur J Orthop Surg Traumatol

Table 2 HADS-A, HADS-D and K index evaluation

HADS-A p HADS-D p K index p
Mean ± SD (median) Mean ± SD (median) Mean ± SD (median)

Age
B35 9.25 ± 3.13 (10) 0.539a 9.13 ± 5.11 (8.5) 0.875a 3.44 ± 0.81 (4)a 0.001**
36–50 8.39 ± 3.64 (7) 8.61 ± 3.86 (9) 2.52 ± 1.03 (3)
[50 8.33 ± 3.69 (8) 8.98 ± 3.77 (9) 1.48 ± 0.98 (1)
Level
Transtibial 7.77 ± 3.16 (7) 0.531b 8.57 ± 3.86 (9) 0.901b 2.14 ± 1.13 (2) 0.154b
Transfemoral 9.09 ± 3.89 (9) 9.21 ± 4.01 (9) 1.80 ± 1.24 (2)
** p \ 0.01
a
Kruskal Wallis test
b
Mann–Whitney U test

most common etiologies of the amputation are trauma and 32 % (n = 43) were classified as abnormal according to
vascular disease and that amputations are performed at an the HADS-D. In a study by Seidel et al. [23] of patients
older age due to vascular diseases. In the current study, the with lower-limb amputation, 27 % of the amputees had
most common cause of amputation was vascular disease at depression. Hawamdeh et al. [10] reported that 20 %
a rate of 57 %, whereas in a study by Dogan et al. [7] (n = 11) of the amputees were abnormal according to the
which evaluated 475 amputees, trauma was reported to be HADS-D scores, and there was no significant difference
the main cause of amputation (40.2 %). The mean age was between the amputation level and HADS-D scores. The
over 50 years old in both groups in the current study. In a results of the current study confirm those results. In the
study by Rommers et al. [22], it was stated that 94 % of the present study, it was observed that amputees with trauma
amputations were performed as a result of vascular dis- and TF were more likely to have higher HADS-A scores
eases and the mean age was older than 65 years in 79 % of and were more anxious than those with TT and vascular-
the study population. Settakorn et al. [24] reported that the related amputations. Patients undergoing TF amputation
major cause of lower-limb amputation was vascular dis- as a result of trauma may need psychological support, and
eases (51 %). The data of this present study confirm the therefore, by taking the necessary precautions, anxiety
results of Settakorn et al. and Rommers et al. [22, 24]. problems of the patient following amputation would be
Several studies have been published in the literature on able to be reduced.
the prevalence of depression and anxiety in individuals The Medicare K level is a predictor of prosthetic satis-
with limb loss. In the present study, 33.3 % of the faction. Several factors, such as age, time since amputation,
amputees had abnormal anxiety. Hawamdeh et al. [10]. number and severity of comorbidities, residual limb pain
reported that of 56 amputees, 37.5 % (n = 21) had an and prosthetic fit, have been shown to have a significant
abnormal anxiety (HADS-A) score. When the TT ampu- impact on the amputee patient’s ability to ambulate with a
tations were compared with the TF amputations in respect prosthesis [7, 11–16]. In the present study, there was a
of HADS-A, there was a statistically significant difference significant correlation between age and K level. With
between the amputation level and HADS-A scores. The increased patient age, the K level reduced (p = 0.001,
HADS-A scores of the TF amputees were significantly p \ 0.01), but there was no correlation between the
higher than those of the TT amputees (p = 0.043, amputation level and K level. In a study by Gailey et al. [9],
p \ 0.05). This may result from the etiology of the it was stated that as the mean age of the participants
amputation, as trauma has higher psychological sequelae decreased, the K level increased correspondingly. This was
than vascular-related amputations [8]. In the present an expected result and reflects the outcome of the amputee
study, trauma was the main cause of 50 % (n = 35) of patient population as described throughout the literature
the TF amputees, and 35.4 % of the TT amputees. The [16, 18, 20]. In the present population, the amputee patients
current study also confirmed the findings of the study by with higher ages had comorbidities such as diabetes mel-
Hawamdeh [10] which reported that patients with limb litus or cardiopulmonary comorbidities. It is well known
loss due to trauma were more anxious than amputees with that as comorbidities increase, the K level status is
vascular diseases. In the current study of 135 amputees, diminished [2].

123
Eur J Orthop Surg Traumatol

Gender and age had an impact on outcomes following comparison to men with surgical lower part amputation. Mil Med
amputation. Hawamdeh et al. [10] and Pezzin et al. [19] 168:106–109
5. Cavanagh SR, Shin LM, Karamouz N, Rauch SL (2006) Psy-
reported that women were more likely to suffer anxiety and chiatric and emotional sequelae of surgical amputation. Psycho-
depression after amputation surgery. In the present study, somatics 47:459–464
gender and age had no impact on HADS-A or HADS-D 6. Dillingham TR, Pezzin LE, MacKenzie EJ (2002) Limb ampu-
scores, and these results were in line with those of Bradway tation and limb deficiency: epidemiology and recent trends in the
United States. South Med J 95:875–883
et al. [3]. However, the results of the current study cannot 7. Doğan A, Sungur İ, Bilgiç S, Uslu M, Atik B, Tan O et al (2008)
be generalized, as females only comprised 17.8 % Amputations in eastern Turkey(Van): a multicenter epidemio-
(n = 24) of the study group. Similar gender representations logical study. Acta Orthop Traumatol Turc 42:53–58
have been reported by Cavanagh et al. [5] as 25 %, and by 8. Fisher K, Hanspal RS (1998) Phantom pain, anxiety, depression,
and their relation in consecutive patients with amputated limbs:
Muzaffar et al. [17] as 21 %. It would be beneficial to case reports. BMJ 316:903–904
conduct further research on this topic. The dominance of 9. Gailey RS, Roach KE, Applegate EB, Cho B, Cunniffe B,
the male gender could be a result of a patriarchal society, Licht S, Maguire M, Nash MS (2002) The amputee mobility
and also the use of tobacco and alcohol is prevalent among predictor: an instrument to assess determinants of the lower-
limb amputee’s ability to ambulate. Arch Phys Med Rehabil
males. 83:613–627
The current study had some limitations, namely the wide 10. Hawamdeh ZM, Othman YS, Ibrahim AI (2008) Assessment of
variety in age, the low number of patients and the low anxiety and depression after lower limb amputation in Jordanian
representation of females in the study sample. Other patients. Neuropsychiatr Dis Treat 4:627–633
11. Johnson V, Kondziela S, Gottschalk F (1995) Pre and post-am-
demographic characteristics (marital status, employment putation mobility of trans-tibial amputees: correlation to medical
and education status) could affect the depression and problems, age and mortality. Prosthet Orthot Int 19:159–164
anxiety levels of the amputees. The Medicare K level 12. Keagy B, Schwartz J, KotbM Burnham S, Johnson G (1986) Lower
scoring system is a subjective and nonstandardized method. extremity amputation: the control series. J Vasc Surg 4:321–326
13. Kerstein M, Zimmer H, Dugdale F, Lerner E (1975) What
The lack of a standardized and objective system for influence does age have on rehabilitation of amputees? Geriatrics
assigning K levels makes possible both over- and under- 30:67–71
prescription of prosthetic components. 14. Leung EC-c, Rush PJ, Devlin M (1996) Predicting prosthetic
In conclusion, the results of the current study confirm rehabilitation outcome in lower limb amputee patients with the
functional independence measure. Arch Phys Med Rehabil
the prevalence of psychological disorders after lower-limb 77:605–608
amputation. The data also show higher HADS-A scores 15. Melchiorre P, Findley T, Boda W (1996) Functional outcome and
with traumatic transfemoral amputation. Therefore, ade- comorbidity indexes in the rehabilitation of the traumatic versus
quate psychiatric evaluation and rehabilitation should be the vascular unilateral lower limbamputee. Am J Phys Med
Rehabil 75:219–224
applied to all amputees with lower-limb amputations, 16. Moore T, Barron J, Hutchinson F, Golden C, Ellis C, Humphries
especially in cases of young, traumatic, above-the-knee D (1989) Prosthetic usage following major lower extremity
amputations. Orthopedic surgeons should give more amputation. Clin Orthop 238:219–224
attention to the psychological status and should be a part of 17. Muzaffar N, Mansoor I, Hafeez A, Margoob M (2012) Psychi-
atric comorbidity in amputees with average sociodemographic
a multidisciplinary team. status and the role of theologic and family support in a conflict
zone. Australas J Disaster Trauma Stud 1:31–38
Compliance with ethical standards 18. Ng E, Berbrayer D, Hunter G (1996) Transtibial amputation:
preoperative vascular assessment and functional outcome.
Conflict of interest None. J Prothet Orthot 8:123–129
19. Pezzin LE, Dillingham TR, MacKenzie EJ (2000) Rehabilitation
and the long-term outcomes of persons with trauma-related
amputations. Arch Phys Med Rehabil 81:292–300
References 20. Pohjolainen T, Alaranta H, Karkkainen M (1990) Prosthetic use
and functional and social outcome following major limb lower
1. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J (1961) An limb amputation. Prosthet Orthot Int 14:75–79
inventory for measuring depression. Arch Gen Psychiatry 21. Radloff L (1977) The CES-D Scale: a self-report depression scale
4:561–571 for research in the general population. Appl Psychol Meas
2. Beekman C, Axtell L (1987) Prosthetic use in elderly patients 1:385–401
with dysvascular above-knee and through-knee amputations. 22. Rommers GM, Vos LD, Groothoff JW, Schuiling CH, Eisma WH
Phys Ther 67:1510–1516 (1997) Epidemiology of lower limb amputees in the north of The
3. Bradway JK, Malone JM, Racy J, Leal JM, Poole J (1984) Psy- Netherlands: aetiology, discharge destination and prosthetic use.
chological adaptation to amputation: an overview. Orthop Pros- Prosthet Orthot Int 21:92–99
thet 38:46–50 23. Seidel E, Lange C, Wetz HH, Heuft G (2006) Anxiety and
4. Cansever A, Uzun O, Yildiz C, Ates A, Atesalp AS (2003) depression after loss of a lower limb. Orthopade
Depression in men with traumatic lower part amputation: a 35:1154–1158

123
 Eur J Orthop Surg Traumatol

24. Settakorn J, Rangdaeng S, Arpornchayanon O, Lekawanvijit S, 26. Zidarov D, Swaine B, Gauthier-Gagnon C (2009) Quality of life
Bhoopat L, Attia J (2005) Why were limbs amputated? An of persons with lower-limb amputation during rehabilitation and
evaluation of 216 surgical specimens from Chiang Mai Univer- at 3-month follow-up. Arch Phys Med Rehabil 90:634–645
sity Hospital. Thail Arch Orthop Trauma Surg 125:701–705 27. Zigmond AS, Snaith RP (1983) The hospital anxiety and
Epub 2005 Oct 8 depression scale. Acta Psychiatr Scand 67:361–370
25. Snaith RP, Zigmond AS (1986) The hospital anxiety and
depression scale. Br Med J 292:344 (Clin Res Ed)

123
 931

Prosthesis Evaluation Questionnaire for Persons With Lower

Limb Amputations: Assessing Prosthesis-Related Quality of Life
Marcia W Legro, PhD, Gayle D. Reiber, PhD, Douglas G. Smith, MD, Michael de1 Aguila, MSc,
Jerrie Larsen, RN, David Boone, MPH, CP

N 1989, 3 OF EVERY 1,000 US residents without diabetes

ABSTRACT. Legro MW, Reiber GD, Smith DG, de1Aguila
M, Larsen J, Boone D. Prosthesis evaluation questionnaire for
persons with lower limb amputations: assessing prosthesis-
Iwere(0.3%) and 29 of every 1,000 persons with diabetes (2.9%)
estimated to be living with a lower limb amputation.’ In
related quality of life. Arch Phys Med Rehabil 1998;79:93 1-8. 1990 lower limb amputations were performed on more than
105,300 individuals in US short-stay hospita1s.l The same year,
Objective: To develop a self-report questionnaire for persons additional amputations were performed in veterans, military,
with lower limb amputations who use a prosthesis. The Indian health, and charitable orthopedic hospitals. During
resulting scales were intended to be suitable to evaluate the periods of international conflict, combat or war-related amputa-
prosthesis and life with the prosthesis. The conceptual frame- tions would be added to this count. Among these persons,
work was health-related quality of life. indications for an amputation corresponded to the age of the
Design: Multiple steps of scale development, terminating individual-younger persons experienced more congenital,
with test-retest of the Prosthesis Evaluation Questionnaire malignancy, and trauma-related amputations, whereas older
(PEQ) by mail. persons experienced amputations due to multiple pathophysi-
Source of Sample: Records from two Seattle hospitals. ologic mechanisms (eg, diabetes, ischemia, and infection).2
Patients: Ninety-two patients with lower limb amputations After lower limb amputations, survival rates and years lived
who varied by age, reason for amputation, years since amputa- by persons with an amputation vary by age and by the
tion, and amputation level. indication for surgery. Persons with a new amputation face a
Main Outcome Measures: The 10 scales used were 4 complex set of tasks as they return to their earlier activities.3
prosthesis function scales (Usefulness, Residual Limb Health, Depending on their earlier status, mobility may have been
Appearance, and Sounds), 2 mobility scales (Ambulation and dramatically compromised. LaPlante4 presented data identify-
Transfers), 3 psychosocial scales(Perceived Responses,Frustra- ing the loss of a lower extremity as the factor producing the
tion, and Social Burden), and 1 Well-being scale. Validation highest risk for functional disability. Orthopedic surgeons,
measures were the Medical Outcomes Study Short Form-36, rehabilitation specialists, and prosthetists continually seek to
the Social Interaction subscale from the Sickness Impact develop procedures, equipment, and training to improve the
Profile, and the Profile of Mood States-short form. likelihood of a successful return by patients to their previous
Results: Nine PEQ scalesdemonstrated high internal consis- level of function.
tency. All met test-retest criteria for comparing group results. Although the patient’s input is recognized today as one
Validity was described based on methods used to gather original essential aspect of evaluating care, few of the current methods
items, distribution of scores, and comparison of scores with of evaluating treatment of persons with lower limb amputations
criterion variables. have systematically quantified their self-reported status. There
Conclusions: The PEQ scales displayed good psychometric are two areas in which the patient’s perspective has been
properties. Future work will assess responsiveness of PEQ reported: the use and function of the prosthesis and health-
scales to changes in prosthetic components. We conclude that related quality of life (HRQL).
they will be useful in evaluation of prosthetic care. Several measures were developed for follow-up studies to
0 1998 by the American Congress of Rehabilitation Medi- ascertain which patients are using the prosthesis they received
cine and the American Academy of Physical Medicine and during rehabilitation. 5-7The Prosthetic Profile of the Amputeesz6
Rehabilitation is thorough but lengthy, and while reliability and validity of the
items have been reported, the questions have not been scaled.
The Houghton instrument7 sums the scores of 12 items and
identifies a score of 9 as “rehabilitated.” The items have face
validity but their psychometric properties (reliability and valid-
From the Division of Health Services Research and Development, Puget Sound
Health Care System, Seattle Division (Dr. Legro, Dr. Reiber, Mr. de1 Aguila); the
ity) have not been reported. This instrument assessesthe use of
Departments of Health Services (Dr. Legro, Dr. Reiber, Mr. de1 Aguila, Mr. Boone), the prosthesis but not its qualities.
Eoidemiolozv (Dr. Reiber. Mr. de1 Azuila). Orthomedics (Dr. Smith. Mr. Boone). and Another instrument, the Functional Independence Measure
Rehabilitation Medicine (Mr. Boone), kersity of Washington; the Orthoiaedic (FIM), has been used since 1983 in the Uniform Data System.8
Surgery Unit, VA Puget Sound Health Care System, Seattle Division (Dr. Smith, Ms.
Larsen); and Prosthetics Research Study (Dr. Smith, Mr. Boone), Seattle, WA.
While the FIM is designed as a provider rating instrument,
Submitted for publication September IO, 1997. Accepted in revised form March 6, responses have been collected directly from patients addressing
199x. six domains of function (locomotion/mobility, eating, personal
Supported by grants from Department of Veterans Affairs, Health Services R&D care, bowel and bladder control, cognition, and social function),
and Rehabilitation R&D; and Prosthetics Research Study.
No commercial party having a direct financial interest in the results of the research
some of which are of more interest than others to persons with
supporting this article has or will confer a benefit upon the authors or upon any lower limb amputations who have completed the rehabilitation
orzanization with which the authors are associated. phase. The Motor Score of the FIM has been reported to
‘reprint requests to Marcia Legro, PhD, Senior Research Scientist, MEDTAP correlate with the Houghton score (both measured at dis-
International Inc., 2101 Fourth Avenue, Suite 2200, Seattle, WA98121.
0 1998 by the American Congress of Rehabilitation Medicine and the American
charge).9 However, when Kelly and JessoplO assessed the
Academy of Physical Medicine and Rehabilitation properties of the FIM in a vocational rehabilitation setting they
0003-9993/98/790X-4648$3.00/0 found ceiling effects-that is, most patients scored at the

Arch Phys Med Rehabil Vol 79, August 1998

932 PROSTHESIS EVALUATION QUESTIONNAIRE, Legro

highest level in many categories. Such an instrument leaves no by a loo-mm line bounded by two anchor phrases denoting the
place to measure improvement over time. extremes of possible answers (eg, never and all the time).
Finally, some researchers have reported patient evaluations Respondents indicated their answer by making a mark across
of their prostheses”-l4 using questions designed for the specific the line. This format was chosen because the wide variety of
study. Further information is not available about the properties topics to be assessedprecluded a common set of responses.
of the questions. We did not find an instrument of reasonable Appropriate extremes were provided for each question with the
length that quantified patient assessmentof the prosthesis and most negative situation on the left and the most positive on the
that had been tested for its reliability and validity. right. Pilot testing indicated to us that the respondents under-
The second important aspect of the patient’s status is a stood the direction of the choices, and the specific wording of
broader concept, health-related quality of life. It can be the anchors helped focus on the meaning of each question.
hypothesized that the quality of the prosthesis will have an Figure 1 presents the sample question provided as an example
effect on the patient’s perception about his or her life. The for scoring. The “hand logo” appeared on correspondence with
multidimensional concept of HRQL has evolved to define field study participants.
major domains of life such as physical, psychologic, social/role Field study design and sample. To test the questionnaire
functioning, and well-being. l5 Few such reports have been with a variety of persons, lists of patient names were obtained
published about persons with amputations. One HRQL instru- from a level-one county trauma center and the Veterans
ment, the Medical Outcomes Study: Short Form-36 (SF-36)16 Administration hospital in Seattle, WA. Inclusion criteria
has been used with persons with lower limb amputations.17 This specified that respondents were to: (1) have had a unilateral
instrument has been designed for use with adults, sick or well, amputation at the Symes level (ankle disarticulation) or higher
so that scores for patients with one problem can be compared one or more years earlier; (2) use their prosthesis at least 5 days
with scores for patients with other diseases or with healthy a week; (3) read English; (4) provide informed consent; and (5)
adults. Smith and colleagues17 compared SF-36 scores of complete the test instrument twice. The field study protocol was
persons with amputations to published normal age-matched reviewed and approved by our university’s human subjects
scores. They reported that in three of the four categories of committee.
physical health in the SF-36 (physical function, role limitations Field study proceduies. One hundred forty-four potential
because of physical health, and pain), those persons with participants were first notified about the project by mail and
traumatic lower limb amputations using prostheses had statisti- then received a telephone call to confirm eligibility and interest.
cally lower (poorer) scores than age-matched persons without Of the 126 persons who passed the eligibility screen, 114 agreed
amputations. No significant differences were found for role to participate. All were mailed a packet containing a consent
limitation due to emotional problems, social functioning, form, the PEQ, a questionnaire covering demographic and
mental health, energy/fatigue, or health perception. No instru- clinical information, and three standardized scales. Upon the
ment was found in the literature with which to examine return of the respondent’s consent form and first completed
amputation- or prosthesis-related quality of life. questionnaire, a second copy was mailed to test the temporal
The purpose of our project was to develop a self-report stability of the answers. Those who did not return their
instrument (questionnaire) that would (1) be specific to persons questionnaires within a month were contacted by telephone
with lower limb amputations and (2) measure small differences and/or mail.21
in prosthesis function and major life domains related to the Validation instruments. Three instruments were chosen to
prosthesis function. Such an instrument could be used to validate aspects of the PEQ. The SF-36 is a well-established
compare the effects of different types of prostheses or different measure of general health status that has proven reliability and
methods of care. This paper reports the development of a validity. Age and disease state adjusted scores for the subscales
reliable and valid questionnaire for the evaluation of prosthetic of this measure have been published.22 The subscalesphysical
care by quantifying the evaluations of persons with amputations function and social function were chosen for validation of the
with regard to their prostheses and their prosthesis-related new PEQ [Link] SicknessImpact Profile (SIP) is a general
quality of life. health status measure composed of 12 categories, one of which
is social interaction.23 Although single categories are not
METHODS usually reported individually, this particular scale was chosen to
Developing an item pool. Conforming to the standard validate the psychosocial scales in the PEQ because it is a
methodology for the development of questions for research specifically compatible concept. The third validation measure
purposes,i2 a small group of clinicians and researchers devel- was the Profile of Mood States-short form (POMS-sf), which
oped a 16-item list about prosthesis function (eg, fit, weight,
and color) and pilot-tested it with 22 patients. In line with the
conceptual framework of HRQL, additional items were sought.
These were gathered from published research,5,6from a multidis-
ciplinary group of health professionals, from participants in a
support group for persons with amputations in Seattle, and from
an Internet group for persons with amputations. Thirty-three How important is it to you to have a good prosthesis?
items were added to the original list. A second question was
asked for each item about the importance of the topic to the
NOT AT ALL EXTREMELY IMPORTANT
individual. Both the content and format were again pilot tested
with local patients and with persons on an Internet mailing list.
Over the PAST FOUR WEEKS, rate how happy you have been with your
Final changes were made and the resulting questionnaire, the current prosthesis.
ProsthesisEvaluation Questionnaire (PEQ), was professionally
formatted and printed for the next phase of development, a field
study. EXTREMELY UNHAPPY EXTREMELY HAPPY
Response format. A linear analog scale response format
was chosen for the PEQ.18-20Each question was accompanied Fig 1. Sample question and response format.

Arch Phys Med Rehabil Vol 79, August 1998

 PROSTHESIS EVALUATION QUESTIONNAIRE, Legro 933

Table 1: Sample Characteristics: Demographic rating scale). The calculation included summing scores for the
County
questions in the scale and then dividing that sum by the number
VA Trauma of questions answered by the respondent. The resulting scores
Overall Hospital Center were ordered from 0 (worst) to 100 (best). This required that
Distribution (%) (N = 92) (n = 58) (n = 34) some scale scoreswere reversed, eg, a Usefulness score of 100
Age distribution (yrs) is equal to “very useful” and a Frustration score of 100 is equal
20-39 13.0 3.4 29.4 to “no frustration.”
40-59 46.7 41.4 55.9 The final version of each scale was submitted to reliability
60+* 40.2 55.2 14.7 and validity testing using the computation of Cronbach’s
Gender male* 85.9 94.8 70.6 alpha,28 Pearson product-moment correlation coefficients,2g
Marital status intraclass correlation coefficients,30 and principle component
Never married 12.0 6.9 20.6 factor analysis using varimax rotation.18
Married 62.6 65.5 55.5
Widowed or divorced 26.1 27.6 23.5
RESULTS
Living alone+ 19.6 25.9 8.8
Racial group
Sample Characteristics
White, not Hispanic 83.7 81.0 88.2 Ninety-two persons (73% of those eligible) returned a signed
Black 5.4 6.9 2.9 consent form and a questionnaire, and 81 (88%) of these
Other 10.9 12.1 8.8 persons also completed the retest. Tables 1 and 2 provide
Education (yrs) demographic and clinical descriptions of the 92 participants.
4 through 8 7.6 10.3 2.9 The sample ranged in age from 20 to 87, with 40.2% over 60
9 through 12 41.3 37.9 47.1
13 through 16 38.0 39.7 35.3 Table 2: Sample: Clinical and Prosthetic Characteristics
17 or more 13.0 12.1 14.7
County
Work status employed 37.0 32.8 44.1 VA Trauma
Of those not working (n = 58), Distribution Overall Hospital Center
why not employed 1%) (N = 92) (n = 58) (n = 34)
Can’t find work 3.4 2.6 5.3
Self-reported health compared to
Disabled 39.7 38.5 42.1
others the same age
Retired 43.1 51.3 26.6
Excellent 16.3 17.2 14.7
Homemaker 1.7 2.6 0
Above average 13.2 13.8 11.8
Student 3.4 0 10.5 Average 46.7 41.4 55.9
Other reason given 8.6 5.1 15.8
Below average 19.6 20.7 17.6
Current drivers (% of those
Poor 4.3 6.9 0
who ever drove) 90.0 91.2 87.9
Self-reported comorbidities (ever
*p 5 ,001, + p 5 .05; two-tailed tests (x2 test for proportion of diagnosed*)
dichotomous variables; Mantel-Haenszel x2 test for categorical vari- Diabetes* 27.8 35.1 15.2
ables).
Heart attack 14.4 19.6 5.9
Respiratory disease 15.4 13.8 18.2
Cancer* 10.0 15.5 0.0
produces a summary measure of mood state.24The overall score
Stroke 4.3 5.2 2.9
has been reported for many disease states.25Current mood state
Congestive heart failure 3.4 3.7 3.0
was expected to correlate with well-being as reported in the
None of the above 51.1 51.1 48.9
PEQ.
Level of amputation
Data management. Data entry for the linear analog scales
Transfemoral 25.0 31.0 14.7
was completed using a digitizing tablet with a marking pen that
Knee disarticulation 3.3 3.4 2.9
directly entered information into a data management program.a
Transtibial 63.0 56.9 73.5
This high-speed technology was accurate and reduced data
Ankle disarticulation (Symes) 8.7 8.6 8.8
entry time over other methods of data entry used for this
Time since amputation*
response format. The analysis was completed using SPSS 6.1
1 through 5yrs 38.0 24.1 61.8
for Windowsb
6 through IOyrs 15.3 3.5 35.3
Data analysis: scale identi&ation. After developing the
11 through 20yrs 8.7 12.1 2.9
item pool and administering those items to a field sample of
21 or more years 38.0 60.3 0
persons with amputations, the next step involved testing
Range: I-53yrs
whether the responses to the items fit into the categories
Mean (SD): 18 (17.2)
previously identified in the conceptual framework.26J7 Before
Reason for amputation*
the statistical analysis of the data, the study team had individu-
Trauma (including war injuries) 66.6 59.6 78.8
ally and then collectively categorized all the questions by life
Chronic disease+ 40.7 46.6 30.3
domains. During a lengthy, iterative process of reviewing
Congenital problem* 3.3 0 9.1
(1) descriptive analyses, (2) correlational and factor analyses,
Tumor 1.1 1.8 0
and (3) the responses to the “importance” questions, the group
made modifications to the content of each [Link] choosing * Multiple answers allowed.
between similar items, we were more likely to include an item if + Chronic diseases included diabetes, ulcers, infections, and vascular
more respondents attached greater importance to it. disease.
*p 5 .05, 5 p i .OOl; two-tailed tests, (x2 test for proportion for
When the content of a scale was agreed upon, a single score dichotomous variables; Mantel-Haenszel x2 for categorical vari-
for that scale was calculated for each respondent (a summed ables).

Arch Phys Med Rehabil Vol 79, August 1998

934 PROSTHESIS EVALUATION QUESTIONNAIRE, Legro

years old. The group were predominately men (85.9%) and not of prosthesis function (Usefulness, Residual Limb Health,
working (63%) for various reasons, the most common being Appearance, and Sounds) and two scales concerned mobility
“retired” (43%) and “disabled” (39.7%). Nearly 51% of the (Ambulation and Transfers). In the psychologic and social
respondents reported schooling beyond 12th grade. Ninety domains, three scales were developed (Perceived Responses,
percent of those who ever had a license to drive reported that Frustration, and Social Burden), and one scale was constructed
they were currently driving. to assessoverall Well-being. Each scale that was developed had
Fifty-one percent of the sample reported no medical comor- a single underlying concept, that is, each scale conformed to a
bidity, whereas 27.8% had diabetes and 15.4% had respiratory single principal-factor analysis with varimax rotation.26 Table 3
disease. The most common level of amputation (63%) was lists sample items from each scale.
transtibial (below the knee), although 25% were transfemoral Table 4 summarizes psychometric information about the 10
(above the knee). While 38% of subjects had experienced their scales, including the number of questions in each scale, the
amputation within the last 5 years, an equal number had had the mean, the standard deviation, the range of scoresfor each scale,
surgery more than 20 years earlier. the percentage of respondents who scored at either extreme of
the scale (floor and ceiling effects), and the skew of the score
Reduction of Items (how far the scores were from a normal distribution around the
The scale development analysis employed the responses of mean).
all 92 persons who completed the first questionnaire since the
intent was to have a wide representation of persons with lower Reliability Tests
limb amputations. Completion rates for most items was very Two aspects of scale reliability were examined: internal
high. Within the physical domain, four scalesaddressed aspects consistency and temporal stability. The results of each test can

Table 3: Sample Items in the 10 PEQ Scales

Item Rating

Prosthesis function
Usefulness The fit of your prosthesis From terrible to excellent
The comfort of your prosthesis while standing still when From terrible to excellent
using your prosthesis
Residual limb health How much of the time your residual limb was swollen to the From ail the time to never
point of changing the fit of your prosthesis
How much you have sweat in your prosthesis From extreme amountto not at ail
Appearance The appearance of your prosthesis (how it has looked) From terrible to excellent
How limited your choice of clothing was because of your From worst possible to not at ail
prosthesis
Sounds How often your prosthesis has made sounds (belching, From always to never
squeaking, clicking, etc.)
How bothersome these sounds were to you From extremely bothersome to not at ail

Mobility
Ambulation Your ability to walk in close spaces when using your pros- From cannotto no problem
thesis
Your ability to walk on slippery surfaces (eg, wet tile, snow, From cannotto no problem
a rainy street or boat deck) when using your prosthesis
Transfers Your ability to shower or bathe safely From cannotto no problem
Your ability to get in and out of a car when using your pros- From cannot to no problem
thesis

Psychosocial experience
Perceived responses How often the desire to avoid strangers’ reactions to your From ail the time to never
prosthesis makes you avoid doing something you other-
wise would have done
How a family member (otherthan your partner) has From very poorly to very well
responded to your prosthesis
Social burden How much a burden your prosthesis has been on your From extremely burdensome to not at ail
partner or family member
Your ability to take care of someone else (eg, your partner, From cannotto no problem
a child, a friend)
Frustration How frequenty you were frustrated with your prosthesis From ail the time to never
Think of the most frustrating event and rate what you felt at From extremely frustrated to not at ail
that time

Well-being
Well-being How satisfied you have been with how things have worked From extremely dissatisfied to extremely satisfied
out since your amputation
Your quality of life From worst possible life to best possible life

Respondents are asked to rate the item over the past 4 weeks.

Arch Phys Med Rehabil Vol 79, August 1998

 PROSTHESIS EVALUATION QUESTIONNAIRE, Legro 935

Table 4: Psychometric Summary of the PEQ Scales (N = 92)

Temporal
Number Mean* internal Stability*
Scale Name of Items (SD) Range % = 0 % = 100 Skew Consistency+ (n= 61)

Physical domain: prosthetic

function scales
Usefulness 8 72.0 (20.5) 9-l 00 0 2 -.7 39 66 t.78, .Sl)
Residual limb health 6 71.8 (23.3) 5-100 0 2 -3 .80 .79 i.68, .86)
Appearance 5 73.3 (20.7) S-l 00 0 7 -1.0 .73 .84 (.76, .SO)
Sounds 2 62.7 (31.2) o-100 1 10 -3 .78 .84 t.75, 30)
Physical domain: mobility scales
Ambulation 8 68.5 (22.0) 6-100 0 2 -.7 .a9 30 (.84, 94)
Transfers 2 84.3 (18.9) 21-100 0 25 -1.4 .47 .73 t.58, .83)
Psychologic and social domains:
psychosocial issues
Perceived responses 5 SO.6 (16.1) 29-100 0 17 -2.2 .89 .56 c.36, .71)
Frustration 2 36.6 (32.7) o-100 22 1 .4 .a2 .64 t-47, .77)
Social burden 3 20.6 (24.5) o-97 10 0 1.4 .83 .81 c.69, .88)
Global: well-being
Well-being 2 71.6 (24.4) a-100 0 8 -.7 .a7 .89 t.80, 93)

* All scales are scored so that 100 indicates the best situation, eg, most useful, least frustrating, etc.
+ Cronbach’s alpha.
* lntraclass correlation estimate for mean ratings with approximate 95% confidence intervals.30

be found in table 2. The internal consistency of each scale was strong negative correlation (r = -52) with the SIP social
tested by computing a Cronbach’s alpha.28This statistical test interaction score. This is as expected since a high Social Burden
indicates whether each item in the scale is contributing to the score indicates experiencing no social burden and a low score
variance in the overall score. Cronbach’s alphas for the 10 on the SIP subscale indicates no problem with social interac-
scalesranged from .73 to .89 with one exception-Transfers. tions. The PEQ scale of Well-being showed a moderate,
The second reliability test determined the degree to which the negative correlation (r = -.49) with the total score on the
scores were stable over time for subjects who had not experi- POMS-sf. Again, this is appropriate since a high score on the
enced a change in health or prosthesis. Since 20 respondents Well-being scale is a positive response and a low score on the
reported major changes in their prosthetic status during the time POMS-sf indicates low mental distress.
between the two questionnaires, they were dropped from the
stability analysis. Thus, 61 persons’ responses were retained for
the retest analysis. The mean retest period was 30 days. The DISCUSSION
intraclass correlation (ICC) estimates for the mean scoresof the The development of a group of scalesto describe the function
first and second administration of the scaleswere calculated.30 of a lower limb prosthesis and the prosthesis-related quality of
The results ranged from F = .79 to F = .90 with two life of the person with an amputation has been needed in the
exceptions-Perceived Responses and Frustration. field of physical medicine and rehabilitation. The PEQ de-
scribed in this article was formulated based on the concept of
Validity HRQL as a multidimensional concept. Ten scales were devel-
To demonstrate whether the scales could differentiate be- oped to assessthe physical and psychosocial life domains and
tween groups of people whose scale scoreswould be expected one global measure of well-being.
to be different, scores for the 10 scales were calculated for To develop an instrument that would generalize to the broad
participants grouped by gender, age, presence or absence of range of persons who have experienced amputations for various
comorbidities, level of amputation, and years since the amputa- reasons, the items were drawn from a wide range of sources.
tion (table 5). Statistical differences were found between men The sample recruited for the field study met the objective of
and women on two scales. Men reported significantly better recruiting young and old individuals, individuals whose ampu-
Ambulation and women reported significantly greater Social tations were done recently or long ago, and men and women.
Burden (lower scores).Two scales-Residual Limb Health and Two groups that we did not attempt to include in this project
Frustration-differed significantly across age groups such that were children and persons with bilateral lower limb amputa-
younger patients identified more problems with their residual tions. Many of the questions in the scaleswould apply to these
limbs and greater frustration than participants who were 40 persons; however, because of additional and unique issues,
years or older. Ambulation scoresdiffered significantly between administering the scalesto either group would require modifica-
those with and without comorbidities such that those with no tions.
comorbidities reported better Ambulation. Finally, no statisti- The reliability of each scale was tested in two ways. The first
cally significant differences were noted by amputation level or method, a statistical test of the internal consistency of the items
years since amputation. in the scale, Cronbach’s alpha, assigns a number between 0 and
Table 6 contains the correlations between criterion measures 1 to indicate how much of the variation in the scale score is
and the relevant PEQ [Link] correlations, between r = .49 accounted for by each item. It also takes into account the
and r = .61, are statistically significant and in the direction number of items in the scale. Scales with alphas of .70 are
expected. For example, the Ambulation scale of the PEQ was acceptable for group comparisons.27All but one of the scalesin
strongly correlated (r = .61) with the SF-36 subscaleof physical this paper were above this threshold. The two-item Transfer
function. The Social Burden scale of the PEQ demonstrated a scale did not reach alpha = .70, and now additional items have

Arch Phys Med Rehabil Vol 79. August 1998

936 PROSTHESIS EVALUATION QUESTIONNAIRE, Legro

Table 5: Mean PEQ Scale Scores (o-100) for Subgroups of Persons With Lower Limb Amputations (N = 92)

Amputation Level (n) Years Since Amputation (n)

Age Group(n)
Gender(n) Comorbidities (n) Transfemoral (23) 53 years (2.3)
20-39 (12) Knee disarticulation (3) 4 through 8 (22)
OVerall Men (79) 40-59 (43) None (47) Transtibial (58) 9 through 28 (24)
Mean ScoreX Women (13) 60 or more (37) 1 or more (45) Ankle disarticulation (8) 29 or more (21)
Prosthetic function scales
Usefulness 73 66 74 65 70
72 (20.5) 66 74 70 82 72
71 75 76
63 72

Residual Limb Health 72 52+ 73 76 67

71.8 (23.3) 70 71 70 91 64
77 69 77
72 77

Appearance 75 62 75 68 69
73.3 (20.7) 63 77 72 90 75
72 74 75
75 74

Sounds 63 56 61 53 66
62.7 (31.3) 58 62 64 56 62
64 65 60
79 62

Mobility scales
Ambulation 70* 66 73* 66 65
68.5 (22.0) 56 73 63 76 65
65 68 73
70 71

Transfers 85 79 86 84 85
84.3 (19.0) 80 87 82 94 85
83 84 87
83 79

Psychosocial experiences
Perceived Responses 91 80 91 90 91
90.6 (16.1) 88 92 90 99 88
92 90 89
92 94

Frustration 35 62* 38 46 47
37.6 (32.7) 43 33 35 18 34
33 34 30
35 36

Social Burden 181 30 18 23 25

20.5 (24.5) 34 16 24 16 27
22 20 18
19 12

Well-being scale
Well-being 73 58 73 63 66
71.6 (24.4) 64 77 70 67 69
71 75 77
75 74
Bold type indicates that the means differ in a significant manner within this category.
* A higher score reflects a better situation.
+ p 5 .Ol, two-tailed ttest or ANOVA Ftest.
* p 5 .05, two-tailed ttest or ANOVA ftest.

been identified31 and are being tested in a study under way in coefficient, quantifies temporal stability or the similarity of
Seattle. If these items prove psychometrically useful, they will answers given by a person at two points in time when the person
be included in future versions of the PEQ. has not changed in relation to the item being tested. All scores
The second test of reliability, the intraclass correlation were at an acceptable level; however, two were lower than the

Arch Phys Med Rehabil Vol 79, August 1998

 PROSTHESIS EVALUATION QUESTIONNAIRE, Legro 937

Table 6: Correlations between PEQ Scales and Criterion Scales general support and assistance with recmitment; and to our 92
respondents who generously and patiently filled out long question-
Correlation
PEQ Scale Criterion Scale (4
naires about a major personal event in their lives.

Ambulation SF-36: physical function .61

Social Burden SF-36: social function .59 References
Social Burden SIP: social interaction -.52* 1. Reiber GE, Boyko E, Smith DG. Lower extremity ulcers and
Well-being POMS-sf: total score -.49* amputations in individuals with diabetes. In: Harris MI, editor.
Diabetes in America. 2nd ed. Bethesda (MD): National Institutes
* Negative correlations are present because the PEQ scales are
of Health; 1995. Publication No.: 95-1468. p. 409-27.
ordered so that the best situation is equal to 100, whereas the SIP
and POMS scores are ordered so that the worst situation is equal to 2. Reiber GE. Who is at risk of limb loss and what to do about it? J
100. Rehabil Res Dev 1994;31:357-62.
3. Williamson GM, Schultz R, Bridges MW, Behan AM. Social and
rest. Perceived Responses and Frustration were two concepts psychological factors in adjustment to limb amputation. J Sot
that might be expected to vary more rapidly than the other Behav Personality 1994;5:249-68.
scales. 4 LaPlante MP. Disability risks of chronic illnesses and impairments.
Proving the validity of a scale is an ongoing process and Disability Statistics Report, Report 2. Washington (DC): US
refers not only to the meaning of the scale but to its use in a Department of Education, Office of Special Education and Reha-
particular population. The purpose of testing validity was to bilitative Services, National Institute on Disability and Rehabilita-
gather evidence that the scale was measuring what it was tion Research; 1991. p. l-39.
intended to measure. The validity of the 10 PEQ scales was 5. Gaufhier-Gagnon C, Griese M-C. Prosthetic profile of the amputee
examined in a variety of ways. The extended process of questionnaire: validity and reliability. Arch Phys Med Rehabil
development contributed to face validity. By consulting persons 1994;75:1309-14.
6. Grise MCL, Gauthier-Gagnon V, Martineau GG. Prosthetic profile
with lower limb amputations and their caregivers, we gained
assurance that the questions in the PEQ represent the major of people with lower extremity amputation: conception and design
issues that are faced by persons with lower limb amputations as of a follow-up questionnaire. Arch Phys Med Rehabil1993;74:862-
70.
they live with their prostheses. I. Houghton A, Allen A, Luff R, McCall I. Rehabilitation after lower
Another way to check the validity of a new scale is to
limb amputation: a comparative study of above-knee, through knee
examine the scores reported by different subgroups of persons. and G&i-Stokes amputations. Br J Surg 1989;76:622-4.
Although causation cannot be inferred with cross-sectional 8. Center for Functional Assessment Research. Guide for use of the
data, several plausible significant differences were found be- uniform data set for medical rehabilitation including the Func-
tween groups of respondents (table 5). The fact that younger tional Independence Measure. Buffalo (NY): State University of
persons were experiencing more problems with Residual Limb New York; 1991.
Health could well have been because they often attempt to 9. Leung C-CE, Rush PJ, Devlin M. Predicting rehabilitation out-
accomplish more activity while using their prostheses, thus come in lower limb amputee patients with the Functional Indepen-
causing blisters and sores to develop. That group also reported dence Measure. Arch Phys Med Rehabil1966;77:605-8.
experiencing greater Frustration. The significant difference in 10. Kelley S, Jessop EG. A comparison of measures of disability and
Ambulation scores between those with and without comorbidi- health status in people with physical disabilities undergoing
ties confirmed another difference we would have expected. vocational rehabilitation. J Public Health Med 1996; 18: 169-74.
A third measure of validity is to compare the performance of 11. Sheredos SJ. Report on the evaluation of the VA/SEATTLE ankle.
a new scale against that of a criterion variable, sometimes J Rehabil Res Dev 1991;28:79-90.
spoken of as a “gold standard.” Unfortunately, in the field of 12. Wirta RW, Mason R, Calvo K, Golbranson FL. Effect on gait using
HRQL there are few gold standards, but there are instruments various prosthetic ankle-foot devices. 3 Rehabil Res Dev 1991;28:
that have been well tested and reported in a variety of 13-24.
populations. They measure specific aspects of HRQL and can 13. Murray DD, Hartvikson WJ, Anton H, Hommonay E, Russell N.
be used to make some judgment about validity. Three of these With a spring in one’s step. Clin Prosthet Orthot 1988;12:128-35.
standardized measures were chosen to test the validity of the 14. Menard MR, Murray DD. Subjective and objective analysis of an
relevant new scales. The significant correlation of the three energy-storing prosthetic foot. J Pro&et Orthot 1989;1:220-30.
15. Patrick DL, Erickson P. Health status and health policy. New York:
published scales-SF-36, POMS-sf, and SIP-SI-are further
Oxford University Press; 1993.
evidence that the PEQ scales measure what they are intended to 16. Ware JE, Sherboume C. The MOS 36-item short-form health
measure. survey (SF-36): I. Conceptual framework and item selection. Med
The PEQ is being used in an intervention study to test the Care 1992;30:473-83.
responsiveness to change, or sensitivity, of each scale. Judging Il. Smith DC, Horn P, Malchow D, Boone DA, Reiber GE, Hansen, Jr
from the strong psychometric properties of the scales as ST. Prosthetic history, prosthetic charges, and functional outcome
described in this paper, the report of the intervention study will of the isolated, traumatic below-knee amputee. J Trauma 1995;38:
complete the evaluation of the extremely useful new instru- 44-7.
ment. At this point, the 10 scales of the PEQ are qualified to be 18. Miller MD, Ferris DG. Measurement of subjective phenomena in
used with persons with lower limb amputations for evaluating primary care research: the visual analogue scale. Fam Pratt Res J
their prostheses and prosthesis-related quality of life. With the 1993;13:15-24.
description of the responsiveness of the scales to change, the 19. Guyatt GH, TownsendM, BermanLB, KellerJL. A comparisonof
scaleswill be useful to describe changes that occur with various likert and visual analogue scales for measuring change in function.
interventions. J ChronDis 1987;40:1129-33.
20. Scott J, Huskisson EC. Graphic representation of pain. Pain
Acknowledgments: Special appreciation is expressed to Linda 1976;2:175-84.
Laing and the employees of Prosthetics Research Study for computer 21. Dillman DA. Mail and telephone surveys: the total design method.
support and pilot testing questions; to Dee Malchow, MN, RN, for New York: John Wiley and Sons; 1978.

Arch Phys Med Rehabil Vol79, August 1998

938 PROSTHESIS EVALUATION QUESTIONNAIRE, Legro

22. Ware JE Jr, Snow KK, Kosinski M, Gandek B. SF-36 Health 28. Cronbach LJ. Coefficient alpha and the internal structure of tests.
Survey: manual and interpretation guide. Boston (MA): The Psychometrika 1951;16:297-334.
Health Institute, New England Medical Center; 1993. 29. Ferguson GA. Statistical analysis in psychology and education. 5th
23. Bergner M, Bobbitt RA, Carter WB, Gilson BS. The Sickness ed. New York: McGraw-Hill Book Co.: 1981.
Impact Profile: development and final revision of a health status 30. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater
measure. Med Care 1981;19:787-805. reliability. Psycho1 Bull 1979$6:420-S.
24. Shacham S. A shortened version of the Profile of Mood States. J 3 1. Roorda LD, Roebroeck ME, Lankhorst GJ, van Tilburg T, Bouter
Pers Assess 1983;47:305-6. LM. Measuring functional limitations in rising and sitting down:
25. Curran SL, Andrykowski MA, Studts JL. Short form of the Profile development of a questionnaire. Arch Phys Med Rehabil1996;77:
of Mood States (POMS-SF): psychometric information. Psycho1 663-9.
Assess 1995;7:80-3.
26. DeVellis RF. Scale development: theory and applications. New- Suppliers
bury Park (CA): Sage Publications; 1991. a. Lockheed Drawing Slate; Calcomp Digitizer Products, Anaheim,
27. Nunnally JC, Bernstein IH. Psychometric theory. 3rd ed. New CA.
York: McGraw-Hill; 1994. b. SPSS, 444 North Michigan Avenue, Chicago, IL.

Arch Phys Med Rehabil Vol79, August 1998

 Clinical Rehabilitation
[Link]

Mobility of people with lower limb amputations: scales and questionnaires: a review
Gerardus M Rommers, Luc D W Vos, Johan W Groothoff and Willem H Eisma
Clin Rehabil 2001 15: 92
DOI: 10.1191/026921501677990187

The online version of this article can be found at:

[Link]

Published by:

[Link]

Additional services and information for Clinical Rehabilitation can be found at:

Email Alerts: [Link]

Subscriptions: [Link]

Reprints: [Link]

Permissions: [Link]

Citations: [Link]

>> Version of Record - Jan 1, 2001

What is This?

Downloaded from [Link] at Maastricht University on June 23, 2014

Clinical Rehabilitation 2001; 1 5: 92–102

Mobility of people with lower limb amputations:

scales and questionnaires: a review
Gerardus M Rommers Rehabilitation Centre, Revalidatie Friesland, Beetsterzwaag, Luc DW Vos Rehabilitation Centre,
Beatrixoord, Haren, Johan W Groothoff Northern Centre for Healthcare Research, University of Groningen and Willem
H Eisma Department of Rehabilitation, University Hospital Groningen, Northern Centre for Healthcare Research,
University of Groningen, The Netherlands

Received 25th June 1999; returned for revisions 10th August 1999; revised manuscript accepted 8th January 2000.

Objective and design: A systematic literature review to compare mobility

scales used for lower limb amputees. A literature search was carried out by
computerized search of biomedical literature including Medline and Embase.
The studies included were published between 1978 and 1998 and including
the following keywords: amputation, artificial limbs, prosthesis, lower limb,
activities of daily living, mobility.
Results: Thirty-five studies were identified; 19 had a measurement of
separate levels of mobility comparable to each other. Sixteen studies used
ordinal and ratio scales without separate levels of mobility. The widest range
of measurement found was the scale from ‘walking with prosthesis without a
walking aid’ to ‘totally confined to bed’. The Stanmore Harold Wood mobility
scale was published most frequently. None of the 35 studies presented give
a continuous measurement of mobility.
Conclusion: A multitude of measurement scales and questionnaires are
available for differ in methods and measuring range. Measuring mobility by a
scale has been shown to have limitations. Several authors did extensive
research but they all measure only a number of aspects of mobility.
Consensus about the measurement of mobility of lower limb amputees is not
available in the recent literature.

Introduction neurological disease.1–4 The key to independence

for this group is their walking ability and their
In the Netherlands and Northern Europe, over ability to move in and around their homes.3
90% of all lower limb amputations are performed Limited indoor walking ability allows transfers
for the treatment of vascular occlusive disease; from wheelchair to bed or toilet facilities to
about 45% of these lower limb amputations are ensure independence and self-esteem. Limited
related to diabetes mellitus.1,2 About 80% of the outdoor walking ability gives the amputee the
patients are over 60 years of age and have more possibility of taking part in social activities in the
or less co-morbidity in vascular, respiratory and local community. It includes transfers from
wheelchair into transport facilities, taking ramps
Address for correspondence: GM Rommers, Rehabilitation
and uneven pavements.
Centre De Hoogstraat, Rembrandtkade 10, 3583 TM Utrecht, Analysis of the available mobility instruments
The Netherlands. e-mail: [Link]@[Link] in this field is essential to compare results of the
© Arnold 2001 0269–2155(01)CR386OA
Downloaded from [Link] at Maastricht University on June 23, 2014
 Mobility of lower limb amputees 93

rehabilitation treatment for this growing group of Inventory regarding mobility

amputees. A multitude of measurement scales Table 1 compares the studies published from
and questionnaires are available but they differ 1978 onwards. In Table 1a6–9,11–25 studies using
in methods and measuring range. A systematic mobility scales with distinct qualitative levels of
review of these instruments gives an overview of mobility are given (e.g. walking without an aid,
the mobility scales used and will produce guide- with help of a crutch or frame, or wheelchair
lines for the best mobility scale for rehabilitation use). In Table 1b26–41 studies using scales with
treatment. ordinal scores are given (e.g. walking outside
Our primary goals in this study are: the house, community walker, carry out several
1) to give a review of different mobility scales household activities, driving a car, etc.).
for lower limb amputees in the studied liter- The columns in Table 1 are as follows:
ature; and • Population The total number of patients
2) to compare the range of measurement of the included in the study. This in order to follow
mobility scales. correctly the ‘intention-to-treat principle’ at
the start of the study. Most studies showed
results for a selected population, for example,
Methods patients fitted with a prosthesis only.
• Age The age of the participating patients
A complete literature search on Medline from was recorded if this was stated in the original
1978–1998 and on Embase from 1988 to 1998 was study.
carried out. For the search, we used the key- • Amputation level The amputation level is
words: amputation, artificial limbs, prosthesis, given according to the rules of the Inter-
lower limb, activities of daily living, mobility, national Society for Prosthetics and Orthotics
questionnaire and combinations of these topics. (ISPO) consensus conference in 1990.42
Review articles found were also checked in the Bilateral amputees are often of several
database and included in the search. amputation levels and only a few studies6,13
The selected summaries were classified accord- gave a separate description of this group.
ing to the following data: In most studies results for unilateral and
1) A measurement scale for mobility of lower bilateral amputees were put together.
limb amputees. • Reason amputated The cause for amputa-
2) Publication period between 1978 to 1998. tion is given according to the description in
3) Use of walking aids stated. the original publication.
• Mobility scale The mobility scale described
The studies presented were reviewed in order is given with the distinct ordinal levels of
to analyse the mobility scales used. The mobility mobility. If a ratio or interval score is used
scales included performance in locomotion as for time, walking speed or distance, it is noted
well as walking distance and walking speed. Mea- separately. In an ordinal scale, it is stated
suring mobility of lower limb amputees was that items in the scale stand in some kind of
related to the use of walking aids as this is an relation to each other. There is no true
item frequently used to measure mobility at zero point and the intervals between the
home after rehabilitation treatment. Personal items are not equal. The scales are presented
communication with several authors was used to in their original form and if an author used
classify several walking aids not mentioned in the a scale of another author, this is stated
questionnaires. Several studies combined mea- separately.
surement of mobility together with question-
• Questionnaire This gives information about
naires used to get additional information of
the use of a questionnaire in order to collect
prosthetic use and independence in ADL.
information about mobility items in the
studies separately from the mobility scale. To
measure mobility, several studies used the
Barthel index5 together with Russek’s classi-
Downloaded from [Link] at Maastricht University on June 23, 2014
 94

Table 1 Mobility scales compared in the study

Author Population Age Amputation Reason for Mobility Questionnaire Use of aids Stairs
level amputation scale

(a) Studies using mobility scales with distinct qualitative levels of mobility
Volpicelli6 103 29–94 Bilateral Vascular, Ordinal 6 levels – Crutch, crane, walker, +
(1983) diabetes, trauma wheelchair, bed
7
Narang 500 2–90 TF, KD, TT Trauma, illness Ordindal 5 levels + Crutch, wheelchair, –
(1984) no prosthesis
GM Rommers et al.

8
Helm 257 38–95 TF, TT Vascular, Ordinal 4 levels Crutch, frame, –
(1986) bilateral diabetes wheelchair, no prosthesis,
others cosmetic
Kullman9 452 8–90 Not given Vascular Ordinal 5 levelsb + Not given –
(1987) diabetes, tumour,
others
Stern11 a 238 mean TF, TT, Vascular, Ordinal 5 levelsb – Crutch, walker, –
(1988) 66 bilateral diabetes no prosthesis
Pinzur12 46 Not KD Vascular, Ordinal 6 levels – No –
(1988) given diabetes
Wolf13 18 55–83 Bilateral Vascular Ordinal 8 levels – Walking aids, wheelchair, –
(1989) TF/TT assistance
Siriwardena14 598 50–70+ TF, KD, TT Vascular Ordinal 6 levels – Crutch, frame, wheelchair –
(1991)
Pohjolainen15 155 14–87 TF, TT Vascular, tumour, Ordinal 7 levels + Crutch, frame, –
(1991) trauma wheelchair,
no prosthesis
Hanspal16 100 60–89 TF, TT Not given Ordinal 6 levels + Crutch, stick, frame –
(1991) wheelchair, cosmetic,
assistance
Hepp17 198 Mean 85% unilateral Vascular Ordinal 7 levels – Crutch, wheelchair, –
(1991) 63 15% bilateral no prosthesis,
bed
Houghton18 a 440 39–90 TF, TT, GS, KD Vascular Ordinal 6 levelsc + Crutch, stick, frame –
(1992) bilateral wheelchair, cosmetic

Downloaded from [Link] at Maastricht University on June 23, 2014

assistance
Datta19 a 41 31–84 Bilateral Vascular, Ordinal 6 levelsd – Crane, crutch, walker, +
diabetes, trauma wheelchair
20
Zijp 61 38–91 Not given Not given Ordinal 3 levels – Crutch, frame, wheelchair –
(1992)
c
Lachman21 a 11 40–82 TF, TT Vascular, Ordinal 6 levels – Crutch, stick, frame –
(1993) infection, wheelchair, cosmetic,
artroplasty assistance
Campbell22 a 210 43–96 TF, TT, GS Vascular Ordinal 6 levelsc – Crutch, stick, frame, –
(1994) cosmetic
d
Johnson23 a 120 25–89 TT Vascular, trauma Ordinal 6 levels + Crutch, cane, walker, –
(1995) wheelchair, bed
c
Kanellopoulos24 a 93 42–93 TF, TT Vascular Ordinal 6 levels + Crutch, stick, frame –
(1996) bilateral wheelchair, cosmetic,
assistance
Burger25 519 Mean HD, TF, KD Trauma Ordinal 3 levels + Crutch, cane, wheelchair +
(1997) 54,4 TT
 (b) Studies using mobility scales with ordinal scores
Kegel26 134 10–90 TF, TT Vascular, trauma, Ordinal score + Crutch, frame, wheelchair +
(1978) bilateral tumour, congenital
27
Day (1981) 2400 Not given Not given Not given Ordinal score + Walking aid, wheelchair +
28
Steinberg 114 65+ TF, TT Not given Ordinal score – Crutch, cane, wheelchair, –
(1985) bilateral 3 levels cosmetic
29
Beekman 55 Mean TF, TT Diabetes, other Ordinal score + Crutch, cane, walker, +
(1988) 65,4 causes ratio scale: time wheelchair
walking speed
interval score:
distance
Lavan30 146 65+ TF, TT Vascular, Ordinal score Clinical Not given +
(1988) bilateral other causes investigation
Chan31 47 65+ TF, TT, TM, Vascular, tumour, Ordinal score + Not given, cosmetic –
(1990) bilateral trauma, diabetes
Brodzka32 24 54–95 Bilateral Vascular Ordinal score Telephone Crutch, frame, wheelchair –
(1990) TT–TT interview
Collin33 40 50–81 TF, TT, Vascular, Ordinal score + Not given except +
(1992) bilateral diabetes ratio scale: time home visit wheelchair use
Hagberg34 59 50+ TF, KD, TT Vascular and Ordinal score + Crutch +
(1992) other reasons
Nissen35 46 42–95 TF, TT Diabetes and Ordinal score + Wheelchair and other –
(1992) bilateral others index measure sources
Walker36 114 2–67 HD, TF, TT, Trauma Ordinal score, + Not given +
(1994) TM, bilateral interval score:
distance
Gauthier- 89 24–87 TF, TT Vascular, Ordinal score, + Crutch, cane, frame +
Gagnon37 diabetes, tumour, interval score:
(1994) trauma distance
Sapp38 132 23–85 TF, TT Not given Not given + Cane, quad cane, –
(1995) crutches, walker
Datta39 53 60–80 TF, TT Vascular, Ratio scale: time + Walking aids –
(1996) diabetes, trauma
Legro40 114 20–87 TF, KD, TT, Vascular, tumour, Visual analogue + Not given –
(1998) Syme trauma, scale
congenital prosthesis
evaluation quest
Traballesi41 144 mean TF Vascular, Ordinal score + Walking aids +

Downloaded from [Link] at Maastricht University on June 23, 2014

(1998) 68 ± 10 diabetes Rivermead
Mobility Index

Amputation level: HD, hip disarticulation; TF, transfemoral; KD, knee disarticulation; TT, transtibial; GS, Gritti Stokes; TM, transmalleolar.
a
Refers to a scale previously used by another author as stated.
b
Based on scale of Russek (1961).10
c
Based on scale of Hanspal and Fisher (1991).16
d
Based on scale of Volpicelli et al. (1983).6
Mobility of lower limb amputees
95
96 GM Rommers et al.

fication,10 the Nottingham extended ADL items in a seventh class at the end of the contin-
index,3 or the Frenchay Activity Index.39 uum (Bed).
• Use of aids The walking aids used for mea- The list of authors is in the same sequence as
surement of the mobility scale or stated in the in Table 1. Because the study of Siriwardena and
questionnaire are given. If the item activity Bertrand is chosen as an inventory for the other
with or without a prosthesis is stated, this is studies, we put it at the top of the list.
especially noted, since mobility without a Some authors used the classification as used in
prosthesis for self-care is of vital importance the WAI and are put together in the same line
for individual ADL. of the table. Others used the scale used by Volpi-
• Stairs The item ‘climbing stairs’ is especially celli et al.6 or Russek10 and are put together in the
noted if included in the questionnaire since same line as the original author. The original
climbing stairs is one of the most demanding numbering of the scales by the different authors
tasks for the lower limb amputee and is there- was in sequence of 1–6 upwards, with best mobil-
fore noted separately. ity grade 1 to worst mobility grade 6; or 3–1 or
Table 2 gives an overview of the range of 6–1 downwards with best mobility grade 6 to
measurement of the different mobility scales as worst mobility grade of 1 in the individual stud-
given in Table 1a.6–9,11–25 In order to give good ies. In order to give a clear overview, the indi-
comparison we used the study by Siriwardena vidual scale gradations are replaced by dots. The
and Bertrand14 to classify the different mobility dots are placed in the corresponding classes
scales. This Walking Ability Index (WAI) scale related to the WAI.
was designed to measure the ability of the After WAI 6 (wheelchair use) Wolf et al.,13
amputee to cross a distance of 10 feet (3 metres) Pinzur et al.12 and Hepp et al.17 include items
in an ordinary room with the use of walking towards ‘wheelchair use with assistance’ and an
aids if necessary. The amputee shows a normal item ‘fully bedridden’. The total panorama of
moving pattern, as practised at home. WAI 1 is mobility from ‘fully mobile’ with a prosthesis
fully mobile with a prosthesis and normal towards ‘totally confined to bed’ is covered.
walking pattern. WAI 2 is fully mobile with a In the scales without the specific use of walk-
prosthesis and an abnormal walking pattern, ing aids, as in Pinzur et al.,12 Wolf et al.13 and
but without any walking aid. WAI 3 is mobility Hepp et al.17 there is a sliding mobility scale.
with one cane or crutch. WAI 4 is mobility with There is no sharp distinction between the sub-
two canes or crutches. In this item, we also classes possible towards the classes defined by the
included a delta roller or a rollator. WAI 5 is WAI. Therefore we also used the subclasses in
mobility with a frame. WAI 6 is unable to order to cover the actual mobility range as given
cross the 10 feet other than with a wheelchair. by the mentioned studies. It gives valuable infor-
The advantage of this scale is that it covers mation about the mobility of lower limb
the actual performance and not what the amputees and is therefore included. As we stated
amputee could or should perform at maximum earlier in the qualitative analysis section the
endurance. scales comprise ordinal scaled classes. All 12
In Table 2 the first line shows the continuous scales but one8 start with the item ‘mobility with-
line of mobility of an amputee from fully mobile out a walking aid’ and all studies include class VI
with a prosthesis without a walking aid towards ‘mobility with a wheelchair’ at the right-hand end
completely bedridden without a prosthesis. This of the range of measurement.
is a continuous line without intervals or sub-
classes. The second line shows the individual Qualitative analysis
items of the WAI by Siriwardena and Bertrand.14 In order to analyse the different studies in a
After WAI 6 (wheelchair use) Pinzur et al.,12 qualitative way we compared the studies in Table
Wolf et al.13 and Hepp et al.17 include items 1. We tried to find a mobility scale with a maxi-
towards wheelchair use with assistance and an mum scale range and the best detail in measur-
item ‘fully bedridden’. Because there was no ing the separate levels of mobility of lower limb
actual use of the prosthesis we included these amputees. The inventory of the literature showed
Downloaded from [Link] at Maastricht University on June 23, 2014
 Table 2 Comparison of scales working towards a continuous mobility scale from fully mobile with a prosthesis without walking aid towards totally
bedridden

Fully mobile with prosthesis Bedridden

No aid No aid 1 cane/crutch 2 canes/crutches Walker Wheelchair Bed

normal walking abnormal walking frame

Siriwardena14 I II III IV V VI

Volpicelli6 • • • • • •
Johnson23/Datta19
Narang7 • • • • •
Helm8 • • • •
Russek10 • • • • • •
Kullman9/Stern11
Pinzur12 • • • • • • •
Wolf13 • • • • • • • •
Pohjolainen15 • • • • • • • •
Hanspal16 • • • • • •
Lachmann21
Campbell20/Houghton18
Kanellopoulos24

Downloaded from [Link] at Maastricht University on June 23, 2014

Hepp17 • • • • • • • • •
Zijp20 • • • • • •
Burger25 • • • • •
Mobility of lower limb amputees
97
98 GM Rommers et al.

studies with seven levels of mobility12,13,17 as the amputation level is also included in the studies
widest measuring range. This covered the full by other authors.7,14,24,25,34,40 Special studies
range of mobility from ‘fully mobile with an arti- including all amputation levels in the lower limb
ficial limb without aids’, to ‘totally confined to are limited: only the study by Walker et al.36
bed’. includes all amputation levels. Two studies focus
In order to add detail between these seven lev- on bilateral amputees only.6,19 Nineteen stud-
els of mobility we subdivided the seven classes ies6–9,11–25 give separate levels of mobility. The
into six subclasses. This enabled us to refine the scales have distinct levels of measurement and
table and to include more detailed descriptions are ordinal scaled.43
of the mobility items used in the scales studied. Several authors9,11,13,19,21–25 use a scale previ-
Since all the scales were ordinal, we spaced the ously used by other authors. Datta et al.19 and
classes equally. In this way we were able to Johnson et al.23 also use the scale by Volpicelli et
compare the measuring range of the individual al.6 Datta et al. used the same classification but
studies. defined walking distance in feet instead of using
In our opinion, by conducting this qualitative the original item – walking blocks. Pohjolainen
analysis, we give a reliable synopsis of the mea- and Alaranta15 used the scale designed by Narang
surement scales studied. et al.7 but extended it with two categories to iden-
tify indoor and outdoor mobility. In the study by
Reliability and validity Kullmann9 as well as the scale used by Russek,10
The scales all measured mobility of lower limb a Barthel score was used to measure ADL. If the
amputees. None of the studies studied the inter- original Russek score was used, only four items
or intra-observer reliability. Test–retest reliabil- of the Barthel score could be found. The study
ity was carried out in several studies. Measure- by Hanspal and Fisher16 used the Stanmore
ments were performed several times on the same Harold Wood mobility scale, later used by
artificial limb users over time. This recorded any several other authors in the UK.18,21,22,23 ‘Using
change in mobility over time with the same stairs’ is included in the scale developed by
mobility test. Volpicelli et al.,6 and used by Datta et al.19 and
The construct of a seven-class instrument for Johnson et al.23
measuring mobility was designed because no ade- Table 1b26–41 gives an overview of the mobility
quate measurement scale existed that covered the of lower limb amputees without a distinct scale
wide range of mobility of artificial limb users. measuring mobility. This is done a less specific
None of the studies used statistical analysis to way than in the previous studies of Table 1a. In
compare the mobility measurements used in the the questionnaires, items about mobility are
studies. often included. For example, questions used may
include the number of hours of prosthetic use,
walking speed and time to reach a specific dis-
Results tance or a visual analogue scale (VAS) to
measure walking possibilities.40 This gives
We found 35 studies in recent literature between information about mobility but is not compara-
1978 and 1998 (Table 1). ble with the previously mentioned scales.6–9,11–25
The population studied varies considerably. Traballesi41 used the Rivermead Mobility
Nine of the studies (25%) included more than Index.48 This scale is used to measure mobility
200 patients, with a range of 210–2400. As but was developed for patients with head injury
expected, most studies have a majority of elderly and stroke and not for artificial limb users.
vascular amputees. Studies including trauma or Climbing stairs with a prosthesis is a demand-
tumour amputees have a wider age range and ing task and in the Barthel Index5 this is the
include more people under 60 years of age. final and most demanding item. In 14 stud-
Almost all studies include transfemoral (TF) ies3,6,25–27,29,30,32–34,36,38,41 this item is included, but
amputees. The study by Pinzur12 focuses primar- only Volpicelli et al.6 used this item in the mea-
ily on knee disarticulation (KD) amputees. This surement scale for mobility. The other studies
Downloaded from [Link] at Maastricht University on June 23, 2014
 Mobility of lower limb amputees 99

inquired about this item in the additional ques- lower limb amputees in the studied literature and
tionnaire or in the ADL index. (2) to compare the range of measurement of the
Table 2 compares range of mobility measured mobility scales. Table 1 gives the review of the
by the studies.6–9,11–25 All but one study start with literature between 1978 and 1998 and Table 2
normal walking without walking aids (WAI I). shows the different ranges of measurement of the
The study by Helm et al.8 starts with the item: individual scales studied. The scales by Pinzur et
‘Patient wears prosthesis all day, walks alone al.,12 Wolf et al.13 and Hepp et al.17 had the widest
even outdoors. At times uses one cane outdoors ranges of measurement.
but not indoors. Does not use a wheelchair’. We The study of Hanspal and Fisher16 used the
therefore put the dot in between class II and III. Stanmore Harold Wood mobility scale. Several
Russek10 and Hepp et al.17 use the item: ‘Walk other authors in the UK18,21,22,24 used this scale,
with prosthesis with a walking aid’, but not stat- and it was (with five publications) the most fre-
ing the quality of walking. We therefore included quently published scale of all the evaluated stud-
an extra dot in between I and II. To include the ies. If the item ‘Using stairs’ is preferred, the
item: ‘Use of wheelchair with assistance’ we put scale from Volpicelli et al.6 is the best to use. We
an extra dot between WAI VI and think that this item is important because it is the
‘Bed’.6,13,15,16,18,21,22,24 highest achievement in indoors prosthetic use.
In the most extensive used scale by Hanspal We found that a multitude of measurement
and Fisher16 and others18,21,22,24 for the item scales and questionnaires are available but that
‘Wears prosthesis only for transfers or to assist they differ in methods and range of measure-
nursing; walks only with a carer’, an extra dot is ment. Several difficulties were faced:
placed in between WAI V and VI.
1) There are difficulties in measurement. All
items used were ordinal scaled. This implies
Discussion that items in the scale stand in some kind of
relation to each other. There is no true zero
This study gives an overview of mobility scales point and the intervals between the items are
for lower limb amputees. A complete literature not equal. Most of the presently used dis-
search of the electronic literature databases Med- ability and health status measures are of this
line (from 1978 to 1998) and Embase (from 1988 type.37 For comparison of the scales, a more
to 1998) was carried out. or less arbitrary interval is chosen. It limits
Our primary goals in this study were: (1) to the possibility of testing and there is limited
give a review of different mobility scales for statistical analysis possible. In order to solve
this problem we used in Table 2 a seven-class
measure with a maximum range of measure-
Clinical messages ment. In this way, we tried to give an accu-
rate description and comparison of the
• Measurement scales of mobility of lower individual studies. We realize that this is an
limb amputees differ considerably in range effort to solve measurement problems and we
of measurement and are only partially made a compromise towards adjustments of
comparable. We need to establish a mobil- the individual scales. A continuous measure-
ity scale with a wide range of measurement, ment tool for mobility of the lower limb
with enough detail to actually measure the amputee is not available.
differences over time. More precise mea- 2) Functional mobility of lower limb amputees
surement techniques need to be developed can differ because of the additional health
to give a proper comparison of mobility status of the individual. Due to medical prob-
items. lems, mobility changes over time. One day a
• A real consensus about the use of mobility person may be able to walk with a stick and
scales of lower limb amputees is not avail- on another day a wheelchair may be needed
able in the recent literature. because of physical or prosthetic problems.
Measuring the mobility of an amputee is
Downloaded from [Link] at Maastricht University on June 23, 2014
100 GM Rommers et al.

therefore a sliding measurement over time. If detail to measure individual changes over time
a global division is made, as done by Russek10 for the individual artificial limb user. The avail-
or Volpicelli et al.6 with items such as ‘house- able studies cannot properly be compared.
hold walker’ and ‘a community walker’, it Therefore, we cannot select the most efficacious
gives some idea of mobility but can hardly be items and measurement scales for everyday reha-
compared with other studies. Burger et al.25 bilitation treatment.
studied the mobility of traumatic lower limb We need to establish a mobility scale with a
amputees and solved the problem by using a wide range of measurement, with enough detail
questionnaire but perceived the same prob- to actually measure the differences over time.
lems as mentioned above. The construct of a seven-class measurement scale
3) Mobility measurements of walking speed and is an effort to compare the studies in a more
hours of prosthesis use are interesting data, detailed way.
but give no actual information about the We conclude that in this study we found:
mobility of the individual. Mobility without 1) Thirty-five mobility scales for lower limb
the use of a prosthesis is even more restricted amputees. They differ considerably in range
and extensive use of a wheelchair is needed. of measurement and are only partly compa-
Measurement in this context cannot be com- rable to each other.
pared with results for walking amputees. 2) Measuring mobility by a scale has been
Most easy to compare are those items related shown to have limitations. Several authors
to walking aids but they are less informative have done extensive research but they all
about the total time of prosthetic use in measure only certain aspects of mobility.
mobility. Step counting, as used by Holden 3) The most frequently published mobility scale
and Fernie,44 gives information in a quantita- was the Stanmore Harold Wood mobility
tive way but does not connect activity and scale.16
mobility together. 4) None of the 35 studies presented give a con-
Continuous measurement of mobility with a pros- tinuous measurement of mobility. A continu-
thesis, as developed by Stam et al.45 and Buss- ous mobility registration instrument needs to
mann and Stam,46 can give additional information be developed.
about mobility during activities. With measuring 5) A real consensus about measurement of
devices on the patient, they are able to detect mobility of lower limb amputees is not avail-
changes in position of the body over time. In this able in the recent literature.
way a ‘’continuous’ mobility registration from
bed, transfer, sitting and walking is possible, Acknowledgements
together with prosthetic and wheelchair use. The authors would like to thank Mr RE Stew-
Measurement of movement in several direc- art, statistician, Northern Centre for Healthcare
tions can distinguish between making a transfer Research, University of Groningen, for analysis
from bed and chair, sitting and walking. Compa- of data. Mrs AC Gunter-Buitenkant, librarian of
rable studies have been done by Kochersberger the General Hospital de Tjongerschans, Heeren-
et al.,47 but they could only distinguish between veen for her kind support in collecting the liter-
poor, moderate and good mobility in an elderly ature. We would like to thank Rehabilitation
population. Friesland, Beetsterzwaag for their support.
In this article, we give an overview of the pub-
lished mobility scales for lower limb amputee
ambulation. We conclude that there is no ade- References
quate tool available for measuring mobility of
lower limb amputees. There is no consensus 1 Houtum van WH, Lavery LA. Regional variation in
the incidence of diabetes-related amputations in the
about the ideal measurement scale and a number Netherlands. Diabetes Res Clin Pract 1996; 31:
of authors have developed their own measure- 125–32.
ment system. The measurement system should 2 Rommers GM, Vos LDW, Groothoff JW, Schuiling
have a wide range of measurement with enough CH, Eisma WH. Epidemiology of lower limb
Downloaded from [Link] at Maastricht University on June 23, 2014
 Mobility of lower limb amputees 101

amputees in the north of the Netherlands: aetiology, een beenamputatie. Tijds Gerontol Geriatr 1992; 23:
discharge destination and prosthetic use. Prosthet 54–59.
Orthot Int 1997; 21: 92–99. 21 Lachmann SM. The mobility outcome for amputees
3 Collin C, Collin J. Mobility after lower limb with rheumatoid arthritis is poor. Br J Rheumatol
amputation. Br J Surg 1995; 82: 1010–11. 1993; 32: 1083–88.
4 Ebskov B, Ebskov L. Epidemiology. In: Murdoch G, 22 Campbell WB, Johnston JA St, Kernick VFM,
Bennett Wilson Jr. eds. Amputation: surgical practice Rutter EA. Lower limb amputation: striking the
and patient management. Oxford: balance. Ann R Coll Surg Engl 1994; 76: 205–209.
Butterworth-Heinemann, 1996: 23–29. 23 Johnson VJ, Kondziela S, Gottschalk F. Pre and
5 Mahoney FI, Barthel DW. Functional evaluation: post-amputation mobility of trans-tibial amputees:
the Barthel index. Md Med J 1965; 14: 61–65. correlation to medical problems, age and mortality.
6 Volpicelli LJ, Chambers RB, Wagner Jr, FW. Prosthet Orthot Int 1995; 19: 159–54.
Ambulation levels of bilateral lower-extremity 24 Kanellopoulos G, Sabaharwal A, MacGregor C,
amputees. J Bone Joint Surg 1983; 65A: 599–605. Cooper GG, Engeset J. Major lower limb
7 Narang IC, Mathur BP, Singh P, Jape VS. amputation for vascular disease in the Grampian
Functional capabilities of lower limb amputees. area: the outcome of rehabilitation. J R Coll Surg
Prosthet Orthot Int 1984; 8: 43–51. Edinburgh 1996; 41: 114–15.
8 Helm P, Engel T, Holm A, Kristiansen VB, 25 Burger H, Marincek C, Isakov E. Mobility of
Rosendahl S. Function after lower limb amputation. persons after traumatic lower limb amputation.
Acta Orthop Scand 1986; 57: 154–57. Disabil Rehabil 1997; 19: 272–77.
9 Kullmann L. Evaluation of disability and of results 26 Kegel B, Carpenter ML, Burgess EM. Functional
of rehabilitation with the use of the Barthel index capabilities of lower extremity amputees. Arch Phys
and Russek’s classification. Int Disabil Stud 1987; 9: Med Rehabil 1978; 59: 109–20.
68–71. 27 Day HJB. The assessment and description of
10 Russek AS. Management of lower extremity amputee activity. Prosthet Orthot Int 1981; 5:
amputees. Arch Phys Med Rehabil 1961; 42: 23–28.
687–703. 28 Steinberg FU, Sunwoo I, Roettger RF. Prosthetic
11 Stern PH. Occlusive vascular disease of lower limbs: rehabilitation of geriatric amputee patients: a
diagnosis, amputation surgery and rehabilitation. Am follow-up study. Arch Phys Med Rehabil 1985; 66:
J Phys Med Rehabil 1988; 67: 145–54. 742–45.
12 Pinzur MS, Smith DG, Daluga DJ, Osterman H. 29 Beekman CE, Axtell LA. Prosthetic use in elderly
Selection of patients for through knee amputation. J patients with dysvascular above-knee and
Bone Joint Surg 1988; 70A: 746–50. through-knee amputations. Phys Ther 1987; 67:
13 Wolf E, Lilling M, Ferber I, Marcus J. Prosthetic 1510–16.
rehabilitation of elderly bilateral amputees. Int J 30 Lavan J. Rehabilitation of the elderly amputee. Irish
Rehabil Res 1989; 12: 271–78. Med J 1991; 84: 91–93.
14 Siriwardena GJA, Bertrand PV. Factors influencing 31 Chan KM, Tan ES. Use of lower limb prosthesis
rehabilitation of artheriosclerotic lower limb among elderly amputees. Ann Acad Med Singapore
amputees. J Rehabil Res Dev 1991; 28: 35–44. 1990; 6: 811–16.
15 Poljolainen T, Alaranta H. Predictive factors of 32 Brodzka WK, Thornhill HL. Long term function of
functional ability after lower limb amputation. Ann persons with atherosclerotic bilateral below knee
Chir Gynaec 1991; 80: 36–39. amputation living in the inner city. Arch Phys Med
16 Hanspal RS, Fisher K. Assessment of cognitive and Rehabil 1990; 71: 895–900.
psychomotor function and rehabilitation of elderly 33 Collin C, Wade DT, Cochrane GM. Functional
people with prostheses. BMJ 1991; 302: 940. outcome of lower limb amputees with peripheral
17 Hepp W, de Jonge K, Palenker J. Rehabilitation- vascular disease. Clin Rehabil 1992; 6: 13–21.
sergebnisse des amputierten GefäBpatiënten. 34 Hagberg E, Berlin ÖK, Renström P. Function after
Orthopaedietechnik 1991; 8: 566–70. through-knee compared with below-knee and above
18 Houghton AD, Taylor PR, Thurlow S, Rootes E, knee amputation. Prosthet Orthot Int 1992; 16:
McColl I. Success rates for rehabilitation of 168–73.
vascular amputees: implications for preoperative 35 Nissen SJ, Newman WP. Factors influencing
assessment and amputation level. Br J Surg 1992; reintegration to normal living after amputation. Arch
79: 753–55. Phys Rehabil Med 1992; 73: 548–51.
19 Datta D, Nair PN, Payne J. Outcome of prosthetic 36 Walker CRC, Ingham RR, Hullin MG, McCreath
management of bilateral lower limb amputees. SW. Lower limb amputation following injury: a
Disabil Rehabil 1992; 14: 98–102. survey of long-term functional outcome. Injury 1994;
20 Zijp EM, Rasenberg EMC, van den Bosch JSG. 25: 387–92.
Reactivering in een verpleeghuis bij patienten met 37 Gauthier-Gaganon C, Grisé M-C. Prosthetic profile
Downloaded from [Link] at Maastricht University on June 23, 2014
102 GM Rommers et al.

of the amputee questionnaire: validity and reliability. University Press, 1995.

Arch Phys Med Rehabil 1994; 75: 1309–14. 44 Holden JM, Fernie GR. Extent of artificial limb use
38 Sapp L, Little CE. Functional outcomes in a lower following rehabilitation. J Orthop Res 1987; 5:
limb amputee population. Prosthet Orthot Int 1995; 562–68.
19: 92–96. 45 Stam HJ, Eijskoot F, Bussmann JBJ. A device for
39 Datta D, Ariyaratnam R, Hilton S. Timed walking long term ambulatory monitoring in transtibial
test – an all-embracing outcome measure for lower amputees. Prosthet Orthot Int 1995; 19: 53–55.
limb amputees? Clin Rehabil 1996; 10: 227–32. 46 Bussman JBJ, Stam HJ. Validity and reliability of an
40 Legro MW, Reiber GD, Smith DG, Del Aguila M, instrument for long-term ambulatory measurement
Larsen J, Boone D. Prosthesis evaluation of mobility activities. Proceedings of the 5th
questionnaire for persons with lower limb European Congress on Research in Rehabilitation,
amputations: assessing prosthesis-related quality of 28 May–1 June 1995, Helsinki, Finland. J Rehabil Sci
life. Arch Phys Med Rehabil 1998; 79: 931–38. 1995; 8: 65.
41 Traballesi M, Brunelli S, Pratesi L, Pulcini M, 47 Kochersberger G, McConnell E, Kuchibhatla MN,
Angioni C, Paolucci S. Prognostic factors in Pieper C. The reliability, validity, and stability of a
rehabilitation of above knee amputees for vascular measure of physical activity in the elderly. Arch
diseases. Disabil Rehabil 1998; 20: 380–84. Phys Med Rehabil 1996; 77: 793–95.
42 Murdoch G, Jacobs NA, Wilson Jr, AB. Report of 48 Collen FM, Wade DT, Robb GF, Bradshaw CM.
ISPO consensus conference on amputation surgery. The Rivermead Mobility Index: a further
University of Strathclyde; ISPO Denmark 1992. development of the Rivermead Motor Assessment.
43 Bowling A. Measuring health: a review of quality of Int Disabil Stud 1991; 13: 50–54.
life measurement scales. Milton Keynes: Open

Downloaded from [Link] at Maastricht University on June 23, 2014

882

Prosthetic Profile of People With Lower Extremity Amputation:

Conception and Design of a Follow-up Questionnaire
Marie-Claude L. Grik, MSc, Christiane Gauthier-Gagnon, MSc, Georges G. Martineau, BSc, LPED

ABSTRACT. GrisC MCL, Gauthier-Gagnon C, Martineau GG. Prosthetic profile of people with lower extremity
amputation: conception and design of a follow-up questionnaire. Arch Phys Med Rehabil 1993;74:862-70.
l A questionnaire, the Prosthetic Profile of the Amputee (PPA), is described both in terms of its conceptualization and
design. The objectives of the questionnaire (mail and telephone versions) were to evaluate and to determine the factors
potentially related to prosthetic use by a person with a lower extremity amputation after discharge from rehabilitation.
The PRECEDE theoretical model (Predisposing Reinforcing and Enabling Causes in Educational Diagnosis and Evalua-
tion) was adapted to the present study and served as a guide for the conceptual basis of the instrument. A multidisciplin-
ary group of 13 experts and another group of eight persons with lower limb amputation were involved in the identifica-
tion, classification, and priority rating of the factors related to prosthetic use. The design of the questionnaire was
structured according to Dillman’s Total Design Method. To ensure the content validity of the instrument, the pertinence
of each question with regard to the questionnaire’s objectives was assessed by the group of experts. Pretesting of the
questionnaire was then conducted with a random sample of 26 people with lower extremity amputations. The preliminary
version of the instrument has already demonstrated the potential for clinical and research use.
0 1993 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and
Rehabilitation

KEY WORDS: Amputation; Artifical limbs: Lower extremity; Questionnaire

In the past decade the annual incidence of lower extrem- medical supervision is considerably reduced, what becomes
ity amputation in the province of Quebec (Canada) has of the LEA subject? In Quebec, approximately 1,100 indi-
averaged 34 per 100,000 inhabitants.’ As reported in other viduals with LEA require prosthetic fitting and training an-
countries,* the incidence of this surgical intervention has nually.’ Following their discharge from the rehabilitation
remained relatively stable over the years, despite improved center, little information is documented regarding the fol-
health education and advances in medical and surgical tech- low-up status of their prosthetic use. Does the LEA subject
nologies. 1,2The aging phenomenon observed in the general continue to wear the prosthesis and what are the factors that
population is thought to be the primary explanation for this contribute to its use or disuse?
stability.’ In industrialized countries, peripheral vascular To address these questions and to determine the needs of
disease is responsible for more than 70% of the lower limb this clientele, follow-up research is of the utmost impor-
amputations, 3-8the majority performed on persons 60 years tance. As evidenced by a recent literature review of 15 stud-
of age and above. 6,g-1oThus, given the growing proportion ies10*15S28on prosthetic use by LEA subjects following dis-
of elderly, for which the frequency of vascular disease is the charge, the percentages of “successful prosthetic users”
highest among the different age groups in the popula- within each sample varied considerably, ranging from 47%
tion ‘J’ the lower extremity amputation (LEA) population to 96%. These varying results may be due to methodologi-
conhnues to be a priority for rehabilitation professionals, cal differences (eg, sample size and selection, evaluation
particularly in terms of their need for prosthetic training. instruments used) and semantic inconsistencies. The se-
Although prosthetic fitting of the older patient was once a mantic inconsistencies noted are mainly related to the
controversial issue, recent studies have shown positive reha- various interpretations of the concept of “successful pros-
bilitation outcomes at discharge12-14among this elderly pop- thetic use” at follow-up. For example, this concept has been
ulation. However, once the patient returns home, where the defined either as “daily use of the prosthesis,“10*‘5-17“daily
or weekly wearing of the prosthesis,“18,‘g “locomotor capa-
bilities with the prosthesis,“20-22 or “use of the prosthesis for
From EC& de rkniaptation, Facultt de medecine, (Ms. GrisC. Ms. Gauthier-
Gagnon), Insitut de Readaptation de Montreal (Ms. Gris&) Fact&C des sciences de locomotor activities.“23-28 Moreover, some authors have
l’education (Mr. Martineau), Universite de Montreal, Montreal, Quebec, Canada. used interchangeably the terms “capability to perform” and
Submitted for publication May 17, 1992. Accepted July 3. 1992.
This research supported by grant 87 1354 from the Fonds de la recherche en same “actual performance or use.” Yet, measures of capability
du Quebec, Montreal. Quebec. Canada. The Corporation Professionelle des Phy- are not necessarily indicative of the actual behavior.*’
siotherapeutes du Quebec also provided assistance in the form of a Gwen Gower Rees We believe that a measure of the “daily or weekly pros-
Bursary to Ms. Grik.
No commercial party having a direct or indirect interest in the subject matter ofthis thetic wearing” would depict more accurately the actual
article has or will confer a benefit upon the authors or upon any organization with behavior of the older LEA subject, because this definition
which the authors are associated.
Reprint request to Christiane Gauthier-Gagnon. MSc, Director ofthe Physiother-
could include the cosmetic user. Nonetheless, further dis-
apy Programme, Ecole de readaptation, Faculte de medecine. Universite de crimination between the different types of users (cosmetic
Montreal. C.P. 6 128. succursale A, Mont&al. Quebec. Canada H3C 337. versus functional) should be measured in terms of the ac-
10 1993 by the American Congress of Rehabilitation Medicine and the American
Academy of Physical Medicine and Rehabilitation tual locomotor activities performed with the prosthesis. In
0003-9993/93/7408-0223$3.00/O addition, the identification of factors encouraging or dis-

Arch Phys Med Rehabil Vol74, August 1993

 PROSTHETIC PROFILE OF THE AMPUTATION SUBJECT, Grid 883

couraging prosthetic use could offer a better appreciation of (questionnaire) to be used for follow-up studies on LEA
the frequency of prosthetic wearing and of the functional subjects, specifically to better understand the use of prosthe-
locomotor level obtained with the prosthesis. sis by LEA subjects in the province of Quebec and to iden-
Several authors have attempted to identify the factors tify the factors related to prosthetic use after discharge from
associated with the use of the artificial limb by the LEA a rehabilitation center.
subject. Through examination of anecdotal or descriptive This article presents the Prosthetic Profile of the Ampu-
data, several factors have been identified; however, only a tee (PPA) questionnaire, in terms of its conception and de-
limited number have been found to be statistically related sign based on the aforementioned recommendations. Such
to specific prosthetic outcomes. 5.16.19,21,26,30,31 Research in
a questionnaire will eventually serve to ( 1) evaluate the pros-
this area is nascent and most of the studies focus primarily thetic use of the LEA population, in terms of the weekly
on the physical factors related to prosthetic use, the psycho- frequency of prosthetic wear and functional level of pros-
logical, social, and environmental factors being poorly rep- thetic use, and (2) identify the factors related to pros-
resented. thetic use.
More extensive and comprehensive studies of the factors
related to prosthetic use need to be undertaken. However, METHODOLOGY AND RESULTS
when the instruments used with the LEA population during A self-administered mail questionnaire, with its tele-
follow-up evaluations were reviewed, no appropriate evalu- phone version, were designed to evaluate the factors related
ation tool was found that could evaluate the frequency and to prosthetic use or nonuse by LEA subjects. A user was
functional level of prosthetic use as well as identify the fac- defined as a person who wears a prosthesis at least once a
tors related to prosthetic use. Several of these instruments, week for either a functional or cosmetic purpose. The ques-
such as the Bar-the1 Index, PULSES profile, and ESCROW tionnaire was specifically developed for the adult unilateral
scale were selected by 0’Toole,32 to evaluate the functional above-knee and below-knee LEA subject who has com-
changes over time, in 60 vascular LEA subjects. However, pleted a prosthetic rehabilitation program and has returned
these instruments as well as other physical and multidimen- home. These persons represent 93% of the LEA population
sional functional assessment instruments, such as Katz’s fitted and trained with a prosthesis annually in the province
ADL index,33 the Kenny self-care evaluation,34 the Func- of Quebec.’
tional Independence Measure,35 the Functional Assessment Benson and coworkers43 describe four phases necessary
Inventory, 36and the Level of Rehabilitation Scale,” do not for the development of an evaluation instrument: the plan-
discriminate between prosthetic users and nonusers. ning, the construction (including content validity), the reli-
Hence, wheelchair users or unipedal walkers may obtain ability, and the validation phases. This paper presents the
the same global scores as prosthetic bipedal walkers. first two phases of the development of the PPA question-
Other instruments, specifically designed for the LEA pop- naire. The subsequent phases are the scope of a second
ulation, include the stepcounter’ and Day’s Activity paper (Part II) in preparation. Because each subsequent
Score.38 Step counting, using a miniature electronic counter methodological step depends on the results of the preceding
attached to the prosthesis, provides an objective and accu- one, the methodology and results sections of this paper
rate measure of the functional aspect of prosthetic use, but have been combined.
does not totally describe a patient’s activity level.5 This type
of instrumentation can also be costly when applied to a PLANNING PHASE-FACTOR IDENTIFICATION
large population. Day’s Activity Score provides more de-
tailed information about LEA subjects’ locomotor activi- During the planning phase, factors potentially related to
ties, yet the evaluation report sheet has to be completed by prosthetic use were identified by health professionals and
an interviewer. However interesting, these two instruments LEA subjects through the use of the PRECEDE (Predispos-
do not evaluate the factors related to prosthetic use. ing Reinforcing and Enabling Causes in Educational Diag-
Recently, for reasons of cost-efficiency,3g,40 mail and tele- nosis and Evaluation) theoretical model.
phone questionnaires have been preferred’6*‘8~2’*27~28,30~4’,42
to the more conventional face to face interviews used in Theoretical Model
Day’s study. Although questionnaires are now considered The PRECEDE model was originally developed by
valuable alternative instruments, most of the previously Green and coworkers44 as a framework for the planning.
mentioned mail and telephone questionnaires provide in- implementation, and evaluation of health education pro-
complete information about the factors related to pros- grams.45-48For our study, the model was adapted and used
thetic use, and some present methodological limitations to objectively identify all factors potentially related to the
that give rise to certain biases. use of prostheses by LEA subjects.
To control for methodological biases when developing a The interest of the PRECEDE model resides in its “diag-
questionnaire. proper procedures must be respected. When nostic chain theory,” which includes three categories of fac-
appropriate, a conceptual framework should be considered, tors assumed to have the potential for affecting voluntary
the content of the questionnaire should incorporate ex- health behaviors and assumed to be modifiable by educa-
perts’ considerations, and recognized design techniques tional interventions.
should be followed. In this study, the specific voluntary behavior was defined
As the first part of a larger ongoing research project, our in terms of prosthetic use. Because behavior is a multifac-
objective was to develop a suitable evaluation instrument eted phenomenon, it is seen as a function of the collective

Arch Phys Med Rehsbil Vol74, August 1893

884 PROSTHETIC PROFILE OF THE AMPUTATION SUBJECT, Gris6

influence of three categories of factors: predisposing, en- each subgroup, some of the rankings were changed. The
abling, and reinforcing factors. Although not mutually ex- final results from the two subgroups were then collected
clusive, these factors are described by Green44 as follows: and compared. When the rankings were not identical, the
Predisposing factors are factors antecedent to behavior numbers attributed to the factors were again combined and
that provide the rationale or motivation for the behavior. a new order of priority was established.
included are knowledge, attitudes, beliefs and values. These procedures were repeated for all 12 subcategories.
[Also included in this category are health status and so- Because it is believed that participants can prioritize only
ciodemographic characteristics. ] Enabling factors are fac- five to nine factors at a time with some degree of reliabil-
tors antecedent to behavior that allow a motivation or ity,4g each subcategory was presented on a separate sheet of
aspiration to be realized. Included are personal skills, [re- paper and included a maximum of nine factors. Following
quired for successful performance of the behavior] and these prioritizations, the least relevant factors (eg, domi-
resources as well as community resources. Reinforcing nant side, ethnic group, LEA subject’s perception of health
factors are factors subsequent to behavior that provide condition, care given to the prosthesis) were discarded.
the continuing reward, incentive or punishment for a be- Once the prioritizations within the subcategories were
havior and contribute to its persistence or extinction. In- completed, the subcategories were rank ordered in a sim-
cluded are social as well as physical benefits and tangible ilar way.
as well as imagined or vicarious rewards. Method 2: Informal method with a group of LEA
subjects. Eight LEA subjects representative of the target
population who had completed a prosthetic program at the
Methods of Factor Identification Montreal Rehabilitation Institute, were interviewed individ-
Following the model’s recommendation, three methods, ually. Face-to-face or nonstructured telephone interviews
both formal and informal, were used to identify the factors were conducted by a research member. Questions were
potentially related to prosthetic use and to sort them in open-ended regarding opinions about the advantages/dis-
order of priority into the predisposing, enabling, or rein- advantages to using the prosthesis, the difficulties faced
forcing categories or subcategories. with the prosthesis once returned home, and the present
Method 1: Informal method with a group of experts. A level of prosthetic use. These interviews helped identify the
group of 13 experts (both anglophones and francophones), familiar vocabulary used by the target population.
composed of health professionals from six different medical Method 3: Formal method with the LEA population. A
facilities in the Greater Montreal area and two representa- literature search of studies reporting the prosthetic use of
tives from a provincial LEA support group, was formed. the LEA subjects guided the working sessions with the
The group included two physiatrists, two physiotherapists, group of experts and the interviews with the LEA subjects.
two occupational therapists, two prosthetists, a social
worker, a psychologist, a nurse, an individual with a vascu- Factors Identified as Potentially Related
lar LEA, and an individual with a traumatic LEA. to the Prosthetic Use by LEA Subjects
The group met four times with the first meeting devoted Tables 1,2, and 3 present a summary of the factors iden-
to generation of the factors related to the prosthetic use by tified using the three methods of identification described
the LEA subjects once they returned home. The verbal above. The factors are ranked in order of priority within
brainstorming technique was used. each of the 12 subcategories of factors.
Factors identified at the first meeting were categorized Table 1 presents the predisposing factors grouped into
and ranked at the following three meetings, using the nomi- “motivation,” “rehabilitation program,” “demographic
nal group technique, slightly modified from Delbecq’s ini- characteristics,” and “physical health status” subcategories.
tial description. 4gThe meaning of each factor was clarified
and their classification into 12 subcategories of the predis-
Table 1: Summary of the Predisposing Factors
posing, enabling, and reinforcing categories of factors was
Potentiallv Related to Prosthetic Use
discussed. Redundant factors were deleted, whereas others
were grouped together. Motivation Rehabilitation program (RP)
To prioritize the remaining 46 factors within their respec- I. Adaptation to I. Time interval between
tive subcategory, the following sequence of procedures was amputation amputation and admission to
2. Self-image* RP
used. The experts were individually allowed five minutes to 3. Adaptation to prosthesis 2. Preprosthetic training
numerically rank, in writing, the factors of a single subcate- Demographic characteristics 3. Status (inpatient or outpatient)
gory. The numerical value of 1 was assigned to the factor 1. Living arrangements during RP
considered most influential to the behavior of prosthetic 2. Place of abode 4. Duration of prosthetic training
use. The subsequent ascending numbers were attributed to 3. Current vocational status Physical health status
4. Education 1. Level of amputation
the other factors according to their relative importance. 5. Learning abilities 2. Cause of amputation
The experts were then divided into two subgroups to facil- 6. Socioeconomic level 3. Medical conditions
itate discussion of the individual factor rankings. Rankings 4. Condition of nonamputated
assigned to each factor were recorded and compiled. The Leg
5. Condition of stump
factor with the smallest numerical total represented the 6. Variation in body weight
most important factor. The preliminary list of rank-ordered
factors was discussed, and after reaching a consensus within * Factors deleted during the construction phase ofthe PPA questionnaire.

Arch Phys Med Rehabil Vol74, August 1993

 PROSTHETIC PROFILE OF THE AMPUTATION SUBJECT, Grise 865

Table 2: Summary of the Enabling Factors Potentially Table 4: Priority Order Established for the 12 Subcategories
Related to Prosthetic Use of Predisposing, Enabling, and Reinforcing Factors

Locomotor abilities with the Integration level of locomotor 1. Physical health status
prosthesis abilities 2. Motivation
I. Walking abihties (indoors, 1. Walking distance 3 Locomotor abilities with the prosthesis
outdoors) 2. Assistance (technical aids, 4: Integration level of locomotor abilities
2. Standing up and assistance or supervision 5 Rehabilitation program
maintenance of the required) 6: Satisfaction with the prosthesis
standing position 3. “automatism” of 7. Social environment
(standing balance versus prosthetic walking 8. Demographic characteristics
number of falls) 4. Walking speed (compared 9. Physical environment
3. Abiiity to don/doff the to wheelchair use or IO. Resources and accessibility tn resources
prosthesis unipedal walking) II. Social interaction
4. Walking abilities while Physical environment I’. Maintenance of physical status and abilities*
carrying objects (architectural barriers)
Resources and accessibility to I. Uneven ground * Subgroup of factors deleted during the construction phase of the PP,JI
resources 2. Inclement weather questionnaire.
1. Medical resources. 3. Distance to cover
community resources* 4. Stairs (especially without
2. Accessibility to different hand-rail)
questionnaire was primarily based on the predisposing, en-
types of transportation 5. Types of doors to open abling, and reinforcing factors presented in tables 1. 2, and
(in public places)* 3 with the exception of the following factors. Factors such
as “self-image, ” “learning abilities, ” “family role,” and “ac-
* Factors deleted during the construction phase of the PPA questionnaire.
curate perception of significant others concerning LEA
subject’s abilities with the prosthesis” (tables 1 and 3) had to
Table 2 presents the enabling factors classified under the be discarded because they constituted complex concepts
following subcategories: “locomotor abilities with the pros- that were difficult to operationalize in the context of a mail
thesis,” “integration of locomotor abihties,” “resources or telephone questionnaire. These factors were also the sub-
and accessibility to resources,” and “physical environ- ject of controversy among the group of experts who could
ment.” not agree upon a common definition; this being especially
In table 3, factors grouped as to “satisfaction with the true for the “family role.” This latter factor also led to com-
prosthesis, ” “social environment” and “interaction” and munication difficulties with the respondents.
“maintenance of physical status and abilities” subcategor- Furthermore, some of the factors that ranked last in the
ies of the reinforcing factors are presented. prioritization were eliminated to shorten the questionnaire.
Table 4 presents the 12 subcategories of factors listed in These included “maintenance of physical status and abili-
order of importance. ties,” “community resources,” and “types of doors to
open” (tables 2 and 3).
CONSTRUCTION PHASE- Examples of questions included in the PPA question-
THE PPA QUESTIONNAIRE naire to evaluate the predisposing, enabling, and reinforc-
ing factors are given in figure 1.
Content of the Questionnaire In addition to the factors being evaluated, the question-
The PPA questionnaire contains 44 questions and can be naire also includes questions pertaining to the behavior of
completed in approximately 25 minutes. The content of the prosthetic use. The actual prosthetic use was defined by the
researchers in terms of the weekly frequency of prosthetic
Table 3: Summary of the Reinforcing Factors wearing. This variable is calculated by multiplying the num-
Potentially Related to Prosthetic Use ber of hours the prosthesis is worn per day by the number of
days it is worn per week. To allow for discrimination be-
Satisfaction with the prosthesis Social environment
I. Discomfort, or unsatisfactory Significant others’ tween the functional and cosmetic user, two questions refer-
tit perception of the LEA ring to the functional level of prosthetic use were included
2. Appearance (cosmetic) subject’s abilities with the in the questionnaire. The latter variable was operational-
3. Problems related to the fixed prosthesis* ized as the daily percentage of displacements performed
prosthetic heel height Significant others’
4. Excessive stump perspiration acceptance of amputation
while wearing the prosthesis both indoors and outdoors.
5. Mechanical problems and prosthesis The questions that served to evaluate the prosthetic use by
(including noises from Familv role of the LEA LEA subjects are presented in figure 2.
prosthesis) subject* To reduce possible memory and knowledge biases from
6. Weight Physical help provided by the respondents, an information chart was also developed
1. Effect of prosthetic wear on significant other (family)
stump and phantom pain Maintenance of physical status that serves to gather precollected data from the respon-,
Social interaction and abilities dent’s medical chart. Therefore, data concerning the “reha-
I. Use of different types of 1. Medical follow-up.* bilitation program,” the “physical health status,” and “gen-
transportation medical compliance* eral prosthetic locomotor skills at the time of discharge”
2. Recreational and sporting 2. Regularity with home
activities exercises*
can then be collected before the questionnaire is adminis-
tered to the LEA subject. Although not considered influen-
* Factors deleted during the construction phase of the PPA questionnaire. tial factors by the group of experts and the LEA representa-

Arch Phys Med Rehabil Vol74, August 1993

866 PROSTHETIC PROFILE OF THE AMPUTATION SUBJECT, Gris6

4.4 At the present time, do you have any of the following problems with your SZWIW?

“Check q the appropriate &”

a) OFfasional stump pain ........................................
b) Constant stump pain ........................................... 1
c) Phantom pain (the missing part of your leg
is painful) ............._................................................ El
d) Wounds or sores ....................,.............................
e) Other problems: B
“specify”

Q.11 Whether or not you wear your prosthesis, at the present time, would you say that you are
“&&” to do the following activities “y$lTH YO-IS ON”?

“Check 0 fOr each statement *

NO YES, IF YES, IF SOMEONE w

SOMEONE IS NEAR ME ALONE
HELPS ME

a) Get up from a chair ..................... 0 ........... 0 ............ 0 ........................ 0

d) Walk in the house ....................... q ........... q .....__.....q ........................ q

e) Walk outside on even
ground .................,.......................... cl ........... Cl ............ 0 ........................ 0
I
m) Go down a few steps (stairs)
m a hand-rail ....................... cl ........... cl ............ Cl ...............*........ cl
n) Walk while carrying
q ..................**....cl
an object ...........*............................. Cl ........... 0 ,...........

QJ3 Thinking about the people who are close to you, can you indicate the answer which best
describes their acceptance of your...

“Check qw for each statements”

DONDTACCEPT ACCEPTD. ACCEPTR ACCEFT ,T ACCEPT IT

ITATALL SLIGHTLY MODERAlELY 0ulTE WELL COMPLETELY

a) ... amputation?.. q ................. Cl ........... 0 ................. Cl ................. Cl Fig l-Examples of questions from the
b) ... prosthesis? .... q ................. 0 ........... 0 ................. 17 ................. 0 PPA questionnaire pertaining to the fac-
tors potentially related to prosthetic use.

tives, information pertaining to age, sex, and description of associated with the behaviors of prosthetic use or nonuse,
the prosthesis have been included in the information chart. was reduced by the addition of a neutral introductory state-
ment in the “prosthetic use” section of the questionnaire,
which reassured the respondents that both behaviors were
Design of the Questionnaire
considered acceptable.
The PPA questionnaire was constructed following Dill- Furthermore, the measurement scales were diversified to
man’s Total Design Method. 5o A cover letter or introduc- minimize response set (checking the same response choice
tory text was included in the mail and telephone versions of in series independently of the questions asked).
the PPA questionnaire, which stated the purpose of the Sequence of questions. For design considerations and to
study for which the questionnaire will eventually be used, respect a logical sequence of inquiry, the questions included
assured confidentiality and mentioned the importance of in the PPA questionnaire were divided into six basic sec-
the respondent’s participation. Concise instructions on tions, where questions that were similar in content and type
how to complete the questionnaire were provided, and all were grouped together. These include “the physical condi-
the questions were worded in simple and familiar terms and tion,” “the prosthesis,” “the prosthetic use,” “the environ-
offered mutually exclusive and exhaustive alternative an- ment, ” “the leisure activities,” and “general information.”
swers. Appropriate skip patterns were used for the screening As recommended in the literature,5G52 the easy, salient,
questions, whereas appropriate transitional statements and nonthreatening questions (eg, problems with the non-
were used each time a new section was introduced to main- amputated leg and the stump, and the medical condition),
tam the respondents’ motivation in completing the ques- were placed at the beginning of the questionnaire, whereas
tionnaire. the more difficult and perhaps more threatening ones (eg,
The social desirability bias (perception of right or wrong), the locomotor abilities with the prosthesis and the actual

Arch Phys Med Rehabil Vol74, August 1993

 PROSTHETIC PROFILE OF THE AMPUTATION SUBJECT, Grid 867

Q.l2 For the following question, w&e the appropriate number in the designated space.

I wear my prosthesis day(s) a week.

I wear my prosthesis approximately hour(s) a day.

MofmacrlrcticlLFe

Q.14 During the day, when you have to move about INTHE up~~~&~ely what percen-
tage of your displacements are done ...
‘Check q f&r each statement ”

almost .................................................. almost

0% 25% 50% 75% 100%
(all your dis-
placements)

a) ... in a wheelchair? ........................ El . . . cl ... Cl . .. q . . cl

b) ... walking a your prosthesis
(technical oids can be used)? . 0 ...... 0 ..... q .. ...... 0 . . 0
c) ... walking m your

0 ...... q .. Cl .,...,...., 0 .,_.,...,..0

Q.16 During the day, when you have to move about OUTSIDE, 4ppmGna@ what percentage of
your displacements are done ...

“Check •! for each statement”

almost ................................................... almost

0% 25% 50% 75% 100%
(of your dis-
placements)

a) ... in a wheelchair? .....................,.. cl ...... Cl ..... Cl . .. . Cl . . . Cl

b) ...
Cl ...... 0 ..... 0 .. . Cl . . .. . El
c) ... walking w your
Fig 2-Questions from the PPA ques- prosth&s (rech&al aids CM
tionnaire pertaining to the behavior of be used)? .................................... 0 ...... 0 ..... 0 . . 0 . . 0
prosthetic use.

prosthetic use) were inserted in the middle of the question- gree of satisfaction regarding the prosthesis and the signifi-
naire. Demographic questions were placed at the end. cant others’ acceptance of the amputation and prosthesis
Type of questions and measurement scales. Ninety- (fig 1, Q.33). In these examples, the response categories
three percent of the questions were of closed-ended or semi- ranged from “not at all” to “completely satisfied or ac-
closed type. Although these types of questions are more cepted.” Ordinal scales were also used to evaluate the loco-
difficult to construct, they are easier to analyze, less de- motor abilities with the prosthesis. To operationalize this
manding to the respondent, and less subject to interviewer factor, a locomotor capabilities index composed of 14 dif-
and coder variance than open-ended questions.” ferent locomotor activities, selected primarily from the lo-
Various levels of measurement were used to operational- comotor disabilities classification ofthe World Health Orga-
ize the different information sought by the closed and semi- nization,53 was developed (fig 1, Q. 1 I). Four-point ordinal
closed questions. They included nominal, ordinal, and ra- scales, ranging from “not able” (numerical value of 0) to
tio scales of measurement. Moreover, the mail and “able to accomplish the activity alone” (numerical value of
telephone versions of the PPA questionnaire had to be de- 3). were used to evaluate the capability of the LEA subject
veloped simultaneously to ensure that the chosen types of to perform each of the activities. Furthermore, a composite
scales applied to both versions. measure representing the global locomotor abilities level of
Nominal scales with dichotomous (“yes or no”) and mul- the LEA subject was calculated by adding the individual
tiple choice answers were primarily used to gather factual scores assigned to each activity. for a possible maximum
information such as the nonamputated leg or stump condi- score of 42.
tion (fig 1, Q.4), the accessibility to resources, or the walk- Ratio scale measures, in percentages, were used to evalu-
ing aid required with the prosthesis. ate the daily proportion of displacements done with the
To better discern the qualitative intensity of a particular prosthesis, indoors and outdoors, compared to other means
response, ordinal scales were used. For example, five-point of displacement (fig 2, Q. 14 and 16). This type of quantita-
ordinal scales were used to measure the LEA subject’s de- tive measure was also used to evaluate the proportion of

Arch Phys Med Rehabil Vol74, August 1993

888 PROSTHETIC PROFILE OF THE AMPUTATION SUBJECT, Grist+

sedentary versus active activities in the LEA subject’s daily Finally, the four versions of the questionnaire (English,
routine. French, mail, and telephone) were preceded to facilitate
computerize data entry.
Translation of the Questionnaire
DISCUSSION
The LEA population in Quebec is composed of French
The absence of a recognized and valid evaluation instru-
speaking and English speaking persons and, therefore, the
ment of prosthetic use and the factors related to its use by
PPA questionnaire needed to be available in both lan-
the LEA subjects prompted the elaboration of a new instru-
guages. Preliminary translation from French into English
ment that would reflect the actual prosthetic profile of the
was done individually by two experts from the multidisci-
amputee subject following discharge from the rehabilita-
plinary group of experts, who were aware of the overall
tion center.
objective of the questionnaire and intent behind each ques-
The PRECEDE theoretical model proved to be a useful
tion. Using the back-translation method,54 a translator,
framework for the identification, categorization, and prior-
blind to the original questionnaire, translated the English
itization of the potential factors related to the LEA subjects
questions back into French. The back-translation version
behavior, specifically the use or nonuse of the prosthesis,
(French) was compared with the original one (French) and
Once the information, generated from the brainstorming
any discrepancies were noted. No questions needed to be
session with the experts, and the preliminary interviews
redrafted completely; however, the wording of some ques-
with the eight LEA subjects was coherently organized into
tions had to be restructured until both versions were judged
the three categories of factors proposed by the model, the
equivalent by the researchers. Although cross-language
elimination of redundant items became much easier. The
equivalence54 is also recommended when translating an in-
use of the nominal group technique also greatly facilitated
strument, the PPA questionnaire could not be tested using
the factor prioritization process and the deletion of the least
this method because fluently bilingual (French and English)
relevant factors identified. A disadvantage of the nominal
LEA subjects could be not recruited.
group technique, however, is that the rankings attributed to
the factors are based on individual subjective opinions. In
Qualitative Evaluations of the Questionnaire an attempt to compensate for this inherent subjectivity, the
Content validaiion. As part of the construction phase factor ranking was based upon a concensus of opinions
of the PPA questionnaire, the content of the instrument from experts.
was evaluated by the multidisciplinary group of experts Although efforts were made to include most of the factors
who met once again four months later. The questionnaire summarized in tables 1,2, and 3 in the questionnaire, oper-
was assumed to have content validity once all the experts ationalization and communication problems arising during
believed that the questions reflected the specific objectives the construction phase caused further elimination of some
of the questionnaire. 55The experts also commented on the factors. However, even following the elimination of these
relevancy of the questions, the appropriateness of the rating factors, the questionnaire was found to have content valid-
scales, and gave their overall appreciation of the instru- ity by the group of experts. They considered the domain to
ment. be adequately evaluated by the questionnaire, and that es-
Pre-test. The second qualitative evaluation consisted of sentially all priority subgroups of factors had been included
pretesting the PPA questionnaire with a random sample (N (table 4).
= 26) of anglophone and francophone LEA representatives The group of experts also approved of the operational
of the target population. Institutional approval was ob- definition for the concept of “prosthetic user” (eg, weekly
tained granting permission to consult the Montreal Rehabil- number of hours of prosthetic wearing). Similar definitions
itation Institute’s medical files of all subjects. The 18 men of “prosthetic user” have been previously reported by Med-
and eight women (mean age = 55.3) who had completed a hat’* and Beekman. l9 Dickstein’s statement that “func-
prosthetic training at the Montreal Rehabilitation Institute tional use of the prosthesis should not be the sole deter-
between May 1, 1985 and April 3 1, 1990 were asked to minant for evaluating the outcome of prosthetic
answer the questionnaire either in the presence of a re- rehabilitation” especially when training the older LEA sub-
search member or over the telephone. They were also asked ject, 24 further supp orts a definition of “prosthetic use,”
(when appropriate) to comment on the order of questions, which includes both cosmetic and functional users.
skip patterns, task difficulty, readability, and understanda- The questionnaires presently in existence for the LEA
bility of the questionnaire. In addition, while each subject population’6.19,*1,38,4* influenced the development of the
completed the questionnaire, their interest and attention PPA questionnaire; however, none of their particular ques-
levels were observed. In order to limit completion and com- tions were included in the present instrument in their origi-
prehension difficulties, modifications were subsequently nal form. Because the PPA questionnaire was entirely com-
made to the questionnaire, which included changing the posed of new material, the qualitative evaluations played a
question format from “numbers to be encircled” to “check- fundamental role during the construction phase. Once the
ing boxes” and changing the formulation and presentation content validity was established by the group of experts,
of some of the questions and response choices. further observations and criticisms were obtained from pre-
Based on the comments from the experts and LEA repre- testing of the questionnaire with a sample of 26 LEA sub-
sentatives, during the two qualitative evaluations, the ques- jects. Based on their recommendations, modifications
tionnaire was deemed to have “face validity.“55 made to the questionnaire contributed to the reduction of

Arch Phys Med Rehabil Vol74, August 1993

 PROSTHETIC PROFILE OF THE AMPUTATION SUBJECT. Grist2 889

comprehension distorsion and enhanced accuracy of re- validity of the PPA questionnaire has been established by a
sponses, thereby improving both the communication with group of experts, and the pretesting of the instrument has
the target population and the quality of information re- demonstrated that the questionnaire is simple and can be
ported by respondents. Furthermore, because the vascular easily completed by the LEA population.
LEA subject often suffers from associated visual deficits, A copy of the PPA questionnaire is available upon re-
the letter printing of the questionnaire was changed to one quest.
with a superior readability.
Acknowledgments: The authors wish to acknowledge Brigitte Maheux.
The quality of the information, obtained from either
MD, PhD, and Johanne Otis, MSc. as scientilic advisers. Bonnie Swaine.
mail or telephone questionnaires, is often considered to be MSc. for reviewing the manuscript, as well as the health care professionals
linked to the sociodemographic status of the respondents. and amputee subjects involved in the conception and qualitative evalua-
However, results from Siemiatycki’s studies39140on the qual- tions of the questionnaire. Appreciation is also extended to the Montreal
ity of responses, substitution, and nonresponse biases in Rehabilitation Institute for its collaboration in recruiting subjects.
mail and telephone survey methods, did not demonstrate References
persuasive differences in the validity of responses among I. Minis&e de la santC et des services sociaux. (Ministry of “Health and
socio-demographic subgroups. The results further demon- Welfare”) R&e de I’assurance-maladie du Q&&c. Statistiques I%-
strated that a substitution bias (proxy respondents answer- 1989. Bureau de la statistique du Quebec. QuCbec (QuCbec) 1990.
2. Stern PH. The epidemiology of amputations. Phys Med Rehabil Clin
ing the questionnaire instead of self-respondents) seemed to
North Am 1991:2:253-61.
be present only in the case of telephone surveys. Finally, for 3. Pohjolainen T. Alaranta H. WikstrBm J. Primary survival and pros-
both survey methods, Siemiatycki39 found no evidence of thetic fitting of lower limb amputees. Prosthet Orthot Int lYXY:13:63-
nonresponse bias using the mail or telephone strategies. Y
Mail or telephone strategies, strongly recommended in the 4. Kullman L. Evaluation of disability and of results of rehabilitation
with the use of the Barthel Index and Russek’s classification. Int Dis-
literature,50,52 include the original mailing or call to the re- abil Stud I987:9:68-7 I.
spondents. followed by follow-up mailings or calls. 5. Holden JM, Fernie CR. Extent ofartificial limb use following rehabili-
The questionnaire, originally developed in French, was tation. J Orthop Res 1987:5:562-8.
translated into English to ensure that data could also be 6. Ebskov B. Choice of level in lower extremity amputation-nation-
wide survey. Prosthet Orthot Int 1983:7:58-60.
obtained from the English LEA population of Quebec. A
7. Boontje AH. Major amputations of the lower extremity for vascular
telephone version of the PPA questionnaire was also devel- disease. Prosthet Orthot Int 1980:4:87-9.
oped, to eventually allow us to contact those respondents 8. Stern PH. Skudder PA. Amputee rehabilitation. I. lower-limb ampu-
who are less motivated to complete the mail version of the tations. N Y State J Med 1977:77:1436-40.
questionnaire, and respondents who cannot complete the 9. Kay HW. Newman JD. Relative incidence of new amputations. statis-
tical comparisons of 6000 new amputees. Orthot Prosthet 1975:29:3-
questionnaire due to certain limitations (eg, blindness, illit- 16.
eracy). The effectiveness of a telephone call as a means of 10. Finch DRA, MacDougal M. Tibbs DJ. Morris PJ. Amputation for
additional follow-up to mail survey has also been demon- vascular disease: the experience ofa peripheral vascular unit. Br J Surg
strated.50 1980:67:233-7.
Even though it is believed that possible methodological I I. Hierton T. James U. Lower extremity amputation in Uppsala county
I947- 1065,. Incidence and prosthetic rehabilitation. Acta Scaqd
biases related to the construction of the questionnaire have 1973:44:573-82.
been accounted for, the PPA questionnaire may present 12. Cruts HEP, De Vries J. Zilvold G. Huisman K. Van AlstC JA. Boom
certain limitations. The questionnaire has been developed HBK. Lower extremity amputees with peripheral vascular disease:
specifically for the adult unilateral above and below-knee graded exercise testing and results of prosthetic training. Arch Phys
Med Rchabil 1987:68: 14-9.
LEA subject and can only be used in follow-up studies. The 13. Couch P. David JK. Tilney NL. Crane C‘. Natural history of the Icg
PPA questionnaire, however. was developed to meet the amputee. Am J Surg 1977:133:46Y-73.
specific objectives of an ongoing provincial survey (Qdbec, 14. Kertstein MD. Zimmer H. Dugdale FE, Lerner E. Amputations ofthe
Canada) regarding the prosthetic use and the factors related lower extremity: a study of 194 cases. 4rch Phys Med Rehabil
1974:55:454-Y.
to the behavior of the prosthetic use by the LEA population
15. Hanf LC. O’Connor JB. A follow-up survey of predominantly older
in Quebec. The results of this survey will provide a charac- amputees. Geriatrics 19662 I : 166-73.
teristic profile the LEA population fitted with a prosthesis, 16. Mueller MJ. Delitto A. Selective criteria for successful long term pros-
which in turn will help to improve rehabilitation programs thetic use. Phys Ther 1985:65: 1037-40.
as well as follow-up health care services offered to this clien- 17. Reyes RL, Leahey EB Jr. Elderly patients with lower extremity ampu-
tations: three-year study in a rehabilitation setting. Arch Phys Med
tele. A second paper (Part II) will report the reliability (test- Rehabil 1977158: I 16-23.
retest) and construct validity of the PPA questionnaire. 18. Medhat A. Huber PM. Medhat MA. Factors that influence the level of
activities in persons with lower-extremit) amputation. Rehabil Nurs
CONCLUSION 1990: 15:13-8.
19. Beekman CE. Axtell LA. Prosthetic use in elderly patients with dysvas-
The questionnaire, PPA, was developed to assess the fac-
cular above-knee and through-knee amputations. Phys Thcr
tors potentially related to prosthetic use and the actual pros- 1987;67:1510-6.
thetic use of lower extremity amputee subjects discharged 30. Castronuovo JJ. Deane LM, Deterling R.4. O’Donnell TF. O’Toole
from rehabilitation facilities in QuCbec, Canada. The con- DM. Callow 4D. Below-knee amputation. Rrch Surg I Y8O:l 15: I I84-
ceptualization of the questionnaire incorporated experts’ 7.
21. Kegel B. Carpenter ML. Burgess EM. Functional capabilities oflower
considerations and its construction followed recognized de- extremity amputees. Arch Phys Med Rehabil lY78:59: 109-20.
sign techniques. The questionnaire, both mail telephone 22. Kihn RB. Warren R. Beebe GW. The “geriatric” amputee. Ann Surg
versions. are available in French and English. The content 197’:176:305-14.

Arch Phys Med Rehabil Vol74, August 1993

870 PROSTHETIC PROFILE OF THE AMPUTATION SUBJECT, Grist?

23. Moore TJ, Barron J, Hutchison III F, Golden C, Ellis C, Humphies D. 38. Day HJB. The assessment and description of amputee activity.
Prosthetic usage following major lower extremity amputation. Clin Prosthet Orthot Int I98 1;5:23-8.
Orthop Relat Res 1989;238:219-24. 39. Siemiatycki J, Campbell S. Nonresponse bias and early versus all re-
24. Dickstein R, Halevy E, Meir H, Bigon R. Use of the early walking aid sponders in mail and telephone surveys. Am J Epidemiol
as a “geriatric prosthesis” in the community. J Prothet Orthot 1984a;120:291-301.
1989;1:110-5. 40. Siemiatycki J, Campbell S, Richardson L, Aubert D. Quality of re-
25. Francis W, Renton CJC. Mobility after major limb amputation for sponse in different population groups in mail and telephone surveys.
arterial occlusive disease. Prosthet Orthot Int 1987;11:85-9. Am J Epidemiol 1984b;120:302-14.
26. Steinbert FU, Sunwoo I, Roettger RF. Prosthetic rehabilitation ofgeri- 41. Subbarao JV, McPhee MC. Prosthetic rehabilitation. comparison of
attic patients: a follow up study. Arch Phys Med Rehabil 1985;66:742- the outcome in patients with cancer and vascular amputation of ex-
5. tremities. Orthop Rev 1982;9:43-5 1.
27. Anderson JE. Use of lower limb prostheses: a questionnaire survey. 42. Houghton A, Allen A, Luff R, McCall I. Rehabilitation after lower
Med J Aust 1980;1:125-6. limb amputation: a comparative study of above-knee, through-knee
28. Murray DD, Shears AH, Schaffer JL. Nova Scotia Rehabilitation and Gritti-Stokes amputation. Br J Surg 1989;76:622-4.
Center Amputee Clinic: a review of lower limb amputees. Can J Phys- 43. Benson J. Clark F. A guide for instrument development and valida-
iother 1979;3 1: 16-20. tion. Am J Occ T 1982:36:789-800.
29. Alexander JL, Fuhrer MJ. Functional assessment of individuals with 44. Green LW, Kreuter MW, Deeds SG, Partridge KB. Health education
physical impairments. In: Halpem AS, Fuhrer MJ. Functional assess- planning. a diagnostic approach. Palo Alto, CA: Mayfield. 1980.
ment in rehabilitation. Baltimore: Brookes, 1984. 45. Bertera RL. Planning and implementing health promotion in the
30. Millstein S, Bain D, Hunter GA. A review of employment patterns of workplace: a case study of the DUPONT company experience. Health
industrial amputees: factors influencing rehabilitation. Prosthet Educ Q 1990; 17:307-27.
Orthot Int 1985;9:69-78. 46. Walter HJ. Primary prevention ofchronic disease among children: the
31. Durance JP, Warren WK, Kerbel DB, Stround TWF. Rehabilitation school based “Know Your body” intervention trials. Health Educ Q
of below-knee amputees: factors influencing outcome and costs in 1989;16:201-14.
three uroarammes. Int Disabil Stud 1989: 11:127-32. 47. Eriksen MP. Gielen AC. The application of health education princi-
32. O’Toole DM, Goldberg RT, Ryan B. Functional changes in vascular ples to automobile child restraint programs. Health Educ Q
amputee patients: evaluation by Barthel index, PULSES profile and 1983;10:30-55.
ESCROW scale. Arch Phys Med Rehabil 1985;66:508-11. 48. Downey AM, Frank GC, Webber LS, et al. The implementation of
33. Katz S, Downs TD, Cash HR, Grotz RC. Progress in development of “Heart Smart”: a cardiovascular school health promotion program. J
the Index of ADL. Gerontologist 1970; 1:20-30. Sch Health 1987;57:98- 104.
34. Iverson IA, Silbergerg NE, Stever RC, Schoening HA. The revised 49. Delbecq AL, Van de Ven AH, Gustafson DH. Group techniques for
Kenny self-care evaluation, a numerical measure of independence in program planning. Glenview, IL: Scott, Foresman, 1975.
activities ofdaily living. User guide. Minnesota: Sister Kenny Institute 50. Dillman DA. Mail and telephone surveys, the total design method.
Abbott-Northwestern Hospital, Research & Education Department New York: Wiley, 1978.
Publication-A/V Office, 1973. 51. Sudman S. Bradburn NM. Asking questions, a practical guide to ques-
35. Forer S, Granger CV, Gordon W, et al. The functional independence tionnaire design. San Francisco: Jossey-Bass, 1982.
measure (FIM), Guide to the implementation phase. Buffalo: School 52. Warwick DP, Lininger CA. The sample survey: theory and practice.
of Medecine, Department of Rehabilitation Medecine, State Univer- New York: McGraw-Hill, 1975.
sity of New York at Buffalo, 1986. 53. World Health Organization. International classification of impair-
36. Crewe NM, Athelstan GT. Functional assessment in vocational reha- ments, disabilities and handicaps. Geneva: World Health Organiza-
bilitation: a systematic approach to diagnosis and goal setting. Arch tion, 1980.
Phys Med Rehabil 1981;62:299-305. 54. Del Greco L, Walop W, Eastridge L. Questionnaire development: 3.
37. Carey RG, Posavac EJ. Manual for the level of rehabilitation scale translation. CMAJ 1987;136:8 17-8.
(LORS). Illinois: Office of evaluation and research services, Lutheran 55. Del Greco L, Walop W, McCarthy RH. Questionnaire development:
General Hospital, Park Ridge, 1977. 2. validity and reliability. CMAJ 1987, 136:699-700.

Arch Phys Med Rehabil Vol74, August 1993

Research article EMHJ – Vol. 27 No. 4 – 2021

Experiences and needs of patients with lower limb amputation in

Saudi Arabia: a qualitative study
Noura Abouammoh,1 Wassim Aldebeya2 and Rabab Abuzaid3

Department of Family and Community Medicine, College of Medicine, King Saud University, Riyadh, Saudi Arabia (Correspondence to: N. Abouammoh:
1

nabouammoh@[Link]). 2Department of Orthopaedic Surgery, College of Medicine, King Saud University, Riyadh, Saudi Arabia. 3Rehabilitation
Centre, Surgical and Amputee, Sultan Bin Abdulaziz Humanitarian City, Riyadh, Saudi Arabia.

Abstract
Background: The effect of amputation on the social and psychological well-being of patients has been established. How-
ever, the experiences and needs of amputees for the process of adjustment varies among individuals and cultures.
Aims: This study aimed to explore the experiences and needs of lower limb amputees for social and psychological adjust-
ment in Saudi Arabia, according to their own perspective.
Methods: Thirteen patients with lower limb amputation (mean age 47 years) were recruited from a large rehabilitation
centre in Saudi Arabia for participation in interviews. A focus group discussion with 6 amputees was followed by indi-
vidual, semistructured interviews with 8 amputees (which included 1 from the focus group) between November 2017 and
March 2018.
Results: Patients’ needs and reactions prior to and after amputation were controlled by the surrounding support system.
Hopelessness and depression, body image distress, religious attitude, and family and community support all contributed
to shape the overall patient experience, including psychological and physical adjustment.
Conclusions: Facilitating the re-integration of patients with lower limb amputation patients into their communities, as
well as providing the required support system, is crucial to ensure a healthy adjustment process for amputees.
Keywords: lower limb amputation, patient experiences, patient needs, Saudi Arabia
Citation: Abouammoh N; Aldebeyan W; Abuzaid R. Experiences and needs of patients with lower limb amputation in Saudi Arabia: a qualitative study.
East Mediterr Health J. 2021;27(4):407–413. [Link]
Received: 30/09/19; accepted: 29/04/20
Copyright © World Health Organization (WHO) 2021. Open Access. Some rights reserved. This work is available under the CC BY-NC-SA 3.0 IGO license
([Link]

Introduction in 2018 (10). More than half of these cases were due
Amputation is a treatment of choice in several health to diabetes, and other causes included road traffic
conditions. Many factors shape a patient’s psychological accidents, other accidents and malignancies. Two thirds
reaction to amputation, including age, type and level of of amputees do not receive rehabilitation services, for
amputation, time since amputation, social support, and unreported reasons (11). The economic impact of treating
active coping mechanisms (1–3). amputees is burdensome, and not meeting their needs
Most patients who experience limb amputation may add to its consequences (12).
encounter a series of multifaceted psychological The aim of the current study was to explore the
responses (1). Depression is a common psychological
adjustment experiences of amputees in Saudi Arabia and
reaction among amputees that may linger for 10–20 years
their needs before and after amputation.
after amputation and negatively affect an amputee’s
attempt to adjust socially and psychologically with
their physical situation. (3–5). Amputees experience Methods
devastation and distress, as well as occasional denial,
shortly after amputation (6–8). Poor social support and Study design
self-image are some of the main concerns that prevent a This was a phenomenological study of the lived experi-
healthy coping experience (6).
ences of lower limb amputees within their community.
Experiences and needs before and after amputation, The researchers had an etic perspective; looking at par-
among amputees from Arabic and Islamic cultures have
ticipants’ experiences without preconceptions. Two au-
not been previously studied. Culture plays an important
role in shaping an individual’s lifestyle, beliefs and thors (WA and RA) work in close contact with amputees,
attitudes, as well as their family and social network (9). which made them more familiar with the expressions,
In Saudi Arabia, 3745 cases of amputation were reported sentiments and ways to establish rapport with the par-
from medical rehabilitation centres and departments ticipants.

407
Research article EMHJ – Vol. 27 No. 4 – 2021

Study setting tured interviews were used to handle sensitive issues

(e.g., feelings and spiritual attitudes). Triangulation of
The study was conducted at Sultan Bin Abdulaziz Hu-
data collection techniques, probing questions, more than
manitarian City (SBAHC); a major nonprofit rehabilita-
one coder, and transparency in reporting the results were
tion hospital and medical centre in Riyadh. Patients at
used to ensure trustworthiness of the findings.
this hospital are required to pay for services, but many
receive government support to cover treatment and re- Data analysis
habilitation expenses.
All interviews were audio-recorded, transcribed verbatim,
Sampling and recruitment and analysed in their original Arabic language. Thematic
analysis was used to provide a sequential structure for
This study was approved by the Institutional Review
the data analysis. An iterative approach was adopted in
Board of SBAHC (013/2017/23, August 2017). A focus group
order to cover patient experiences. Familiarization with
discussion was followed by semistructured interviews
the range, depth and diversity of the data was performed
with lower limb amputees from SBAHC between Novem-
by NA and WA. They agreed on an initial coding frame
ber 2017 and March 2018. Participants for the focus group
that was then applied to other transcripts, with the flexi-
were conveniently chosen from the Inpatient Rehabilita-
bility to enable other codes to be added. Key themes were
tion Unit. From the 11 participants who were eligible, 6
developed and quotations were used to provide support-
attended at the time of the focus group, including 1 am-
ing evidence (Table 2). Themes were investigated until
putee who works at the centre.
saturation was achieved. NVivo software (version 11.4.2)
A convenience sample of 3 inpatients and a purposive was used to manage the data. 
sample of 4 more patients were invited to participate in the
semistructured interviews. Interviews were conducted
and analysed 4 weeks after the discussion. Patients were Results
included if they were Saudi nationals aged ≥ 18 years. Six and 8 patients participated in the focus group and
We included individuals of both sexes and with different semistructured interviews (3 men and 5 women, includ-
levels of education, as well as with various times since ing 1 man from the focus group). Participant age ranged
and levels of amputation. Individuals were not chosen if between 26 and 71 years. Time since amputation ranged
they lacked the capacity to provide informed consent or between 4 and 15 years, and levels of amputation var-
had severe mental illness. ied among the participants (9 above knee, 3 below knee
and 1 at ankle level). Six of the participants had bilater-
Data collection al above-knee amputation. Two main themes emerged
The focus group discussion took place in a quiet meeting from analysing the data: experiences and needs before
room at SBAHC and lasted 1 hour. Three interviews took amputation, and experiences in social and psychological
place in the same room and the others were conducted adjustment after amputation (Table 3).
in the patient’s room and lasted 30–45 minutes. Data
collection was conducted by NA in the presence of RA or Experiences and needs before amputation
WA. The topic guide was developed based on guidelines Patient attitude
previously used in similar studies (6,7,9,13) and amended
Differences in reactions between those who were about
according to the aim of this project (Table 1). Semistruc-
to lose their limb at a young age and those who under-
went amputation at an older age were found. A partic-
ipant peacefully noted: “I told the doctor if this is (foot
Table 1 Topic guide for focus group discussion
amputation) what you see as the best for me then…go for
– What does the word amputee mean to you? it! This is what God wants for me.” (male, 67 years, focus
– What was the story behind your limb amputation? group). Another participant added: “I might die at any
– What was your main concern before amputation? (probe: con- time…losing a limb is not a big deal” (male, 73 years). A
cerns about self-image, social functioning and physical func- 26-year-old woman who had her leg amputated at the age
tioning) of 12 years due to bone cancer explained: “.. when I heard
– How was your life before being an amputee? How did it change ‘cut her leg’ I cried hysterically, because that means I will
after? (psychological, social and religious)
never be able to walk again.”
– Can you identify some of the challenges faced by lower-limb
amputees? (challenges on accepting self-image, social challenges, Those who had their limbs amputated due to trauma
physical challenges and psychological challenges) and those who lost them due to vascular disease showed
– What are the strategies you took (are taking) to adapt to your life differences in their reaction to amputation. Participants
after amputation? (probe: rehabilitation support including pros- who underwent surgery to restore the limb before
thesis, family and friends, talking to other amputees, spiritual amputation, were less frustrated. One participant
support, and psychological counselling)
explained: “I was glad that all the suffering will end...”
– What do you think amputees want from the community? (male, 29 years, focus group).
– What do you think professionals and community could do to Consultation with religious men to lead prayers
facilitate integration of new amputees back into the community?
for amputees before their operations was common.
– Is there anything else you would like to add? Although they subsequently underwent amputation,

408
Research article EMHJ – Vol. 27 No. 4 – 2021

Table 2 Process of deriving themes from the focus group discussion and the interviews with amputation patients
Primary codes Categories Themes
– Patients’ attitude before amputation 1. Older patients tended to take a resilient attitude while Experiences and needs before
– Family support before amputation younger patients were more sensitive. amputation
2. Amputees who underwent restorative surgery were,
– Family attitude before amputation unlike those who lost their limb due to trauma, more
accepting of their situation.
3. Consulting religious men for emotional support
4. Need for support/empathy from the family
– Hopelessness and depression 1. Worrying of the unknown due to no knowledge about Experiences in social and
– Spiritually related attitude amputation and its consequences made amputees psychological adjustment after
hopeless. amputation
– Body image distress and self-esteem 2. Some younger participants experienced depression
3. Limb loss was associated with sins
4. Being more religious after amputation
5. Body image consciousness when around spouse,
relatives and strangers causes depression
6. Depending on others
– Family support 1. Parents’ and children’s support Physical and psychological support
– Community support 2. Parents’ close supervision
3. Spouse support (a priori theme)
4. Amputees did not want to be helped/ pitied in public.
5. No work problem (a priori theme)
6. Limitations to functional ability

participants reported that they felt emotional and ried about my future. Being able to walk again has never
physical improvement afterwards. A 51-year-old woman crossed my mind.” A 35-year-old man added: “...My friend
said: “My daughter called a religious man who came to told me that his cousin lost both her legs and she is able
the hospital to read Quran on me and my affected leg…I to walk now with prosthesis… I was so sad I couldn’t even
definitely felt better, my leg started to be less painful. I hold onto that hope.” A 51-year-old woman said: “(a phys-
was so comfortable…”. iotherapist) here from the centre had arranged a meeting
Family support and attitude for me to see one of the prosthesis users ... I could not
believe my eyes when I saw her (the prosthesis user)!” A
Participants’ experience with family support before am- 35-year-old man who had lost both limbs due to a car ac-
putation varied. For example, a 48-year-old woman not- cident noted: “I woke up to find that both my legs were
ed: “Everyone was crying... I was broken from the inside gone!... I was so angry I wished that I died before I got my
and seeing my kids crying was not helping at all. I had to legs amputated!”
find a way to calm them down, when I actually needed
someone to calm me down…I went to my room and cried Nervous breakdown and the development of clinical
for a whole day and at the same time managing to hide depression were mentioned by younger participants.
the distress I was feeling...” In contrast, a 51-yeard-old “I was sad, crying whenever I’m alone...” (female, 26
woman explained that the balance between those who years). “I was diagnosed with depression and I was using
cried and those who were strong for her made the news medications and consulting a psychiatrist for that.” (male,
of the amputation easier to accept: “I can’t forget my 35 years)
dad’s face… it was the first time I saw him crying… my Spiritually related attitudes
daughter was strong and she was focusing on making me
Psychological reaction to amputation was shaped by the
feel better...” Support and empathy were equally needed
patients’ religious beliefs and spirituality. Losing a limb
by participants before amputation. They felt important
was associated with being a sinful person. “I can tell from
to their families and loved ones when they shared their
their looks that they dislike me because they believe that
emotions, and at the same time, they needed people who
I am not a good person and that’s why I got punished by
were strong and able to hide their emotions. This same
Allah (God)…but this is not true” (female, 51). The same
patient added: “At that time, I have sensed how precious
participant added: “I used to constantly think…Why me?
I am to my father.”
What did I do wrong to deserve that?” Reliance on and
Experiences in social and psychological strengthening one’s faith in God had a positive role in
adjustment after amputation patients’ psychological ability to adjust to life after am-
putation. Most participants stated being more religious
Hopelessness and depression after amputation. “I cannot say no…I pray more, and I
Having limited knowledge about amputation, in addi- hope more from God.” (female, 36 years, focus group).
tion to not knowing someone who had been through the According to the participants, being more spiritual was
same situation, contributed to a state of hopelessness a source of hope and strength to be better able to cope
for the amputees. A 32-yeaar-old man noted: “I was wor- with their physical and psychological health. “Every time

409
Research article EMHJ – Vol. 27 No. 4 – 2021

Table 3 Themes and subthemes participant added: “My mum used to make sure that I
Themes Subthemes was being entertained the whole time.” (female, 36 years,
focus group). Two young female participants reported
Experiences and needs before – Patients’ attitude before
amputation amputation needing support; however, too much of that may back-
– Family support and attitude fire on their ability to adjust mentally and physically to
before amputation their new life. A participant who seemed to understand
Experiences in social and – Hopelessness and depression her mother’s concerns noted: “My mum never left me
psychological adjustment after – Spiritually related attitude alone! She used to come with me to school and then to
amputation – Body image distress and self-
esteem the university every day the whole day to make sure that
I don’t need anything. It was tiring for her… and a bit lim-
Physical and psychological – Family support
support – Community support iting for me.” (female, 26 years). Older patients, including
those with spouses, tended to get their physical, psycho-
logical and financial support from their children. Spousal
role in the physical support of amputation patients was
I pray, I feel that Allah is hearing me, and he is support-
also reported; however, this was not as significant as pa-
ing me. It makes me stronger… no, I still haven’t reached
rental support. When asked about his wife’s role as a sup-
the ultimate peacefulness though I did in others’ eyes.”
porter, a 35-year-old man reported: “My wife used to help
(male, 32 years).
my mother to take care of me when I most needed it.”
Body image distress and self-esteem
It was important for the participants to feel
Being conscious about the changes in their appearanc- comfortable in their environment. This included the
es made most of the participants worry about being workplace and other places that required regular visits,
accepted by their community and by their families. A such as supermarkets and government departments.
35-year-old man who had both legs amputated 4 months It was important for the participant to find wheelchair
after getting married noted: “I was worried that my wife ramps, elevators, and disabled bathrooms in the places
would leave me as I don’t look ‘good’ or… mutilated.” A they visited. One participant stated:
51-year-old woman noted: “I did not like to go to malls or “The community make us feel handicapped, but we
see strangers… people keep staring at me as if I were an are not!” (female, 36 years, focus group).
alien!” Looking different from others may restrict am-
Since these facilities are not available everywhere
putees from moving forward with their lives. Although
in Saudi Arabia, people must ask, or be approached
patients were able to assert their value, the struggle was
voluntarily by others for help, which may add to their
deeper. “I admit that I wanted to quit studying at the
psychological instability. A 26-year-old woman noted:
university many times due to that feeling I had. Even if
“Every time someone helps with carrying the wheelchair,
I tried to convince myself to live with my new different
they ask about my leg and how I lost it…other passers-
look peacefully and accept my new self…I am in a con-
by try to hear my story…it is annoying.” Another patient
stant battle from the inside.” (female, 26 years). A 51-year-
added: “I remember an old lady saying, “you are young,
old woman explained that it was the change in her body
you don’t deserve this”. I don’t like the look of pity I get.”
image that put her in a state of depression: “It was impor-
(female, 36 years, focus group).
tant for me to look normal. People talk and stare… it made
me sad and not wanting to see anyone.” Self-esteem was
affected by amputees’ body image and limited physical Discussion
function. Participants reported being hesitant to go back The current study was conducted to explore the needs
to work after amputation due to their looks. One teacher for better physical and psychological adjustment among
noted: “I did not know how I could go back to work and
lower limb amputees. Interviews indicated that patients
meet my young students…I did not want to hear negative
needed a balanced environment for healthy expression
comments…” (female, 51 years). Another noted: “Feeling
of their emotions and that their physical and emotional
dependent after all the years of giving and providing for
symptoms could be alleviated by cultural and spiritual
my family was difficult.” (male, 61 years, focus group).
traditions. Furthermore, this study showed that in con-
In contrast, older participants expressed their surrender
trast to community support, family support was effective
to their situation and tended to respond with words that
for adaptation and improvement of quality of life.
reflected their gratification with what was destined to
happen. Having the opportunity to go through the decision-
making process may lead to higher perceived control
Physical and psychological support and improved adjustment (14). This study showed
Parents and children were mentioned by the participants that previous distress of participants who underwent
as being their main source of support, even among mar- amputation after a series of reconstructive operations
ried patients. One married participant stated: “My father appeared to have more control over their emotions and
was ready to do anything to help me through this…when needs. In contrast, those who underwent amputation
I got home, I found my house to be fully equipped with due to trauma had greater difficulty in accepting their
handicapped facilities.” (male, 35 years). Another single condition shortly after surgery.

410
Research article EMHJ – Vol. 27 No. 4 – 2021

It could be argued that depressive reactions are a In contrast with other research conducted in similar
normal reaction to amputation rather than depression per cultural settings, this study found that parents and/or
se (1–5). Nevertheless, this study showed that depressive children, were the main source of support for lower-limb
reactions could be minimized with patient education. amputees (25–27). Participants in our study reinforced the
Similar to previous studies (15), our results suggested that role of their parents, or their children for older patients, in
meeting with other amputees was the most useful source improving physical and emotional adjustment. Patients
of information and support. This was possibly because did acknowledge their spouses’ role, but seemed to take
emic sources of support and information were more this for granted.
trustworthy for patients.
Our study explained that local cultural and social
Adaptation to body image could be used as a potential context factors may make amputees feel handicapped. It
measure of psychological adjustment to amputation has been noted in a qualitative study that people develop
(8,16). Participants in the present study showed self-
their identity and interpretation of their look from other
consciousness of their postamputation body image.
peoples’ perspective (28). In concordance, participants in
In contrast to participants of previous studies (17,18),
the current study did not like to be pitied or helped by
younger participants showed limited acceptance to their
others when logistics were lacking.
new identity as amputees, and this affected their self-
esteem around their families and loved ones. The experiences of disadvantaged patients who did
Participants considered spirituality as a key element not utilize rehabilitation services could be different from
of the support system utilized in order to adjust to their those in the present study. Future research should look at
new situation. In line with findings of other studies, the experiences of less-advantaged lower limb amputees
we found that participants’ spiritual beliefs stabilized who have poor access to rehabilitation services. The
their lives, provided meaning for the experience of role of spouses and religious/traditional healers in the
disability, assisted them with coping, and enhanced physical and psychological adjustment could be further
their spiritual connection (19,20). Simultaneously, some explored.
amputees believed that going through amputation was
a punishment from God for their sins; the latter belief Conclusions
made them isolate themselves from the community.
Another study on families of intensive care patients in A range of unmet needs, which involved family, commu-
Saudi Arabia found that families believed that illness was nity and religious support, was identified among low-
a test from God, rather than a punishment (19). er-limb amputees. This study calls upon policy-makers to
Family support is the main source of psychological participate in reducing social and psychological implica-
strength among amputees and patients with chronic tions of amputation as a disability, by empowering am-
disease in general (21–24). This is the first study to explore putees to acquire their right of independence and social
patient needs from their family before amputation. inclusion by providing all the necessary logistics. Health-
Participants reported needing a balanced reaction to care professionals should pay more attention to patient
enable them to feel strong and express their emotions needs and support systems by involving families and re-
simultaneously. Potential amputees may suppress their ligious healers in amputees’ treatment plans, in addition
true feelings if their families overlook the support they to introducing success stories and life examples to allevi-
could provide. ate the emotionally destructive news of amputation.

Acknowledgement
The authors are grateful to the Deanship of Scientific Research, King Saud University, for funding through the Vice Dean-
ship of Scientific Research Chairs. The authors are thankful to Professor Abdulrahman Abouammoh for his support and
gratefully acknowledge the patients who participated in this study.
Funding: None.
Competing interests: None declared.

Expériences et besoins des patients amputés d’un membre inférieur en Arabie

saoudite : étude qualitative
Résumé
Contexte : L'effet de l'amputation sur le bien-être social et psychologique des patients a été établi. Cependant, les
expériences et les besoins des personnes amputées pour le processus d'adaptation varient selon les individus et les
cultures.
Objectifs : La présente étude visait à examiner les expériences et les besoins des sujets amputés des membres
inférieurs en matière d'adaptation sociale et psychologique en Arabie saoudite, selon leur propre perspective.

411
Research article EMHJ – Vol. 27 No. 4 – 2021

Méthodes : Treize patients amputés d'un membre inférieur (âge moyen de 47 ans) ont été recrutés dans un grand
centre de réadaptation en Arabie saoudite pour participer à des entretiens. Un groupe de discussion thématique avec
six patients a été suivi d’entretiens individuels et semi-structurés avec huit patients (dont un ayant participé au
groupe de discussion thématique) entre novembre 2017 et mars 2018.
Résultats : Les besoins et les réactions des patients avant et après l’amputation étaient déterminés par le système de
soutien environnant. Le désespoir et la dépression, la détresse liée à l'image corporelle, l'attitude religieuse et le
soutien familial et communautaire ont tous contribué à façonner l'expérience globale du patient, y compris
l'adaptation psychologique et physique.
Conclusions : Faciliter la réintégration des patients amputés des membres inférieurs dans leur communauté, ainsi
que fournir le système de soutien nécessaire, est crucial pour assurer un processus d'adaptation favorisant la santé à
ces derniers.

‫ دراسة نوعية‬:‫جتارب واحتياجات مرىض برت األطراف السفلية يف اململكة العربية السعودية‬
‫ رباب أبو زيد‬،‫ وسيم الدبيان‬،‫نورا أبوعمه‬
:‫اخلالصة‬
‫ غري أن جتارب مرىض البرت واحتياجاهتم من أجل عملية التكيف ختتلف من‬،‫ ث ُبت أن البرت يؤثر عىل السالمة االجتامعية والنفسية للمرىض‬:‫اخللفية‬
.‫ثقافة إىل أخرى ومن فرد إىل آخر‬
‫ هدفت هذه الدراسة إىل التعرف عىل جتارب واحتياجات مرىض برت األطراف السفلية من أجل التكيف اجتامعي ًا ونفسي ًا يف اململكة العربية‬:‫األهداف‬
.‫ وفق ًا لوجهة نظرهم اخلاصة‬،‫السعودية‬
‫ عام ًا) من أحد املراكز الكربى إلعادة التأهيل يف اململكة‬47 ‫ استُعني بثالثة عرش مريض ًا من مرىض برت األطراف السفلية (بمتوسط عمر‬:‫طرق البحث‬
‫ من مرىض البرت (أحدهم من جمموعة الرتكيز) عقب‬8 ‫ وعقدت مقابالت شبه منظمة مع‬.‫ وذلك للمشاركة يف مقابالت معهم‬،‫العربية السعودية‬
.2018 ‫آذار‬/‫ ومارس‬2017 ‫ترشين الثاين‬/‫ وذلك يف الفرتة بني نوفمرب‬،‫ من مرىض البرت يف جمموعة تركيز‬6 ‫مناقشة مع‬
‫ وقد ساهم كل من اليأس واالكتئاب والشعور بالضيق من‬.‫ حتكم نظام الدعم املحيط يف احتياجات املرىض وردود أفعاهلم قبل البرت وبعده‬:‫النتائج‬
.‫ ويشمل ذلك عملية التكيف النفيس والبدين‬،‫صورة اجلسم واملوقف الديني والدعم األرسي واملجتمعي يف تشكيل جتربة املرىض بشكل عام‬
‫ من الرضوري تيسري إعادة دمج مرىض برت األطراف السفلية يف جمتمعاهتم وتوفري نظام الدعم املطلوب هلم لضامن حتقق عملية تكيف‬:‫االستنتاجات‬
.‫مرىض البرت بصورة صحية‬

References
1. Mckechnie PS and John A. Anxiety and depression following traumatic limb amputation: a systematic review. Injury. 2014 Dec;
45(12):1859–66. [Link] PMID:25294119
2. Belon HP and Vigod DF. Emotional adaptation to limb loss. Phys Med Rehabil Clin N Am. 2014 Feb; 25(1):53–74. [Link]
org/10.1016/[Link].2013.09.010 PMID:24287240
3. Bragaru M. Van Wilgen CP, Geertze J.H, Ruij SG, Dijkstra PU, Dekker R. Barriers and facilitators of participation in sports: a
qualitative study on Dutch individuals with lower limb amputation. PloS One. 2013; 8(3):e59881. [Link]
pone.0059881 PMID:23533655
4. Horgan O, MacLachlan M. Psychosocial adjustment to lower-limb amputation: a review. Disabil Rehabil. 2004 Jul 22–Aug 5;
26(14–15):837–50. [Link] PMID:15497913
5. Durmus D, Safaz I, Adıgüzel E, Uran A, Sarısoy G, Goktepe AS, et al. The relationship between prosthesis use, phantom pain
and psychiatric symptoms in male traumatic limb amputees. Compr Psychiatry.2015 May;59:45–53. [Link]
comppsych.2014.10.018 PMID:25764906
6. Furst L, Humphrey M. Coping with the loss of a leg. Prosthet Orthot Int. 1983;7(1):152–6.
7. Murray CD, Forshaw MJ. The experience of amputation and prosthesis use for adults: a metasynthesis. Disabil Rehabil.2013
Jul;35(14):1133–42. [Link] PMID:23033871
8. Holzer LA, Sevelda F, Fraberger G, Bluder O, Kickinger W, Holzer G. Body image and self-esteem in lower-limb amputees. PLoS
One. 2014 Mar 24;9(3):e92943. [Link] PMID:24663958
9. Lupton, D. Medicine as culture: illness, disease and the body. Sage; 2012
10. Statistical year book. Ministry of Health, Saudi Arabia; 2018 ([Link]
[Link], accessed 4 September 2020)

412
Research article EMHJ – Vol. 27 No. 4 – 2021

11. AlSofyani MA, AlHarthi AS, Farahat FM, Abuznadah WT. Impact of rehabilitation programs on dependency and functional
performance of patients with major lower limb amputations: a retrospective chart review in western Saudi Arabia. Saudi Med J.
2016 Oct;37(10):1109–13. [Link] PMID:27652362
12. Al-Thani H, Sathian B, El-Menyar A. Assessment of healthcare costs of amputation and prosthesis for upper and lower extrem-
ities in a Qatari healthcare institution: a retrospective cohort study. BMJ Open. 2019;9:e024963. [Link]
en-2018-024963
13. Senra H, Oliveira RA, Leal I, Vieira C. Beyond the body image: a qualitative study on how adults experience lower limb amputa-
tion. Clin Rehabil 2012 Feb;26(2):180–91. [Link] PMID:21908476
14. Ogden, J. Health psychology a textbook. Open University Press; 2012.
15. Hamill R, Carson S, Dorahy M. Experiences of psychosocial adjustment within 18 months of amputation: an interpretative phe-
nomenological analysis. Disabil Rehabil. 2010;32(9):729–40. [Link] PMID:20302441
16. Cho H, MacLachlan M, Clarke M, Mannan H. Accessible home environments for people with functional limitations: a systemat-
ic review. Int J Environ Res Public Health. 2016 Aug 17;13(8):826. [Link] PMID:27548194
17. Akarsu S, Tekin L, Safaz I, Göktepe AS, Yazıcıoğlu K. Quality of life and functionality after lower limb amputations: comparison
between uni- vs. bilateral amputee patients. Prosthet Orthot Int. 2013 Feb;37(1):9–13. [Link]
PMID:22833517
18. Weiss G. Body images: embodiment as intercorporeality. London: Routledge; 2013
19. Al‐Mutair AS, Plummer V, Clerehan R, O’brien A. Needs and experiences of intensive care patients’ families: a Saudi qualitative
study. Nurs Crit Care. 2014 May;19(3):135–44. [Link] PMID:24118629
20. Treloar LL. Disability, spiritual beliefs and the church: the experiences of adults with disabilities and family members. J Adv
Nurs. 2002 Dec 40(5):594–603. [Link] PMID:12437609
21. Valizadeh S, Dadkhah B, Mohammadi E, Hassankhani,H. The perception of trauma patients from social support in adjust-
ment to lower-limb amputation: A qualitative study. Indian journal of palliative care. 2014 Sep;20(3):229–38. [Link]
org/10.4103/0973-1075.138401 PMID:25191013
22. Oztunc G, Yesil P, Paydas S, Erdogan S. Social support and hopelessness in patients with breast cancer. Asian Pac J Cancer Prev.
2013;14(1):571–8. PMID:23534797
23. Abdullah-zadeh F, Agahosseini S, Asvadi-Kermani I, Rahmani A. Hope in Iranian cancer patients. Iranian J Nurs Midwifery Res.
2011 Fall;16(4):288–91. PMID:23450037
24. Li QP. Mak YW, Loke AY. Spouses’ experience of caregiving for cancer patients: a literature review. Int Nurs Rev. 2013
Jun;60(2):178–87. [Link] PMID:23692000
25. Delgado-Guay MO, Parsons HA, Hui D, Cruz MGDL, Thorney S, Bruera E. Spirituality, religiosity, and spiritual pain
among caregivers of patients with advanced cancer. Am J Hosp Palliat Care. 2013 Aug;30(5):455–61. [Link]
org/10.1177/1049909112458030 PMID:22952129
26. Wen KY, Fang CY, Ma GX. Breast cancer experience and survivorship among Asian Americans: a systematic review. J Cancer
Surviv. 2014 Mar;8(1):94–107. [Link] PMID:24214498
27. Jassim GA, Whitford DL. Understanding the experiences and quality of life issues of Bahraini women with breast cancer. Soc Sci
Med. 2014 Apr;107:189–95. [Link] PMID:24631996
28. Monteleone R, Forrester-Jones R. ‘Disability means, um, dysfunctioning people’: A qualitative analysis of the meaning and ex-
perience of disability among adults with intellectual disabilities. J Appl Res Intellect Disabil. 2017 Mar;30(2):301–15. [Link]
org/10.1111/jar.12240 PMID:26834036

413
 
Cochrane
Library
Cochrane Database of Systematic Reviews

   
Type of incision for below knee amputation (Review)

  Tisi PV, Than MM  

  Tisi PV, Than MM.  
Type of incision for below knee amputation.
Cochrane Database of Systematic Reviews 2014, Issue 4. Art. No.: CD003749.
DOI: 10.1002/14651858.CD003749.pub3.

  [Link]  

 
Type of incision for below knee amputation (Review)
Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

TABLE OF CONTENTS
HEADER......................................................................................................................................................................................................... 1
ABSTRACT..................................................................................................................................................................................................... 1
PLAIN LANGUAGE SUMMARY....................................................................................................................................................................... 2
BACKGROUND.............................................................................................................................................................................................. 3
Figure 1.................................................................................................................................................................................................. 4
Figure 2.................................................................................................................................................................................................. 5
OBJECTIVES.................................................................................................................................................................................................. 6
METHODS..................................................................................................................................................................................................... 6
RESULTS........................................................................................................................................................................................................ 7
Figure 3.................................................................................................................................................................................................. 8
Figure 4.................................................................................................................................................................................................. 9
Figure 5.................................................................................................................................................................................................. 10
DISCUSSION.................................................................................................................................................................................................. 11
AUTHORS' CONCLUSIONS........................................................................................................................................................................... 12
ACKNOWLEDGEMENTS................................................................................................................................................................................ 12
REFERENCES................................................................................................................................................................................................ 13
CHARACTERISTICS OF STUDIES.................................................................................................................................................................. 14
DATA AND ANALYSES.................................................................................................................................................................................... 16
Analysis 1.1. Comparison 1 Two-stage amputation versus one-stage amputation, Outcome 1 Failed primary stump healing....... 17
Analysis 1.2. Comparison 1 Two-stage amputation versus one-stage amputation, Outcome 2 Post-operative infection rate........ 17
Analysis 1.3. Comparison 1 Two-stage amputation versus one-stage amputation, Outcome 3 Reamputation at same level........ 17
Analysis 1.4. Comparison 1 Two-stage amputation versus one-stage amputation, Outcome 4 Reamputation at higher level....... 18
Analysis 1.5. Comparison 1 Two-stage amputation versus one-stage amputation, Outcome 5 Mobility with prosthetic limb........ 18
Analysis 2.1. Comparison 2 Skew flaps amputation versus long posterior flap amputation, Outcome 1 Failed primary stump 18
healing...................................................................................................................................................................................................
Analysis 2.2. Comparison 2 Skew flaps amputation versus long posterior flap amputation, Outcome 2 Post-operative wound 19
necrosis..................................................................................................................................................................................................
Analysis 2.3. Comparison 2 Skew flaps amputation versus long posterior flap amputation, Outcome 3 Reamputation at same 19
level........................................................................................................................................................................................................
Analysis 2.4. Comparison 2 Skew flaps amputation versus long posterior flap amputation, Outcome 4 Reamputation at higher 19
level........................................................................................................................................................................................................
Analysis 2.5. Comparison 2 Skew flaps amputation versus long posterior flap amputation, Outcome 5 Fitted with prosthetic 19
limb........................................................................................................................................................................................................
Analysis 2.6. Comparison 2 Skew flaps amputation versus long posterior flap amputation, Outcome 6 Mobility with prosthetic 19
limb........................................................................................................................................................................................................
Analysis 2.7. Comparison 2 Skew flaps amputation versus long posterior flap amputation, Outcome 7 30-day mortality............ 20
Analysis 3.1. Comparison 3 Sagittal flaps amputation versus long posterior flap amputation, Outcome 1 Failed primary stump 20
healing...................................................................................................................................................................................................
Analysis 3.2. Comparison 3 Sagittal flaps amputation versus long posterior flap amputation, Outcome 2 Reamputation at same 20
level........................................................................................................................................................................................................
Analysis 3.3. Comparison 3 Sagittal flaps amputation versus long posterior flap amputation, Outcome 3 Reamputation at higher 21
level........................................................................................................................................................................................................
Analysis 3.4. Comparison 3 Sagittal flaps amputation versus long posterior flap amputation, Outcome 4 Fitted with prosthetic 21
limb - overall.........................................................................................................................................................................................
Analysis 3.5. Comparison 3 Sagittal flaps amputation versus long posterior flap amputation, Outcome 5 Fitted with prosthetic 21
limb - below knee prosthesis...............................................................................................................................................................
Analysis 3.6. Comparison 3 Sagittal flaps amputation versus long posterior flap amputation, Outcome 6 Fitted with prosthetic 21
limb - PTB prosthesis............................................................................................................................................................................
Analysis 3.7. Comparison 3 Sagittal flaps amputation versus long posterior flap amputation, Outcome 7 Mortality at 3 months.... 21
APPENDICES................................................................................................................................................................................................. 22
WHAT'S NEW................................................................................................................................................................................................. 22
HISTORY........................................................................................................................................................................................................ 22
CONTRIBUTIONS OF AUTHORS................................................................................................................................................................... 23

Type of incision for below knee amputation (Review) i

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

DECLARATIONS OF INTEREST..................................................................................................................................................................... 23
SOURCES OF SUPPORT............................................................................................................................................................................... 23
DIFFERENCES BETWEEN PROTOCOL AND REVIEW.................................................................................................................................... 23
NOTES........................................................................................................................................................................................................... 23
INDEX TERMS............................................................................................................................................................................................... 23

Type of incision for below knee amputation (Review) ii

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

[Intervention Review]

Type of incision for below knee amputation

Paul V Tisi1, Mary M Than2

1Department of Vascular Surgery, Bedford Hospital, Bedford, UK. 2General Surgery, Bedford Hospital NHS Trust, Bedford, UK

Contact address: Paul V Tisi, Department of Vascular Surgery, Bedford Hospital, Kempston Road, Bedford, Bedfordshire, MK42 9DJ, UK.
pvtisi@[Link].

Editorial group: Cochrane Vascular Group.

Publication status and date: New search for studies and content updated (no change to conclusions), published in Issue 4, 2014.

Citation: Tisi PV, Than MM. Type of incision for below knee amputation. Cochrane Database of Systematic Reviews 2014, Issue 4. Art. No.:
CD003749. DOI: 10.1002/14651858.CD003749.pub3.

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

ABSTRACT

Background
Below knee amputation (BKA) may be necessary in patients with advanced critical limb ischaemia or diabetic foot sepsis in whom no other
treatment option is available. There is no consensus as to which surgical technique achieves the maximum rehabilitation potential. This
is the third update of the review first published in 2004.

Objectives
To assess the effects of different types of incision on the outcome of BKA in people with lower limb ischaemia or diabetic foot sepsis, or
both. The main focus of the review was to assess the relative merits of skew flap amputation versus the long posterior flap technique.

Search methods
For this update the Cochrane Peripheral Vascular Diseases Group Trials Search Co-ordinator (TSC) searched the Specialised Register (last
searched 28 March 2013) and CENTRAL (2013, Issue 2).

Selection criteria
Randomised controlled trials comparing two or more types of skin incision for BKA were identified. People with lower limb ischaemia
(acute or chronic) or diabetic foot sepsis, or both, were considered for inclusion. People undergoing below knee amputation for other
conditions were excluded.

Data collection and analysis

One review author identified potential trials. Two review authors independently assessed trial quality and extracted the data. Additional
information, if required, was sought from study authors.

Main results
Three studies with a combined total of 309 participants were included in the review. One study compared two-stage versus one-stage BKA;
one study compared skew flaps BKA versus long posterior flap BKA; and one study compared sagittal flaps BKA versus long posterior flap
BKA. Overall the quality of the evidence from these studies was moderate. BKA using skew flaps or sagittal flaps conferred no advantage
over the well established long posterior flap technique (primary stump healing was 60% for both skew flaps and long posterior flap (risk
ratio (RR) 1.00, 95% confidence interval (CI) 0.71 to 1.42) and primary stump healing was 58% for sagittal flaps and 55% for long posterior
flap (Peto odds ratio (OR) 1.04, 95% CI 0.45 to 2.43). For participants with wet gangrene, a two-stage procedure with a guillotine amputation
at the ankle followed by a definitive long posterior flap amputation led to better primary stump healing than a one-stage procedure (Peto
OR 0.08, 95% CI 0.01 to 0.89). Post-operative infection rate or wound necrosis, reamputation, and mobility with a prosthetic limb were
similar in the different comparisons.

Type of incision for below knee amputation (Review) 1

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Authors' conclusions
There is no evidence to show a benefit of one type of incision over another. However, in the presence of wet gangrene a two-stage procedure
leads to better primary stump healing compared to a one-stage procedure. The choice of amputation technique can, therefore, be a matter
of surgeon preference taking into account factors such as previous experience of a particular technique, the extent of non-viable tissue,
and the location of pre-existing surgical scars.

PLAIN LANGUAGE SUMMARY

Type of incision used for below knee amputation to create a skin flap that maximises healing

Below knee amputation may be necessary for people with critical limb ischaemia caused by advanced vascular disease or diabetic foot
infection (sepsis) where no other treatment option is possible. Keeping the knee joint gives a better chance of walking using an artificial
leg or prosthesis and social independence after the amputation. The surgical technique is important. Bone and deep tissues are generally
treated in a similar way but the type of skin incision varies between techniques. A skin flap is designed to go over the stump, where the main
consideration is to maximise blood supply and healing. A long posterior skin flap and unequal (skewed) anterior and posterior muscle and
skin (myocutaneous) flaps are most often used, although other techniques have been described.

Three randomised controlled studies were identified. Overall the quality of the evidence from these studies was moderate. They were
reported on between 1977 and 1991 and involved a total of 309 participants. Each reported on different comparisons. Below knee
amputation using skew flaps or sagittal flaps provided no advantage over the long posterior flap technique on primary stump healing,
which approached 60% for all groups. In the third study, involving 30 participants with wet gangrene, a two-stage procedure with a
guillotine amputation at the ankle followed by long posterior flap amputation led to better primary stump healing than a one-stage
procedure with delayed skin closure. Post-operative infection rate or wound necrosis, reamputation and mobility with a prosthetic limb
were similar in the different comparisons.

Nearly all the surgeons in the study that looked at skew flap amputation versus the long posterior flap technique were new to the skew flap
operation and so were on a learning curve. Factors which might have influenced the findings include previous experience of a technique,
the extent of non-viable tissue, and location of pre-existing surgical scars.

Type of incision for below knee amputation (Review) 2

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

BACKGROUND tissues in a similar way, although there may be some variation in

the level of bone section and formation of the muscle flaps used
Description of the condition to cover the bone ends. However, the type of skin incision varies
between techniques. The main consideration in designing a skin
Below knee amputation (BKA) has been increasingly used as
flap is to maximise its blood supply in order that healing may occur.
a therapeutic option since the 1950s. This is in preference to
The 'gold standard' BKA is the long posterior flap as popularised by
above knee amputation, which has a more limited chance of
Burgess (Burgess 1968a; Burgess 1968b), see Figure 1. The rationale
achieving successful rehabilitation (Silbert 1950). BKA is necessary
behind this technique is that the poorly vascularised anterior skin
for patients with advanced critical limb ischaemia who cannot
flap is compensated by the relatively well vascularised posterior
be treated with reconstructive vascular surgery (to restore blood
skin flap (Chavatzas 1975). Robinson showed in an uncontrolled
flow to the leg) or in whom vascular surgery has failed. It
study that 69% of patients treated by a long posterior flap walked
may also be necessary for patients with aggressive diabetic
with a prosthetic limb (Robinson 1976). Numerous alternative
foot infections or gangrene, or both; for those with extensive
techniques have been described. These include unequal anterior
venous ulceration; or following major trauma. Guidelines from the
and posterior myocutaneous (muscle and skin) flaps (Haimovici
Vascular Surgical Society of Great Britain and Ireland recommend
1996; Kaufman 1995); equal anterior and posterior myocutaneous
that all patients undergo a vascular surgical assessment prior
flaps (McCullough 1981); equal medial (inner side of leg) and lateral
to amputation (VSSGBI 1996). The current quality improvement
(outer side of leg) flaps (that is sagittal) (Alter 1978; Persson 1974);
framework for major amputation surgery (VSGBI 2010) aims to
a 'laterally based' skin flap (Catre 1997); and creation of a broad
reduce the mortality rate post major amputation surgery to less
posterior flap with burying of the ends of the flap (after removal
than 5% by 2015 and states that "amputation for vascular disease
of the outer skin layer) to provide extra padding for the stump
and diabetes should only be undertaken after formal investigation
(Galvao 1975). The 'skew' flap technique reported by Robinson (see
to the arterial system by angiography (diagnostic conventional
Figure 2) is widely considered to be superior to the long posterior
angiography (DSA), computed tomographic angiography (CTA) or
flap in terms of wound healing and time to full mobility (Harrison
magnetic resonance angiography (MRA)) or specialist ultrasound
1987; Robinson 1982; Robinson 1991). The aim of this review was to
imaging, except when the leg is clearly beyond salvage".
assess the evidence supporting the use of these techniques.
Description of the intervention
The results of BKA are influenced by the surgical technique used.
Most of the described operative techniques treat the bone and deep
 

Type of incision for below knee amputation (Review) 3

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Figure 1.   Long posterior (Burgess) flap.

 
 

Type of incision for below knee amputation (Review) 4

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Figure 2.   Skew flap.

 
How the intervention might work with peripheral arterial disease could maintain independent
ambulation after BKA (Pinzur 1993).
Careful judgement is required in patient selection and decision
making about the amputation level in order to maximise the Why it is important to do this review
rehabilitation potential. There is no doubt that preservation of the
knee joint leads to improved function and social independence, This is the third update of a Cochrane review first published in 2004.
with one report suggesting that 87% of previously mobile patients The main findings from previous versions of the review were that
the choice of amputation technique has no effect on outcome and
Type of incision for below knee amputation (Review) 5
Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

can, therefore, be a simple matter of surgeon preference. Factors (6) Quality of life measures, using formal quality of life
which might influence this finding include previous experience of questionnaires administered either in person or by post
a particular technique, the extent of non-viable tissue, and the
location of pre-existing surgical scars. An update of the review was Search methods for identification of studies
conducted to identify any new evidence since the publication of the
Electronic searches
last version of this review.
For this update, the Cochrane Peripheral Vascular Diseases (PVD)
OBJECTIVES Group Trials Search Co-ordinator (TSC) searched the Specialised
Register (last searched 28 March 2013) and the Cochrane Central
To assess the effects of different types of incision on the outcome of Register of Controlled Trials (CENTRAL) (2013, Issue 2) in The
below knee amputation (BKA) in people with lower limb ischaemia Cochrane Library ([Link]). See Appendix 1
or diabetic foot sepsis, or both. for details of the search strategy used to search CENTRAL. The PVD
Specialised Register is maintained by the TSC and is constructed
The main focus of the review was to assess the relative merits of
from weekly electronic searches of MEDLINE, EMBASE, CINAHL,
skew flap amputation versus the long posterior flap technique.
AMED, and through handsearching relevant journals. The full list
METHODS of the databases, journals and conference proceedings which have
been searched, as well as the search strategies used, are described
Criteria for considering studies for this review in the Specialised Register section of the Cochrane PVD Group
module in The Cochrane Library ([Link]).
Types of studies
Data collection and analysis
Randomised trials comparing two or more different skin incisions
or surgical techniques for BKA were considered for inclusion in this Selection of studies
review.
One review author collated all randomised trials identified from
Types of participants the search strategy for potential inclusion in the review. Additional
information, if required, was sought from the relevant authors to
People with lower limb ischaemia (acute or chronic) or diabetic enable the quality of the trials to be assessed.
foot sepsis, or both, were considered for inclusion without age
restriction. This included patients who had venous disease as Potentially eligible trials were assessed independently by two
well as peripheral arterial disease (PAD). People undergoing review authors to determine the relevance of each study. Ideally,
BKA for lower limb trauma and those requiring amputation for studies should have had sufficient statistical power to detect a
miscellaneous conditions such as bone or soft tissue tumours were difference between treatment groups. Trials were only accepted
excluded. if both review authors agreed on the inclusion criteria being met.
Disagreements were resolved through discussion.
Types of interventions
Data extraction and management
Long posterior flap (Burgess) BKA versus skew flap amputations, or
versus other less commonly used surgical techniques. Data from the trials were extracted independently by two review
authors. The figures were then cross-checked for agreement.
Types of outcome measures Disagreements were resolved through discussion.
Primary outcomes
Assessment of risk of bias in included studies
(1) Primary stump healing, defined as a painless, healed suture line
Two review authors independently assessed the risk of bias in
enabling fitting of a prosthetic limb (if appropriate) and regaining
each included study according to the Cochrane Handbook for
of mobility
Systematic Reviews of Interventions (Higgins 2011). We assessed the
(2) Post-operative infection rate, including the specific incidence of following domains of trial quality: random sequence generation,
methicillin-resistant Staphylococcus aureus (MRSA) infection allocation concealment, blinding, incomplete outcome data,
selective reporting and other bias. We gave trials a quality rating
(3) Rate of reamputation at (a) same level; (b) higher level of 'low risk', 'unclear risk' or 'high risk' of bias for each of these
domains according to the Cochrane Handbook for Systematic
(4) Number of participants mobilising with a prosthetic limb Reviews of Interventions (Higgins 2011). Disagreements were
resolved through discussion.
Secondary outcomes
(1) Number of participants fitted with a prosthetic limb Measures of treatment effect
Results were expressed as Peto odds ratios (OR) with 95%
(2) Thirty-day mortality rate confidence intervals (CI) for dichotomous variables, although for
(3) Length of hospital stay comparisons with a high frequency of events the outcomes were
given as risk ratio (RR) with 95% CI. Results for continuous variables
(4) Symptoms relating to the stump, such as pain and swelling were expressed as standardised mean differences (SMD) with 95%
CIs. For studies where the standard deviation was not given, further
(5) Phantom limb pain analysis was impossible and the studies were therefore not pooled.

Type of incision for below knee amputation (Review) 6

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Unit of analysis issues publication bias, although it was not appropriate to conduct such
an analysis due to the limited number of studies included.
The unit of analysis was the individual patient.
Data synthesis
Dealing with missing data
We used a fixed-effect model meta-analysis for the data analyses of
Missing follow-up data were sought from the original investigators,
the treatment effect.
where possible. If this information was unavailable, the data were
re-analysed using a reasonable range of values for the missing data Subgroup analysis and investigation of heterogeneity
to determine if this affected the overall results.
Subgroup analysis (for example diabetic participants who could
Assessment of heterogeneity be stratified into the presence or absence of peripheral arterial
disease) proved impossible in the trials identified.
Finer points of the analysis were determined by the type and quality
of the data extracted. Heterogeneity of the combined results from Sensitivity analysis
the different studies for each comparison was to be assessed using
a Chi2 test as well as by clinical judgement. However, this did not This was not applicable to this review due to the limited number of
apply in this review as each comparison related to a single study studies included.
only.
RESULTS
Assessment of reporting biases
Description of studies
Both the original and re-analysed results were reported, if
appropriate. A funnel plot was considered to identify any Results of the search
See Figure 3.
 

Type of incision for below knee amputation (Review) 7

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Figure 3.   Study flow diagram.

 
No new relevant studies were identified for this update.

Type of incision for below knee amputation (Review) 8

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Included studies BKA versus long posterior flap BKA (Ruckley 1991); and sagittal flaps
BKA versus long posterior flaps BKA (Termansen 1977). Fisher 1988
Three studies were included in the review. Duration of recruitment
also included participants undergoing above knee amputation
ranged from 28 to 30 months and the study follow-up was two
(AKA) but these data were excluded from the analyses of this review.
months in one study (Fisher 1988), six months in a second study
(Ruckley 1991) and a mean (range) of 11.7 months (3.5 to 22 Excluded studies
months) in the third study (Termansen 1977). All were parallel trials.
Participants, inclusion and exclusion criteria, interventions, and No studies were excluded.
outcomes are described in the table 'Characteristics of included
studies'. The studies examined three comparisons: two-stage Risk of bias in included studies
versus one-stage long posterior flap BKA (Fisher 1988); skew flaps See Figure 4; Figure 5.
 
Figure 4.   Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages
across all included studies.

 
 

Type of incision for below knee amputation (Review) 9

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Figure 5.   Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

 
Allocation Other potential sources of bias
The randomisation method was different in each of the three No concerns over other potential sources of bias were identified
included studies: sealed envelopes (stratified by centre) (Ruckley from the three included studies.
1991); random numbers table (Fisher 1988); and year of birth (even
or odd) (Termansen 1977). Allocation concealment was considered Effects of interventions
adequate in two studies (Fisher 1988; Ruckley 1991) because of
Tests for heterogeneity and sensitivity analyses were not possible
the randomisation methods used and was inadequate in one study
in this review as each comparison related to a single study only.
(Termansen 1977) because the allocation could be deduced from
the date of birth. One study compared two-stage BKA (a guillotine amputation at the
ankle as the primary procedure followed by a long posterior flap
Blinding
BKA with primary skin closure as a secondary procedure) with one-
Blinding was impossible in all studies looking at objective stage BKA (long posterior flap BKA with delayed skin closure) in
outcomes such as primary stump healing as the surgical technique 30 participants with wet gangrene of the foot (Fisher 1988). There
used would be obvious to the observer. was significantly better (100%) primary stump healing in the two-
stage group than in the one-stage group (Peto OR 0.08, 95% CI
Incomplete outcome data 0.01 to 0.89). There was no difference between the two groups in
As detailed below, statistical analysis of 'length of hospital stay' post-operative infection rate, reamputation at the same level, or
could not be undertaken due to failure to report standard reamputation at a higher level. Mobility with a prosthetic limb was
deviations in two included studies (Fisher 1988; Ruckley 1991). 47% in the two-stage group and 54% in the one-stage group. This
was not statistically different (Peto OR 0.87, 95% CI 0.43 to 1.78).
Selective reporting Mean length of hospital stay was 44 days in the two-stage group
and 67 days in the one-stage group. Statistical analysis could not be
No concerns over the selective reporting of data were identified applied to this length of stay as the standard deviation of hospital
from the three included studies.
Type of incision for below knee amputation (Review) 10
Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

stay was not given in the paper. See Data and analyses Comparison however more patellar tendon-bearing prostheses were fitted in
1. the latter group (44% versus 55%) (Termansen 1977).

The Joint Vascular Research Group (JVRG) study randomised 191 In the presence of wet gangrene, there was significantly better
participants in 11 vascular centres to skew flaps BKA (n = 98) or long (100%) primary stump healing in the two-stage group than in the
posterior flap BKA (n = 93) (Ruckley 1991). There was no difference one-stage group, although there was no difference between the two
in primary stump healing between the two groups (60% for both groups in post-operative infection rate, reamputation at the same
the skew flaps and long posterior flap) (RR 1.00, 95% CI 0.71 to level, or reamputation at a higher level (Fisher 1988). Mobility with a
1.42). The rates of post-operative wound necrosis, reamputation prosthetic limb was higher in the one-stage group (54% versus 47%)
at the same level, and reamputation at a higher level were again however this was not statistically significant (Fisher 1988).
no different between the groups. Thirty-day mortality and number
of participants fitted with a prosthetic limb were not different Mobility with a prosthetic limb following below knee amputation
between the groups. Mobility with a prosthetic limb was 60% in the (BKA) appears relatively consistent between studies, at 47% to 60%
skew flaps group and 49% in the long posterior flap group, although (Pinzur 1993).
this was not statistically different (RR 1.22, 95% CI 0.94 to 1.58).
Mean length of hospital stay was 36 days in the skew flaps group and Overall completeness and applicability of evidence
42 days in the long posterior flap group. Statistical analysis could The three included studies were published more than 20 years
not be applied as the standard deviation of hospital stay was not ago. As indicated in the Implications for practice section below, in
given in the paper. See Data and analyses Comparison 2. modern vascular surgery both skew flaps and long posterior flap
below knee amputations continue to be used. Length of stay data
The last study compared 41 participants treated with sagittal flaps
quoted in the included studies look to be outside what would now
BKA to 47 participants with a long posterior flap (Termansen 1977).
be deemed acceptable in modern clinical practice as the limiting
There was no difference in primary stump healing between the
factor for discharge is often social care rather than a requirement
two groups (58% for sagittal flaps, 55% for long posterior flap)
for inpatient medical care.
(Peto OR 1.04, 95% CI 0.45 to 2.43). The rates of reamputation
at the same level, reamputation at a higher level, and mortality One could conclude from Ruckley 1991 that the choice of technique
after three months were no different between the two groups. is a matter of surgeon preference. However, nearly all the members
Overall percentages of participants fitted with a prosthetic limb of the JVRG group were new to the skew flap operation and were
were 78% in the sagittal flaps group and 72% in the long posterior therefore on a learning curve, which may have influenced the
flap group, although this difference was not statistically significant results.
(RR 1.08, 95% CI 0.85 to 1.37). Numbers of participants mobilising
with a prosthetic limb were not stated. However, the study Quality of the evidence
quoted numbers of participants fitted with a patellar tendon-
bearing prosthesis, which one would normally associate with good The overall body of evidence to support one surgical technique over
mobility: 44% in the sagittal flaps group compared to 55% in the another for BKA is limited to three studies which are different in
long posterior flap group (RR 0.79, 95% CI 0.52 to 1.22). See Data their scope. Overall, there is a low risk of selection bias in two out
and analyses Comparison 3. of the three studies; unclear risk of bias for blinding and incomplete
outcome data; and low risk for selective reporting and any other
The predefined secondary outcomes symptoms relating to the sources of bias. Therefore, the overall strength of the evidence
stump, such as pain and swelling, phantom limb pain and quality could be considered as 'moderate'.
of life measures, were not reported in the three included studies.
Potential biases in the review process
DISCUSSION No obvious bias was identified in the review process.
Summary of main results Agreements and disagreements with other studies or
Only a limited number of trials were identified which addressed reviews
these important issues. The most important question to be
To the review authors' knowledge no other evidence has addressed
answered was whether the skew flap technique conferred any
this issue.
advantage over the long posterior flap amputation in terms of
primary stump healing, reamputation rate and return to full The type of prosthesis used was not mentioned by Ruckley 1991.
mobility with a prosthetic limb. The single multicentre Joint Trial participants would have accessed different limb fitting centres
Vascular Research Group (JVRG) study addressed this issue and that would have decided which prosthesis was appropriate for an
found no significant difference between the two techniques in all individual patient. Mobility with a prosthetic limb following BKA
outcome measures (Ruckley 1991). It could be concluded from this appears relatively consistent between studies, at 47% to 60%, and
study that the choice of technique is a matter of surgeon preference. this appears to reflect current clinical practice. This contrasts with
the report from Pinzur which suggests that 87% of participants
There was no difference in primary stump healing, rates of
could maintain independent ambulation following BKA (Pinzur
reamputation at the same level, reamputation at a higher level,
1993).
and mobility after three months between sagittal flaps and the
long posterior flap (Termansen 1977). A higher percentage of
participants were fitted with prosthetic limb in the sagittal flaps
group than in the long posterior flap group (78% versus 72%),

Type of incision for below knee amputation (Review) 11

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

AUTHORS' CONCLUSIONS a technique, the extent of non-viable tissue, and the location of

pre-existing surgical scars, for example from a previous vascular
Implications for practice reconstruction.
Sagittal flaps below knee amputation is seldom used in clinical Implications for research
practice. There is no evidence that this confers any advantage to
the long posterior flap technique. The main question in clinical A further RCT comparing skew flaps to the long posterior flap,
practice is whether the skew flaps technique improves the outcome involving surgeons with experience of both techniques, should be
compared with the long posterior flap. The single multicentre considered. This would negate the possible effects of a learning
randomised controlled trial (RCT) that was identified demonstrated curve, which might have influenced the JVRG study.
no obvious benefit of one technique over the other. We conclude
that there is no evidence to show a benefit of one technique over ACKNOWLEDGEMENTS
the other except in the presence of wet gangrene, where a two-
We would like to thank Mr Michael J Callam for his work on previous
stage procedure has a benefit of better primary stump healing over
versions of this review.
a one-stage procedure. Factors which might influence the choice
of one technique versus the other include previous experience of We would like to thank the Cochrane Consumer Network for
providing a plain language summary.

Type of incision for below knee amputation (Review) 12

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

REFERENCES
 
References to studies included in this review Harrison 1987
Fisher 1988 {published data only} Harrison JD, Southworth S, Callum KG. Experience with the
'skew flap' below-knee amputation. British Journal of Surgery
Fisher DF Jr, Clagett GP, Fry RE, Humble TH, Fry WJ. One-stage
1987;74(10):930-1.
versus two-stage amputation for wet gangrene of the lower
extremity: a randomized study. Journal of Vascular Surgery Higgins 2011
1988;8(4):428-33.
Higgins JPT, Green S (editors). Cochrane Handbook for
Ruckley 1991 {published data only} Systematic Reviews of Interventions Version 5.1.0 [updated
March 2011]. The Cochrane Collaboration, 2011. Available from
Ruckley CV, Stonebridge PA, Prescott RJ. Skewflap versus long
[Link].
posterior flap in below-knee amputations: multicenter trial.
Journal of Vascular Surgery 1991;13(3):423-7. Kaufman 1995
Termansen 1977 {published data only} Kaufman JL. Alternative methods for below-knee amputation:
reappraisal of the Kendrick procedure. Journal of the American
Termansen NB. Below-knee amputation for ischaemic
College of Surgeons 1995;181(6):511-6.
gangrene. Prospective, randomised comparison of a transverse
and a sagittal operative technique. Acta Orthopaedica McCullough 1981
Scandinavica 1977;48(3):311-6.
McCullough NC, Harris AR, Hampton FL. Below-knee
  amputation. Atlas of limb prosthetics. St. Louis: CV Mosby, 1981.
Additional references Persson 1974
Alter 1978 Persson BM. Sagittal incision for below-knee amputation in
Alter AH, Moshein J, Elconin KB, Cohen MJ. Below-knee ischaemic gangrene. Journal of Bone and Joint Surgery (British
amputation using the sagittal technique: a comparison with the volume) 1974;56(1):110-4.
coronal amputation. Clinical Orthopaedics and Related Research
1978;Mar-Apr(131):195-201. Pinzur 1993
Pinzur MS, Gottschalk F, Smith D, Shanfield S, de Andrade R,
Burgess 1968a Osterman H, et al. Functional outcome of below-knee
Burgess EM. The below-knee amputation. Bulletin of Prosthetic amputation in peripheral vascular insufficiency. A multicenter
Research 1968;5:19-25. review. Clinical Orthopaedics and Related Research
1993;Jan(286):247-9.
Burgess 1968b
Burgess EM. The stabilization of muscles in lower extremity Robinson 1976
amputations. Proceedings of the 81st annual meeting of the Robinson KP. Long posterior flap amputation in geriatric
American Orthopaedic Association, 1968 Apr 22-25; Boca Raton patients with ischaemic disease. Annals of the Royal College of
(Florida), Journal of Bone and Joint Surgery (American version) Surgeons of England 1976;58(6):440-51.
1968;50A:1486-7.
Robinson 1982
Catre 1997 Robinson KP, Hoile R, Coddington T. Skew flap myoplastic
Catre MG, Lieberman IH. Laterally based skin flap for below- below-knee amputation: a preliminary report. British Journal of
knee amputation: case report. Journal of Trauma Injury Infection Surgery 1982;69(9):554-7.
and Critical Care 1997;43(5):869-71.
Robinson 1991
Chavatzas 1975 Robinson KP. Skew-flap below-knee amputation. Annals of the
Chavatzas D, Budak D, Jamieson CW. An assessment of value Royal College of Surgeons of England 1991;73(3):155-7.
of long posterior flaps in below knee amputation by skin blood
pressure. Journal of Cardiovascular Surgery 1975;16(6):594-6. Silbert 1950
Silbert S, Haimovici H. Results of midleg amputations for
Galvao 1975 gangrene in diabetics. Journal of the American Medical
Galvao MS. An improved technique for below knee amputation. Association 1950;144:454.
Journal of Cardiovascular Surgery 1975;16(6):603-8.
VSGBI 2010
Haimovici 1996 The Vascular Society of Great Britain and Ireland. Quality
Haimovici H. Below-the-knee amputations. In: Haimovici improvement framework for major amputation surgery. http://
H, Ascer E, Hollier LH, Strandness DE, Towne JB editor(s). [Link]/news-and-press/2010/47-quality-
Haimovici's Vascular Surgery. 4th Edition. Cambridge, [Link].
Massachusetts: Blackwell Science, 1996:1333-40. November 2010 (accessed July 2013).

Type of incision for below knee amputation (Review) 13

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

VSSGBI 1996 Vascular Surgical Society of Great Britain and Ireland. European
Anonymous. Recommendations for the management of chronic Journal of Vascular and Endovascular Surgery 1996;12(2):131-5.
critical lower limb ischaemia. The Audit Committee of the
 
CHARACTERISTICS OF STUDIES

Characteristics of included studies [ordered by study ID]

 
Fisher 1988 
Methods Duration of recruitment to study: 28 months
Duration of follow-up: 2 months after discharge from hospital
Parallel trial
Randomisation method: random numbers table

Participants 30 participants
Age: not stated
Sex: not stated
Inclusion criteria: necrotising wet gangrene of the foot; participants were stratified for presence or ab-
sence of diabetes
Exclusion criteria: suitable for lesser (minor) amputation
Dropouts: none

Interventions Two-stage amputation (guillotine amputation at ankle followed by long posterior flap BKA with prima-
ry skin closure, n = 17) versus one-stage amputation (long posterior flap BKA with delayed skin closure,
n = 13)

Outcomes 1. Primary stump healing

2. Post-operative infection rate
3. Rate of reamputation at same level
4. Rate of reamputation at higher level
5. Length of hospital stay
6. Number of participants mobilising with a prosthetic limb

Notes Above knee amputation cases were excluded from data analyses

Risk of bias

Bias Authors' judgement Support for judgement

Random sequence genera- Low risk Randomisation by random numbers table (Fisher 1988)
tion (selection bias)

Allocation concealment Low risk Adequate measures were undertaken for allocation concealment (random
(selection bias) numbers table)

Blinding (performance Unclear risk Blinding was impossible due to the nature of the procedure undertaken
bias and detection bias)
All outcomes

Incomplete outcome data Unclear risk Statistical analysis of outcome 'length of stay' could not be undertaken due to
(attrition bias) failure to report standard deviations
All outcomes

Selective reporting (re- Low risk No concerns over the selective reporting of data
porting bias)

Other bias Low risk No concerns over other bias

Type of incision for below knee amputation (Review) 14

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

 
 
Ruckley 1991 
Methods Duration of recruitment to study: not stated
Duration of follow-up: 6 months
Parallel trial
Randomisation method: sealed envelopes, stratified by centre

Participants 191 participants in 11 centres (Joint Vascular Research Group)

Age (mean (range)): 70 years (35 to 93) skew flaps; 72 years (39 to 92) long posterior flap
Sex (M:F): 65:33 skew flaps; 60:33 long posterior flap
Inclusion criteria: critical limb ischaemia where no other treatment option available
Exclusion criteria: none stated

Interventions Skew flaps BKA (n = 98) versus long posterior flap BKA (n = 93)

Outcomes 1. Primary stump healing (1 week)

2. Post-operative wound necrosis
3. Rate of reamputation at same level
4. Rate of reamputation at higher level
5. 30-day mortality
6. Length of hospital stay
7. Number of participants fitted with a prosthetic limb
8. Number of participants mobilising with a prosthetic limb

Notes Length of stay (mean (range)): skew flap 36 days (7 to 409); posterior flap 42 days (6 to 385)

Risk of bias

Bias Authors' judgement Support for judgement

Random sequence genera- Low risk Randomisation by sealed envelopes (stratified by centre)
tion (selection bias)

Allocation concealment Low risk Adequate measures were undertaken for allocation concealment (sealed en-
(selection bias) velopes)

Blinding (performance Unclear risk Blinding was impossible due to the nature of the procedure undertaken
bias and detection bias)
All outcomes

Incomplete outcome data Unclear risk Statistical analysis of outcome 'length of stay' could not be undertaken due to
(attrition bias) failure to report standard deviations
All outcomes

Selective reporting (re- Low risk No concerns over the selective reporting of data
porting bias)

Other bias Low risk No concerns over other bias

 
 
Termansen 1977 
Methods Duration of recruitment to study: 30 months
Duration of follow-up (mean (range)): 11.7 months (3.5 to 22)
Parallel trial
Randomisation method: year of birth (even/odd)

Type of incision for below knee amputation (Review) 15

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Termansen 1977  (Continued)
Participants 88 participants
Age (mean in years): 70.0 sagittal flap; 70.5 long posterior flap
Sex (M:F): 24:17 sagittal flap; 22:25 long posterior flap
Inclusion criteria: acute limb ischaemia, critical limb ischaemia, diabetes, mixed arterio-venous ulcera-
tion
Exclusion criteria: none stated

Interventions Sagittal flaps BKA (n = 41) versus long posterior flap BKA (n = 47)

Outcomes 1. Primary stump healing

2. Rate of reamputation at same level
3. Rate of reamputation at higher level
4. Mortality at 3 months
5. Number of participants fitted with a prosthetic limb
- overall
- below knee prosthesis
- patellar tendon-bearing (PTB) prosthesis

Notes No exclusion criteria stated, e.g. those participants not suitable for a long posterior flap (previous vas-
cular surgery) or refusal to enter study

Risk of bias

Bias Authors' judgement Support for judgement

Random sequence genera- High risk Randomisation by year of birth (even or odd)
tion (selection bias)

Allocation concealment High risk Inadequate as this was based on the year of birth
(selection bias)

Blinding (performance Unclear risk Blinding was impossible due to the nature of the procedure undertaken
bias and detection bias)
All outcomes

Incomplete outcome data Low risk No concerns over incomplete outcome data
(attrition bias)
All outcomes

Selective reporting (re- Low risk No concerns over the selective reporting of data
porting bias)

Other bias Low risk No concerns over other bias

BKA: below knee amputation

PTB: patellar tendon bearing
 

 
DATA AND ANALYSES
 

Type of incision for below knee amputation (Review) 16

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Comparison 1.   Two-stage amputation versus one-stage amputation

Outcome or subgroup title No. of No. of Statistical method Effect size

studies partici-
pants

1 Failed primary stump healing 1   Peto Odds Ratio (Peto, Fixed, Totals not selected
95% CI)

2 Post-operative infection rate 1   Peto Odds Ratio (Peto, Fixed, Totals not selected
95% CI)

3 Reamputation at same level 1   Peto Odds Ratio (Peto, Fixed, Totals not selected
95% CI)

4 Reamputation at higher level 1   Peto Odds Ratio (Peto, Fixed, Totals not selected
95% CI)

5 Mobility with prosthetic limb 1   Risk Ratio (M-H, Fixed, 95% CI) Totals not selected

 
 
Analysis 1.1.   Comparison 1 Two-stage amputation versus one-
stage amputation, Outcome 1 Failed primary stump healing.
Study or subgroup Two-stage amputation One-stage amputation Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Fisher 1988 0/17 3/13 0.08[0.01,0.89]

Favours two-stage 0.005 0.1 1 10 200 Favours one-stage

 
 
Analysis 1.2.   Comparison 1 Two-stage amputation versus one-
stage amputation, Outcome 2 Post-operative infection rate.
Study or subgroup Two-stage amputation One-stage amputation Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Fisher 1988 0/17 2/13 0.09[0.01,1.58]

Favours two-stage 0.001 0.1 1 10 1000 Favours one-stage

 
 
Analysis 1.3.   Comparison 1 Two-stage amputation versus one-
stage amputation, Outcome 3 Reamputation at same level.
Study or subgroup Two-stage amputation One-stage amputation Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Fisher 1988 0/17 0/13 Not estimable

Favours two-stage 0.02 0.1 1 10 50 Favours one-stage

 
 

Type of incision for below knee amputation (Review) 17

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Analysis 1.4.   Comparison 1 Two-stage amputation versus one-

stage amputation, Outcome 4 Reamputation at higher level.
Study or subgroup Two-stage amputation One-stage amputation Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Fisher 1988 0/17 1/13 0.1[0,5.19]

Favours two-stage 0.001 0.1 1 10 1000 Favours one-stage

 
 
Analysis 1.5.   Comparison 1 Two-stage amputation versus one-
stage amputation, Outcome 5 Mobility with prosthetic limb.
Study or subgroup Two-stage amputation One-stage amputation Risk Ratio Risk Ratio
  n/N n/N M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Fisher 1988 8/17 7/13 0.87[0.43,1.78]

Favours one-stage 0.1 0.2 0.5 1 2 5 10 Favours two-stage

 
 
Comparison 2.   Skew flaps amputation versus long posterior flap amputation

Outcome or subgroup title No. of No. of Statistical method Effect size

studies partici-
pants

1 Failed primary stump healing 1   Risk Ratio (M-H, Fixed, 95% CI) Totals not selected

2 Post-operative wound necrosis 1   Peto Odds Ratio (Peto, Fixed, Totals not selected
95% CI)

3 Reamputation at same level 1   Peto Odds Ratio (Peto, Fixed, Totals not selected
95% CI)

4 Reamputation at higher level 1   Peto Odds Ratio (Peto, Fixed, Totals not selected
95% CI)

5 Fitted with prosthetic limb 1   Risk Ratio (M-H, Fixed, 95% CI) Totals not selected

6 Mobility with prosthetic limb 1   Risk Ratio (M-H, Fixed, 95% CI) Totals not selected

7 30-day mortality 1   Peto Odds Ratio (Peto, Fixed, Totals not selected
95% CI)

 
 
Analysis 2.1.   Comparison 2 Skew flaps amputation versus long
posterior flap amputation, Outcome 1 Failed primary stump healing.
Study or subgroup Skew flaps Long posterior flap Risk Ratio Risk Ratio
  n/N n/N M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Ruckley 1991 39/98 37/93 1[0.71,1.42]

Favours skew flaps 0.1 0.2 0.5 1 2 5 10 Favours long post flap

 
Type of incision for below knee amputation (Review) 18
Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

 
Analysis 2.2.   Comparison 2 Skew flaps amputation versus long
posterior flap amputation, Outcome 2 Post-operative wound necrosis.
Study or subgroup Skew flaps Long posterior flap Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Ruckley 1991 21/98 23/93 0.83[0.42,1.63]

Favours skew flaps 0.1 0.2 0.5 1 2 5 10 Favours long post flap

 
 
Analysis 2.3.   Comparison 2 Skew flaps amputation versus long
posterior flap amputation, Outcome 3 Reamputation at same level.
Study or subgroup Skew flaps Long posterior flap Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Ruckley 1991 7/98 7/93 0.95[0.32,2.8]

Favours skew flaps 0.1 0.2 0.5 1 2 5 10 Favours long post flap

 
 
Analysis 2.4.   Comparison 2 Skew flaps amputation versus long
posterior flap amputation, Outcome 4 Reamputation at higher level.
Study or subgroup Skew flaps Long posterior flap Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Ruckley 1991 10/98 7/93 1.39[0.51,3.75]

Favours skew flaps 0.1 0.2 0.5 1 2 5 10 Favours long post flap

 
 
Analysis 2.5.   Comparison 2 Skew flaps amputation versus long
posterior flap amputation, Outcome 5 Fitted with prosthetic limb.
Study or subgroup Skew flaps Long posterior flap Risk Ratio Risk Ratio
  n/N n/N M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Ruckley 1991 64/98 50/93 1.21[0.96,1.54]

Favours long post flap 0.1 0.2 0.5 1 2 5 10 Favours skew flaps

 
 
Analysis 2.6.   Comparison 2 Skew flaps amputation versus long
posterior flap amputation, Outcome 6 Mobility with prosthetic limb.
Study or subgroup Skew flaps Long posterior flap Risk Ratio Risk Ratio
  n/N n/N M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Ruckley 1991 59/98 46/93 1.22[0.94,1.58]

Favours long post flap 0.1 0.2 0.5 1 2 5 10 Favours skew flaps

 
 

Type of incision for below knee amputation (Review) 19

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Analysis 2.7.   Comparison 2 Skew flaps amputation versus

long posterior flap amputation, Outcome 7 30-day mortality.
Study or subgroup Skew flaps Long posterior flap Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Ruckley 1991 11/98 16/93 0.61[0.27,1.38]

Favours skew flaps 0.1 0.2 0.5 1 2 5 10 Favours long post flap

 
 
Comparison 3.   Sagittal flaps amputation versus long posterior flap amputation

Outcome or subgroup title No. of No. of Statistical method Effect size

studies partici-
pants

1 Failed primary stump heal- 1   Peto Odds Ratio (Peto, Fixed, 95% CI) Totals not selected
ing

2 Reamputation at same lev- 1   Peto Odds Ratio (Peto, Fixed, 95% CI) Totals not selected
el

3 Reamputation at higher 1   Peto Odds Ratio (Peto, Fixed, 95% CI) Totals not selected
level

4 Fitted with prosthetic limb 1   Risk Ratio (M-H, Fixed, 95% CI) Totals not selected
- overall

5 Fitted with prosthetic limb - 1   Risk Ratio (M-H, Fixed, 95% CI) Totals not selected
below knee prosthesis

6 Fitted with prosthetic limb - 1   Risk Ratio (M-H, Fixed, 95% CI) Totals not selected
PTB prosthesis

7 Mortality at 3 months 1   Peto Odds Ratio (Peto, Fixed, 95% CI) Totals not selected

 
 
Analysis 3.1.   Comparison 3 Sagittal flaps amputation versus long
posterior flap amputation, Outcome 1 Failed primary stump healing.
Study or subgroup Sagittal flaps Long posterior flap Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Termansen 1977 17/41 19/47 1.04[0.45,2.43]

Favours sagittal flaps 0.1 0.2 0.5 1 2 5 10 Favours long post flap

 
 
Analysis 3.2.   Comparison 3 Sagittal flaps amputation versus long
posterior flap amputation, Outcome 2 Reamputation at same level.
Study or subgroup Sagittal flaps Long posterior flap Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Termansen 1977 1/41 4/47 0.33[0.05,1.97]

Favours sagittal flaps 0.01 0.1 1 10 100 Favours long post flap

Type of incision for below knee amputation (Review) 20

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

 
 
Analysis 3.3.   Comparison 3 Sagittal flaps amputation versus long
posterior flap amputation, Outcome 3 Reamputation at higher level.
Study or subgroup Sagittal flaps Long posterior flap Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Termansen 1977 14/41 10/47 1.9[0.75,4.84]

Favours sagittal flaps 0.1 0.2 0.5 1 2 5 10 Favours long post flap

 
 
Analysis 3.4.   Comparison 3 Sagittal flaps amputation versus long
posterior flap amputation, Outcome 4 Fitted with prosthetic limb - overall.
Study or subgroup Favours sagittal flaps Long posterior flap Risk Ratio Risk Ratio
  n/N n/N M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Termansen 1977 32/41 34/47 1.08[0.85,1.37]

Favours long post flap 0.1 0.2 0.5 1 2 5 10 Favours sagittal flaps

 
 
Analysis 3.5.   Comparison 3 Sagittal flaps amputation versus long posterior flap
amputation, Outcome 5 Fitted with prosthetic limb - below knee prosthesis.
Study or subgroup Sagittal flaps Long posterior flap Risk Ratio Risk Ratio
  n/N n/N M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Termansen 1977 25/41 29/47 0.99[0.71,1.38]

Favours long post flap 0.1 0.2 0.5 1 2 5 10 Favours sagittal flaps

 
 
Analysis 3.6.   Comparison 3 Sagittal flaps amputation versus long posterior
flap amputation, Outcome 6 Fitted with prosthetic limb - PTB prosthesis.
Study or subgroup Sagittal flaps Long posterior flap Risk Ratio Risk Ratio
  n/N n/N M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Termansen 1977 18/41 26/47 0.79[0.52,1.22]

Favours long post flap 0.1 0.2 0.5 1 2 5 10 Favours sagittal flaps

 
 
Analysis 3.7.   Comparison 3 Sagittal flaps amputation versus
long posterior flap amputation, Outcome 7 Mortality at 3 months.
Study or subgroup Sagittal flaps Long posterior flap Peto Odds Ratio Peto Odds Ratio
  n/N n/N Peto, Fixed, 95% CI Peto, Fixed, 95% CI
Termansen 1977 4/41 8/47 0.54[0.16,1.83]

Favours sagittal flaps 0.2 0.5 1 2 5 Favours long post flap

Type of incision for below knee amputation (Review) 21

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

APPENDICES

Appendix 1. CENTRAL search strategy

 
 
#1 MeSH descriptor: [Amputation] explode all trees 304

#2 MeSH descriptor: [Knee] explode all trees and with qualifiers: [Surgery - SU] 156

#3 amput*:ti,ab,kw  (Word variations have been searched) 915

#4 below near/3 knee:ti,ab,kw  (Word variations have been searched) 250

#5 #1 or #2 or #3 or #4  1227

#6 MeSH descriptor: [Surgical Flaps] explode all trees 849

#7 transverse near/3 technique:ti,ab,kw  (Word variations have been searched) 7

#8 *flap:ti,ab,kw  (Word variations have been searched) 1418

#9 incision  4167

#10 #6 or #7 or #8 or #9  5723

#11 #5 and #10 in Trials 23

 
WHAT'S NEW
 
Date Event Description

5 August 2013 New search has been performed Searches re-run; no new studies were identified.

5 August 2013 New citation required but conclusions Searches were re-run and no new studies were identified. New
have not changed author joined review team, methods updated to reflect current
Cochrane standards, risk of bias tables completed. Conclusions
not changed.

 
HISTORY
Protocol first published: Issue 3, 2002
Review first published: Issue 1, 2004

 
Date Event Description

23 July 2008 New search has been performed Searches re-run and no new trials found. The review was as-
sessed as up to date.

23 July 2008 Amended Converted to new review format.

14 December 2007 Amended New plain language summary added.

Type of incision for below knee amputation (Review) 22

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 Cochrane Trusted evidence.
Informed decisions.
 
 
Library Better health. Cochrane Database of Systematic Reviews

Date Event Description

3 April 2006 New search has been performed Searches re-run and no new trials found. The review was updat-
ed with no changes other than dates of last search.

 
CONTRIBUTIONS OF AUTHORS
Paul Tisi: identified trials for inclusion; contacted authors for additional information; assessed eligibility and quality of trials; extracted
data; and wrote the review.

Mary Than: assessed quality of the trials and updated the review.

DECLARATIONS OF INTEREST
None known

SOURCES OF SUPPORT

Internal sources
• No sources of support supplied

External sources
• Chief Scientist Office, Scottish Government Health Directorates, The Scottish Government, UK.

The PVD Group editorial base is supported by the Chief Scientist Office.

DIFFERENCES BETWEEN PROTOCOL AND REVIEW
In line with current Cochrane Collaboration guidelines the assessment of the methodological quality of the studies was conducted
according to Higgins 2011.

NOTES
The title of the protocol for this review was: 'Type of skin incision for below knee amputation'.

INDEX TERMS

Medical Subject Headings (MeSH)

Amputation  [*methods];  Diabetic Foot  [*surgery];  Gangrene  [complications]  [surgery];  Ischemia  [*surgery];  Leg  [*blood supply]; 
Randomized Controlled Trials as Topic;  Reoperation;  Surgical Flaps;  Treatment Outcome;  Wound Healing

MeSH check words

Humans

Type of incision for below knee amputation (Review) 23

Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
 ORIGINAL ARTICLES
EUR MED PHYS 1994;30:1-6

Measuring long-term outcome in people

with lower limb amputation:
cross-validation of the Italian versions

A
of the Prosthetic Profile of the Amputee
and Prosthesis Evaluation Questionnaire

C
T® I
G. FERRIERO 1, D. DUGHI 1, D. ORLANDINI 2, T. MOSCATO 2, D. NICITA 2, F. FRANCHIGNONI 1

H ED
Aim. Recently 2 questionnaires have been developed for 1Unit of Occupational Rehabilitation and Ergonomics
IG M
people with lower limb amputation to determine, in Salvatore Maugeri Foundation
follow-up studies, the level of function and extent of Scientific Institute, Veruno (Novara), Italy
prosthetic use, to measure major life domains connect- 2INAIL Prosthesis Center, Vigorso di Budrio (Bologna), Italy
ed with prosthesis function, and to study the factors
potentially related to prosthetic use: the Prosthetic Profile
R A

of the Amputee (PPA) and the Prosthesis Evaluation

Questionnaire (PEQ). The purpose of the present study
was: a) to produce Italian versions of both PPA (PPA-it)
P RV

and PEQ (PEQ-it), using a validated procedure of cross-

cultural translation; b) to analyse and discuss the inter- rate, visual analogue scale format of PEQ, demanding
nal consistency and construct validity of the main sec- scoring procedures) were sub-optimal, and their struc-
tions of the 2 questionnaires, in an Italian population. ture (item selection, response format, scaling properties,
Methods. The PPA questionnaire consists of 44 ques- etc.) would need some refinement and simplification
tions arranged in 6 sections. The PEQ is composed of 82 in order to facilitate a broader clinical use. On the oth-
O E

questions subdivided into 9 scales related to 4 sectors. In er hand, findings for the LCI/5 (PPA) and MO (PEQ) are
order to produce the Italian versions of the PPA and encouraging (particularly regarding the first scale) and
PEQ the forward/backward translation method was confirm their sound practical and psychometric fea-
C N
Y

used. The final versions of the questionnaires were tures.

mailed to 110 patients and 95 of them returned the ques- Key words: Questionnaires - Leg - Amputation.
tionarries.
I

Results. The Cronbach’s α of Locomotor Capabilities

Index (LCI/5), part of the PPA, was 0.97, and those of the
M

9 PEQ-it scales ranged from 0.64 (appearance) to 0.95

(mobility, MO). The LCI/5 and MO correlated highly T he ultimate goal of rehabilitation in people with
lower limb amputation (PLLA) is the best possible
reintegration of these persons in their environment,
with each other (rs=0.81) and with the variables related
to prosthesis use. There was a significant correlation with an optimal physical, mental, emotional, social,
among the PEQ-it domains concerning MO, prosthesis vocational and economic efficiency.1 This intervention
function, psycho-social aspects and well-being. requires a multidisciplinary approach and has as its
Conclusion. Unfortunately, the acceptability and feasi- cornerstone the restoration of locomotor function
bility of both questionnaires (rather low completion through use of prosthetic devices.2 There is an increas-
Submitted for publication September 24, 2004.
ing demand for objective reporting of outcomes in
Accepted February 2, 2005. this field, in order to better understand the main rea-
sons for long-term use/disuse behaviour after dis-
charge from rehabilitation, including the typical prob-
Address reprint requests to: F. Franchignoni, MD, Fondazione Salvatore lems encountered by an amputee at home or in the
Maugeri, Clinica del Lavoro e della Riabilitazione, IRCCS, Via Revislate 13,
28010 Veruno (NO), Italy. E-mail: ffranchignoni@[Link] community.3

Vol. 41 - No. 1 EUROPA MEDICOPHYSICA 1

 FERRIERO MEASURING LONG-TERM OUTCOME IN PEOPLE WITH LOWER LIMB AMPUTATION

In recent years, 2 questionnaires have been devel- 4) physical and social environments: living arrange-
oped and validated for follow-up studies in PLLA: the ments, place of abode, architectural barriers, physical
Prosthetic Profile of the Amputee (PPA) 4, 5 and the help needed etc. (10 items);
Prosthesis Evaluation Questionnaire (PEQ).6 Their 5) leisure: sports and recreational activities (6 items);
aim is to determine the level of function and extent of 6) general information (current vocational status,
prosthetic use (in terms of frequency, duration of use,

A
education, socio-economic status) (5 items). Questions
mobility, activities done etc.), to measure major life are for the most part close-ended or semi-closed, of
domains connected with prosthesis function, and to nominal or ordinal type. In the present study we used

C
study the factors potentially related to prosthetic use. the postal version of the questionnaire, but there is also
The original version of both measures (PPA, PEQ) a telephone version.
demonstrated good reliability and validity,4-9 but val- The PEQ 6, 13 is composed of 82 questions subdi-

T ® DI
idation is an ongoing process and the replication and
vided into 9 scales related to 4 sectors: 1) prosthesis
integration of psychometric studies is always recom-
function: utility (UT, 8 items), residual limb health
mended, particularly when measures have to be used
(RL, 6 items), appearance (AP, 5 items), sounds (SO,
in other than the source context.10
H E 2 items); 2) mobility: ambulation (AM, 8 items), trans-
The purpose of the present study was twofold: a)
fer (TR, 5 items); 3) psycho-social aspects: frustration
to produce Italian versions of both PPA and PEQ using
a validated procedure of cross-cultural translation; b) (FR, 2 items), perceived response (PR, 5 items), social
IG M
to analyse and discuss the internal consistency and burden (SB, 3 items); 4) well being (WB, 2 items). In
construct validity of the main sections of the 2 trans- accordance with Miller et al.,9 we combined the 2
lated questionnaires when used with an Italian pop- mobility scales AM and TR into a single mobility mea-
ulation. sure (MO, 13 items).
In addition the PEQ contains many individual ques-
R A

tions which are not meant to be included in the scale

scores. These are: satisfaction in particular situations
Materials and methods
Y V

(3 items), specific body sensations (stump pain, phan-

tom limb, non-painful sensations etc.) (16 items),
Instruments prosthetic care (3 items), self-efficacy (i.e. ability to do
P R

The PPA questionnaire 4, 5, 11 consists of 44 questions daily activities under difficult conditions) (3 items),
arranged in 6 sections: and questions about the importance of different
aspects of experience with the prosthesis (10 items).
O E

1) physical condition of the individual in terms of

level of amputation, particular health problems and Most questions in the PEQ (76 out of 82) use a visu-
problems with the stump (4 items); al analogue scale (VAS) format, scored as a continu-
C IN

2) satisfaction regarding specific characteristics of the ous numerical variable measured as the distance in
prosthesis and adaptation either to the amputation mm from the left endpoint of the 100-mm line (bound-
or to the prosthesis (5 items); ed by 2 anchor phrases denoting the extremes of pos-
3) use of the prosthesis in terms of time of use, sible answers) to the point at which the respondent’s
M

percentage of displacements done with the prosthe- mark crosses the line. For each scale a summary score
sis, and causes of not-wearing (14 items). One ques- can be calculated giving the arithmetical mean of all
tion, number 11, consists of an index called Locomotor the questions which make up that particular scale
Capabilities Index (LCI) that contains 14 items relat- and which the respondent answered. There is a PEQ
ed to autonomy in postural passages and deambula- guide for scoring and analysis containing detailed
tion and that can be used, if desired, separately from coding instructions for all the questions.13
the PPA. To evaluate subjects’ capability to perform
each of the 14 activities, we adopted the version with Procedure
a 5 point ordinal scale (LCI/5), ranging from 0=not able
CROSS-CULTURAL TRANSLATION
to 4=able to accomplish the activity without aids. The
LCI/5 presents similar psychometric properties to the The procedure used to produce the Italian versions
original version but has a lower ceiling effect and of the PPA and PEQ was the forward/backward trans-
larger effect size;12 lation method,14 involving the following steps:

2 EUROPA MEDICOPHYSICA March 2005

 MEASURING LONG-TERM OUTCOME IN PEOPLE WITH LOWER LIMB AMPUTATION FERRIERO

1) translation into Italian of the original versions of sum of the remaining items (item-total correlation,
the PPA and PEQ: 2 English-Italian translators, native- ITC) and by calculation of Cronbach’s α coeffi-
Italian speakers with 13-19 years of education, were cient.15
involved. Each translator independently translated The construct validity of the PPA-it and PEQ-it was
the PPA and the PEQ, and then compared and dis- analysed as follows:
cussed the result with the other translator in order to

A
1) convergent validity - testing the a priori
arrive at a unified preliminary version; hypotheses to find a significant correlation: a)
2) back-translation of the Italian versions of the between the 2 scales investigating the locomotor

C
PPA and PEQ into English: the preliminary versions capabilities (LCI/5 of the PPA-it and MO of the PEQ-
were given to 2 native-English translators with 13-17 it); b) between both LCI/5 and MO scales and oth-
years of education, who each produced a translation er related constructs of PPA-it: frequency of pros-

T ® DI
into English. These translators were unaware of either thesis use, percentage of displacements done respec-
the methodology or the aims of the study; tively with a wheelchair and with the prosthesis
3) convergence towards a final version of each (indoors and outdoors), distances walked, and use
questionnaire: a multi-disciplinary panel made up of
H E of walking aids; c) among the PEQ-it domains con-
4 bilingual native-Italian experts in the field of cerning MO, UT scale of prosthesis function, psy-
Rehabilitation (2 physiatrists, 1 psychologist, 1 phys- cho-social aspects (FR, PR, SB), and WB. For all
IG M
ical therapist) evaluated the discrepancies between these purposes the Spearman rank-correlation coef-
the back-translations and potential ambiguities result- ficient corrected for ties was used.
ing from semantic and/or idiomatic peculiarities of 2) discriminant validity - we calculated the ability of
the English terms. The preliminary versions thus LCI/5 and MO to respectively distinguish between
obtained underwent a pretesting study, where 8 persons with unilateral below-knee (BK) and above-
R A

patients—upon completion of the questionnaire— knee (AK) amputation, using the Mann-Whitney test.
were interviewed by a psychologist about the clarity,
Y V

simplicity, and potential ambiguity of each item and

their administration procedures. The final versions Results
were defined PPA-it and PEQ-it.
P R

Ninety-five out of 110 study participants (86%)

VALIDATION OF THE PPA-IT AND PEQ-IT returned the envelop with the questionnaires. All of
O E

The final versions of the questionnaires were them answered the PPA-it, though 87% did not com-
mailed to 110 patients who met the preselection cri- plete all the questions (median number of unanswered
teria, i.e. subjects were required to have had an ampu- questions: 5; interquartile range, IQR: 3-11). Seventy-
C IN

tation of the lower limb in the last 5 years; to have no five subjects (79%) answered the PEQ-it but 14 filled
mental diseases or other conditions that would restrict it in incompletely (median number of unanswered
the use of technical aids for walking; to have com- questions: 3; IQR: 2-7).
pleted the rehabilitation period and the prosthetic Table I shows the descriptive statistics of the main
M

training program at least 6 months before the begin- socio-demographic characteristics of the study pop-
ning of the study; and to have been using the pros- ulation. Bilateral amputees (n=12) did not have to fill
thesis when discharged from rehabilitation. If sub- in questions 3-23 of the PPA. Eighty-nine percent
jects did not return the questionnaires within 4 weeks, (n=74) of the 83 unilateral amputees were prosthetic
a second mailing took place, and 20 days later a users, utilising their prosthesis at least 1 day a week.
phone call was made to non-respondents, as a last Sixty-two prosthetic users (84%) used their prosthesis
reminder. daily: 44 (59%) 9 h per day or more. With respect to
active use of the prosthesis for mobility, 54% of pros-
thetic users reported using the prosthesis to perform
Statistical analysis
the majority (75% or more) of ambulatory activities
The internal consistency of the LCI/5 (PPA-it) and indoors, and 75% to perform activities outdoors.
the 9 scales of the PEQ-it was tested—for each scale Seventy-four percent of prosthetic users (n=55) were
separately—both by correlating each item with the able to don the prosthesis alone without difficulty,

Vol. 41 - No. 1 EUROPA MEDICOPHYSICA 3

 FERRIERO MEASURING LONG-TERM OUTCOME IN PEOPLE WITH LOWER LIMB AMPUTATION

TABLE I.—Sample demographics (n=95). Median values and IQR Construct validity
or percentages.
As for convergent validity, the LCI/5 showed a high
Value (IQR)
correlation with MO (rs=0.81). Table III shows the
Age (years) 54 (37-64) correlation of both LCI/5 and MO with the PPA-it vari-
Gender (male) 72% ables related to prosthesis use. Table IV contains the

A
School education (years) 8.7 (3-20) correlation among the PEQ-it domains concerning
Time since amputation (months) 30 (18-38)
MO, UT scale of prosthesis function, psycho-social
Cause of amputation
aspects (FR, PR, SB) and WB.

C
Peripheral vascular disease 35%
Trauma 55% As regards discriminant validity, persons with uni-
Tumour and other 10% lateral BK amputation showed significant lower LCI/5
scores than persons with unilateral AK amputation

T ® DI
Amputation level
Unilateral, above the knee 50% (BK=28.5, AK=44.9; P<0.001). The difference between
Unilateral, below the knee 37% the respective scores of MO was not significant
Bilateral 13% (BK=30.3, AK=36.7; P=0.12).
Current vocational status The sample size in these correlations related to
H E
Employed
Unemployed
29%
71%
validity issues ranged from n=60 to n=65.
IG M
Discussion and conclusions
9% (n=7) alone with difficulty, and 16% (n=12) only In the cross-cultural translation of both the PPA
with supervision or help. and PEQ no semantic difficulties were found. Specific
Table II presents the median values (and IQR) of the
R A

attention was paid to the translation of the anchor

LCI/5 and of the 9 scales of the PEQ-it, in the 70 sub- phrases of the PEQ-it. The LCI/5 had previously
jects who filled in all the domains. been translated and validated in a dedicated study.12
Y V

As for the completion rate of the 2 questionnaires,

Internal consistency 79% of subjects answered the PEQ-it (64% in com-
plete mode) whereas 100% answered the PPA-it (but
P R

The internal consistency (Cronbach’s α coefficient) only 13% in complete mode). These quite low com-
of the LCI/5 was 0.97, and those of the 9 PEQ-it scales pletion rates may be due to the fact that they are
were as follows: prosthesis function - UT=0.83, RL=
O E

both lengthy to fill in.16-18 In addition, the PEQ has a

0.75, AP=0.64, and SO=0.84; MO=0.95 (AM=0.94, TR VAS format which is unfamiliar and not easily under-
=0.90); psycho-social aspects - FR=0.87, PR=0.87, stood (particularly by those with limited education).9
C IN

SB=0.72; WB=0.72. On the other hand, subjects that understand the VAS
Also the item-remainder correlation using Pearson’s structure may find it relatively simple to answer to all
coefficients gave very high values of rs for the LCI/5 the questions (or most of them), due to their repet-
(rs=0.74-0.93; P<0.0001) and MO (rs=0.57-0.83, P< itive format. In any case, we think that the VAS for-
M

0.0001). The ITC values of the other PEQ-it scales mat of PEQ-it should be converted to a numeric rat-
showed the following ranges: prosthetic function - UT ing scale ranging from 0 to 10, as suggested by Miller
(rs=0.52-0.68), RL (rs=0.38-0.57), AP (rs=0.33-0.50), SO et al.,9 in order to reduce problems in administration
(rs=0.72); psycho-social aspects - FR (rs=0.77), PR (rs=0.64- and scoring.10 Again, the PEQ contains items related
0.73), SB (rs=0.49-0.64); WB (rs=0.56); with P<0.005 for to different scales in the same group of questions:13
RL and AP and P< 0.0001 for the other scales. this—together with VAS format—complicates scoring

TABLE II.—Median values (and IQR) of the LCI/5 (PPA-it) and of the 9 scales of the PEQ-it (n=70).
LCI/5 UT RL AP SO MO FR PR SB WB

41 63 67 73 72 61 57 84 70 69
(30-48) (57-77) (53-81) (59-82) (40-95) (43-76) (32-97) (62-94) (51-87) (52-87)

4 EUROPA MEDICOPHYSICA March 2005

 MEASURING LONG-TERM OUTCOME IN PEOPLE WITH LOWER LIMB AMPUTATION FERRIERO

TABLE III.—Correlation of LCI/5 and MO with factors related to pro- 0.47), and only later were other items included in the
sthetic use.
TR domain.6
LCI/5 MO The Cronbach’s α of the other PEQ-it scales ranged
from 0.64 to 0.87, and was slightly lower than in Legro
Frequency of prosthetic use 0.53 0.63
Use of walking aids, indoors -0.60 -0.61
et al.’s paper (0.73-0.89).6 In particular, AP did not
exceed the minimum desired level of internal con-

A
Use of walking aids, outdoors -0.67 -0.62
Distances walked 0.55 0.46 sistency of 0.70 suggested by Nunnally,20 while SB, WB
% of displacements with a wheelchair, indoors -0.54 -0.58 and RL showed quite low values (≤0.75). This is par-
% of displacements with the prosthesis, indoors 0.46 0.47 tially due to the dependence of α not only on the

C
% of displacements with a wheelchair, outdoors -0.49 -0.44
% of displacements with the prosthesis, outdoors 0.46 0.41 magnitude of the correlation among items but also
on the (low) number of the items in the scale (SB, 3

T ® DI
items; WB, 2 items). However, the ITC values showed
that in both AP and RL domains there are items of
TABLE IV.—Correlation between specific domains of the PEQ-it.
quite low homogeneity with respect to the general
MO UT FR PR SB construct, even if they are above the lower limit of 0.20
UT
FR
PR
H E 0.61
0.65
0.67
—
0.54
0.62
—
0.68 —
suggested by Streiner et al.15 Apart from the general
lack of internal consistency of the AP section that
undermines the meaningfulness of this scale, for RL the
IG M
SB 0.78 0.53 0.68 0.74 — item “…rate any ingrown hairs that were on your
WB 0.67 0.53 0.57 0.70 0.76 residual limb” fitted poorly the construct of its scale
(rs=0.38). Overall, these findings suggest that some
PEQ-it scales should undergo structural refinement.
R A

procedures. As regards construct validity, the LCI/5 showed a

Also the PPA-it is difficult to understand (due to strong correlation, as expected, with MO (r=0.81;
the varied and sometimes complex response formats), n=67) indicating the ability of both scales to measure
Y V

especially for older or illiterate persons; Streppel et al. locomotor capabilities with the prosthesis.
reported a similar difficulty in obtaining complete and Again, this finding is in agreement with previous
P R

properly filled out questionnaires.17 Furthermore, the correlations between the 2 tests, ranging from 0.77
PPA has no total or partial scores (except the LCI/5); to 0.83.9 Only the LCI/5 was able to distinguish
hence, to get an impression of the patient’s function- between BK and AK amputees, while the MO failed
O E

ing all questions have to be reviewed.11 in this discriminative task. This indicates, for this
The median values of the PEQ-it scales show that aspect, a psychometric superiority of the LCI/5 over
our population had a poorer mobility than the sam- the MO scale.
C IN

ple of Legro et al.,6 as well as lower scores in some The locomotor capabilities with the prosthesis (as
prosthetic function scales and psycho-social experi- measured by both LCI/5 and MO) correlated signifi-
ences, probably because of the different selection cri- cantly with the other related constructs of PPA-it: fre-
teria used (“use of the prosthesis at least 5 days a quency of prosthesis use, percentage of displace-
M

week” and “only unilateral amputation”). Conversely, ments done respectively with a wheelchair and with
mobility and well-being of 4 patients as well as the the prosthesis (indoors and outdoors), distances
time since amputation (30 vs 38 months) were similar walked, and use of walking aids. Some of these vari-
to those of the non-traumatic amputees in Harness ables were identified by Gauthier-Gagnon as an
et al. study.19 enabling factor (i.e. ability or human resource, which
Both mobility scales – LCI/5 (PPA-it) and MO (PEQ- allows to realise an objective) related to prosthetic
it) - showed α equal to or greater than 0.95, so con- use, as measured by PPA.8 As hypothesised, previ-
firming the good internal consistency reported by ous observations were confirmed and widened by
Miller et al.9 who used the LCI with items in a 4-lev- our results for both PPA-it and PEQ-it. The relationship
el version (α=0.95) and the MO with numeric rating of both LCI/5 and MO with the other variables was on
scales (α=0.89). Conversely, in 1998 Legro et al. cal- average moderate (rs between 0.41 and 0.67): a little
culated separately the α values for the PEQ mobility higher with frequency of prosthesis use and use of
sub-scales AM (8 items; α=0.89) and TR (2 items; α= walking aids, and lower with distances walked and

Vol. 41 - No. 1 EUROPA MEDICOPHYSICA 5

 FERRIERO MEASURING LONG-TERM OUTCOME IN PEOPLE WITH LOWER LIMB AMPUTATION

percentages of displacements. The moderate correla- chometric features reported in previous studies.7-9, 12
tion between the locomotor abilities and the distance
walked without stopping with the prosthesis can be
explained by many other factors influencing perfor- References
mance (age, comorbidity, motivation, residual limb
1. Geertzen JH, Martina JD, Rietman HS. Lower limb amputation.
health, and so on). Hence, the fair level of correlation

A
Part 2: Rehabilitation- a 10 year literature review. Prosthet Orthot
between the locomotor abilities and the percentage of Int 2001;25:14-20.
displacements done (with both prosthesis and wheel- 2. Pernot HF, Winnubst GM, Cluitmans JJ, De Witte LP. Amputees in
Limburg: incidence, morbidity and mortality, prosthetic supply, care

C
chair, both indoors and outdoors) could be partly utilisation and functional level after one year. Prosthet Orthot Int
attributed to the ambiguity of, and the high level of 2000;24:90-6.
abstraction required by, this group of items (“During 3. Rommers GM, Vos LD, Groothoff JW, Eisma WH. Mobility of peo-
ple with lower limb amputations: scales and questionnaires: a

T ® DI
the day, when you have to move…, approximately review. Clin Rehabil 2001;15:92-102.
what percentage of displacements around are done 4. Grisé MCL, Gauthier-Gagnon C, Martineau GG. Prosthetic profile of
walking with…?” Answer option: 0%, 25%, 50%, 75%, people with lower extremity amputation: conception and design of
a follow-up questionnaire Arch Phys Med Rehabil 1993;74:862-70.
100%). Direct interview with some patients confirmed 5. Gauthier-Gagnon C, Grisé MCL. Prosthetic profile of the amputee
H E
their rather low comprehensibility.
The correlation between MO and all the other PEQ-
it domains related to usefulness of prosthesis function,
questionnaire: validity and reliability. Arch Phys Med Rehabil
1994;75:1309-14.
6. Legro MW, Reiber GD, Smith DG, del Aguila M, Larsen J, Boone
D. Prosthesis evaluation questionnaire for person with lower limb
IG M
psycho-social aspects, and well-being was quite good amputation: assessing prosthesis quality of life. Arch Phys Med
Rehabil 1998;79:931-8.
(rs ranging from 0.61 to 0.78). The strongest relation- 7. Gauthier-Gagnon C, Grisé MCL. The locomotor capabilities index:
ship was with SB. This is in line with the high corre- content validity. J Rehabil Outcomes Meas 1998;2:40-6.
lation reported by Harness et al. in PEQ between 8. Gauthier-Gagnon C, Grise MC, Potvin D. Enabling factors related
to prosthetic use by people with transtibial and transfemoral ampu-
R A

both AM and TR scales and SB.19 In fact, mobility can tation. Arch Phys Med Rehabil 1999;80:706-13.
be thought of as a construct highly influencing the 9. Miller CW, Barry Deathe A, Speechley M. Lower extremity pros-
psycho-social aspects and which in turn is itself influ- thetic mobility: a comparison of 3 self-report scales. Arch Phys Med
Y V

Rehabil 2001;82:1432-40.
enced by the prosthesis function. The PEQ psycho- 10. Fitzpatrick R, Davey C, Buxton MJ, Jones DR. Evaluating patient-
social domains showed (as expected) a significant based outcome measures for use in clinical trials. Health Technol
Assess 1998;2:1-74.
P R

correlation both among themselves and with WB.

11. Gauthier-Gagnon C, Grisé MC. Prosthetic Profile of the Amputee.
Care should be taken not to over-generalize these Handbook of documents developped within the framework of a
results, because our population represents a highly prosthetic follow-up study. Montreal (Quebec); 1996.
O E

selected sample (younger and with more amputees 12. Franchignoni F, Orlandini D, Ferriero G, Moscato TA. Reliability,
validity, and responsiveness of the locomotor capabilities index in
of traumatic origin than the general PLLA). adults with lower-limb amputation undergoing prosthetic training.
Nevertheless, the results of the correlational analy- Arch Phys Med Rehabil 2004;85:743-8.
C IN

sis are fully in line with other similar studies con- 13. Guide for the use of the Prosthesis Evaluation Questionnaire.
Seattle: Prosthetics Research Study; 1998.
ducted in different countries and contexts using the 14. Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of
same instruments.4-9, 17, 19 health-related quality of life measures: literature review and pro-
In conclusion, this study shows that both PPA-it posed guidelines. J Clin Epidemiol 1993;46:1417-32.
15. Streiner DL, Norman GR. Health measurement scales. 2nd ed.
M

and PEQ-it—obtained according to a validated pro- Oxford: Oxford University Press; 1995.
cedure of cross-cultural translation—have a good psy- 16. Devlin M, Pauley T, Head K, Garfinkel S. Houghton scale of pros-
chometric equivalence to their original counterparts. thetic use in people with lower-extremity amputations: reliability,
validity, and responsiveness to change. Arch Phys Med Rehabil
But their acceptability (as measured by response rate 2004;85:1339-44.
and time to complete) and feasibility (i.e. difficulty 17. Streppel KR, de Vries J, van Harten WH. Functional status and
prosthesis use in amputees, measured with the Prosthetic Profile
of administration and processing) are sub-optimal, of the Amputee (PPA) and the short version of the Sickness Impact
and their structure (in terms of item selection, response Profile (SIP68). Int J Rehabil Res 2001;24:251-6.
format, scaling properties, and so on) needs some 18. Callaghan BG, Sockalingam S, Treweek SP, Condie ME. A post-
discharge functional outcome measure for lower limb amputees:
refinement and simplification—also through Rasch test-retest reliability with trans-tibial amputees. Prosthet Orthot
analysis—in order to facilitate a broader clinical use. Int 2002;26:113-9.
Despite this, the results concerning the LCI/5 and 19. Harness N, Pinzur MS. Health related quality of life in patients with
dysvascular transtibial amputation. Clin Orthop 2001;383:204-7.
MO are very encouraging (particularly regarding the 20. Nunnally JC, Bernstein IH. Psychometric theory. 3rd ed. New
first scale), and confirm their sound practical and psy- York: McGraw-Hill; 1994.

6 EUROPA MEDICOPHYSICA March 2005

Health and Quality of Life Outcomes BioMed Central

Research Open Access

The Locomotor Capabilities Index; validity and reliability of the
Swedish version in adults with lower limb amputation
Brita Larsson†1, Anton Johannesson*†2, Ingemar H Andersson3 and
Isam Atroshi4,5

Address: 1Department of Rehabilitation Medicine, Hässleholm Hospital, SE-28125 Hässleholm, Sweden, 2Department of Clinical Sciences, Lund
University, Lund, Sweden, Ortopedteknik AB, Kristianstad Hospital, Kristianstad, Sweden, 3Department of Health and Society, Kristianstad
University, Kristianstad, Sweden, 4Department of Clinical Sciences, Lund University, Lund, Sweden and 5Department of Orthopedics, Hässleholm
and Kristianstad Hospitals, Hässleholm, Sweden
Email: Brita Larsson - [Link]@[Link]; Anton Johannesson* - [Link]@[Link];
Ingemar H Andersson - [Link]@[Link]; Isam Atroshi - [Link]@[Link]
* Corresponding author †Equal contributors

Published: 23 May 2009 Received: 9 April 2008

Accepted: 23 May 2009
Health and Quality of Life Outcomes 2009, 7:44 doi:10.1186/1477-7525-7-44
This article is available from: [Link]
© 2009 Larsson et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License ([Link]
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract
Background: The Locomotor Capabilities Index (LCI) is a validated measure of lower-limb
amputees' ability to perform activities with prosthesis. We have developed the LCI Swedish version
and evaluated its validity and reliability.
Methods: Cross-cultural adaptation to Swedish included forward/backward translations and field
testing. The Swedish LCI was then administered to 144 amputees (55 women), mean age 74 (40–
93) years, attending post-rehabilitation prosthetic training. Construct validity was assessed by
examining the relationship between the LCI and Timed "Up-and-Go" (TUG) test and between the
LCI and EQ-5D health utility index in 2 subgroups of 40 and 20 amputees, respectively.
Discriminative validity was assessed by comparing scores in different age groups and in unilateral
and bilateral amputees. Test-retest reliability (1–2 weeks) was evaluated in 20 amputees (14
unilateral).
Results: The Swedish LCI showed good construct convergent validity, with high correlation with
the TUG (r = -0.75) and the EQ-5D (r = 0.84), and discriminative validity, with significantly worse
mean scores for older than younger and for bilateral than unilateral amputees (p < 0.01), and high
internal consistency (Cronbach alpha 0.95). In test-retest reliability the intraclass correlation
coefficient was 0.91 (95% CI 0.79–0.96) but for the unilateral amputees was 0.83 (95% CI 0.56–
0.94). Ceiling effect occurred in 23%.
Conclusion: The Swedish version of the LCI demonstrated good validity and internal consistency
in adult amputees. Test-retest reliability in a small subsample appears to be acceptable. The high
ceiling effect of the LCI may imply that it would be most useful in assessing amputees with low to
moderate functional abilities.

Page 1 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2009, 7:44 [Link]

Background Methods
Patients with severe peripheral arterial disease or diabetes Procedure of translation
may require lower limb amputation and in Scandinavia The procedure of cross-cultural adaptation of the English
these conditions account for more than 90% of all lower version of the LCI to Swedish was done in three steps [14].
limb amputations [1]. The annual incidence of above-foot First, the English version was translated to Swedish (for-
amputation ranges from 20 to 46 per 100,000 inhabitants ward translation) by 3 translators whose first language
[2,3]. In patients with lower limb amputation the primary was Swedish, with one having no medical background.
aim of rehabilitation is to restore walking ability with Based on consensus meeting a final version was created.
prosthesis. Not all patients can receive prosthesis after In the second step, two bilingual persons whose first lan-
amputation. The reported rate of prosthetic use following guage was English independently re-translated the Swed-
lower limb amputation related to peripheral arterial dis- ish version into English (backward translation). Both
ease or diabetes has varied from 32% to 43% [4-6]. In were blinded to the concepts being investigated and one
addition, amputees successfully fitted with a prosthesis had no medical background. Finally, the translations were
may differ in how much they use the prosthesis and in the reviewed by a group consisting of 2 forward-translators, 1
type of activities they can perform with their prosthesis backward-translator and one supervisor and discrepancies
[7]. were resolved to achieve conceptual equivalence with the
original version.
Walking ability with a prosthesis depends on several fac-
tors including patient's physical and mental status [8], the A pre-final version was created and tested on a reference
surgical method used [9], postoperative care, nutrition group of 10 amputees attending training in a special after-
and pain relief [10] as well as the rehabilitation and pros- rehabilitation training unit for amputees. The pre-final
thetic fitting procedures [6]. Lower limb amputation version performed well in the field-testing. However, the
related to peripheral arterial disease or diabetes is usually reference group suggested that a second version be created
performed on elderly patients who have multiple medical with lines between the questions for better readability as
disorders, and the rehabilitation may be compromised by many amputees suffer from poor vision because of high
other illnesses such as stroke and heart failure or vascular age and/or diabetes. A final Swedish version of the LCI
problems involving the contralateral leg. An instrument was then created (Additional file 1). The data from the
that measures walking ability following amputation can field-testing were not used further in the analysis.
therefore be used to trace changes in function related to
comorbidity, treatment or rehabilitation. Validation study
The Swedish version was assessed for validity (convergent
The Locomotor Capabilities Index (LCI) is a 14-item ques- and discriminative) and reliability (internal consistency
tionnaire specifically designed to measure walking ability and test-retest reliability) in a cross-sectional study con-
of lower-limb amputees. The LCI was developed in Can- ducted on a population of lower limb amputees attending
ada in 1993 as part of the Prosthetic Profile of the training after discharge from the hospital rehabilitation
Amputee questionnaire [11,12]. According to its develop- unit with retest follow-up of a small subsample of the par-
ers the LCI "computes the global, basic, and advanced ticipants.
locomotor skills of the lower limb amputee with the pros-
thesis and assesses level of independence" [13]. The LCI Participants
has demonstrated good validity and reliability in adults Participants from our rehabilitation unit (Hässleholm-
with lower limb amputation and it has been found espe- Kristianstad Hospitals) as well as from three other rehabil-
cially useful in daily clinical practices. It has been trans- itation units in Sweden (one in Gothenburg, and two in
lated to several European languages and used in Stockholm) were recruited for this study. The aim of these
international studies [13]. Despite the relatively high inci- training units that are usually located in larger hospitals in
dence of above-foot amputations related to peripheral Sweden is to help amputees who had undergone rehabil-
arterial disease or diabetes in Sweden [6], resulting in itation with prosthesis to maintain their mobility level.
many prosthetic users, no valid and reliable measure of The training program is offered to amputees after the con-
lower limb amputees' physical function with the prosthe- clusion of routine prosthetic rehabilitation and participa-
sis has been available in Swedish. tion is voluntary.

The purpose of this study was to perform a cross-cultural The inclusion criteria for this study were age 40 years or
adaptation of the LCI to Swedish and evaluate the Swed- older, lower limb amputation up to trans-femoral level,
ish version for validity (convergent and discriminative) and that the amputee was fitted with a prosthesis. Data
and reliability in lower limb amputees attending training from all four rehabilitation units included gender, age,
after discharge from the hospital rehabilitation unit. and amputation level, and data for the amputees from

Page 2 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2009, 7:44 [Link]

Hässleholm-Kristianstad also included date of amputa- Two subscales emerge from this general construct; basic
tion and of receiving the prosthesis. One hundred and abilities (7 items) and advanced abilities (7 items). The
fifty five amputees fulfilled the inclusion criteria (67 from items inquire about the ability to perform activities and
Hässleholm-Kristianstad, 71 from Gothenburg, and 17 the level of independence while performing these activi-
from Stockholm), of whom 11 were excluded because of ties. Each of the 14 items is graded on a 4-point ordinal
incomplete data (Figure 1). The participants were repre- scale; 0 (not able to), 1 (yes, with help from other person),
sentative of the amputee population in Sweden with the 2 (yes, with supervision) and 3 (yes, independently). The
most common cause of amputation being peripheral arte- total LCI score is the sum of the item scores and can range
rial disease with or without diabetes and less commonly from 0 (worst) to 42 (best). Similarly, subscale scores for
infection or fracture [6]. basic and advanced capabilities with the prosthesis can
range from 0 to 21. The LCI is intended for self-adminis-
The data were collected from September 2003 through tration but can also be administered in a face-to-face or
December 2007. The study population consisted of 144 telephone interview. The time needed to complete the LCI
amputees; 55 women, mean age 75 (range 40–93) years, is approximately five minutes [11,12].
and 89 men, mean age 73 (range 44–91) years (Table 1).
EQ-5D
All participants from Hässleholm/Kristianstad were The EQ-5D is a measure of health-related quality of life
informed of the aim of the study and gave their written composed of 5 items covering 5 dimensions (mobility,
consent. Data from the other rehabilitation units con- self-care, usual activities, pain/discomfort, and anxiety/
tained no personal identifying information. The study was depression). Each item has 3 response levels: 1 (no prob-
approved by the Local Ethics Committee. lem), 2 (some problems), and 3 (unable to do for the first
3 items, or severe problems for the last 2 items). The pref-
Questionnaires and mobility test erence weights for the EQ-5D index have been generated
Locomotor Capabilities Index previously in the United Kingdom from a random general
The LCI consists of 14 items that measure one general population sample using the time trade-off method of
construct, the locomotor capabilities with the prosthesis. health evaluation [15]. The EQ-5D index ranges from 1.0

155 amputees attending after-rehabilitation training

Hässleholm/Kristianstad (HLM/KSD) 67
Gothenburg (GBG) 71
Stockholm (SHM) 17

11 excluded (incomplete data)

HLM/KSD 3, GBG 6, SHM 2

144 amputees
x Internal consistency
x Discriminative validity I (age groups)

21 non trans-
tibial amputees
excluded

123 trans-tibial amputees 40 amputees* 20 amputees*

Discriminative validity II Convergent validity I (TUG) x Convergent validity II (EQ-5D)
(unilateral vs. bilateral) x Test-retest reliability

* If an amputee completed the LCI on more than one occasion data from the first occasion were used
in the analysis of internal consistency and discriminative validity
Figure
Flow diagram
1 of the participants in the validity and reliability analyses
Flow diagram of the participants in the validity and reliability analyses.

Page 3 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2009, 7:44 [Link]

Table 1: Characteristic of the study population

Discriminative validity I Discriminative validity II Convergent validity I Convergent validity II

& &
Internal consistency Test-retest reliability

Number of amputees 144 123 40 20

Age, mean (range) yrs 74 (40–93) 74 (40–93) 74 (41–89) 76 (41–91)
Women, n (%) 55 (38) 50 (41) 15 (38) 10 (50)
Unilateral amputees, n (%)
TT 110 (76) 110 (89) 40 (100) 13 (65)
TF/KD 18 (13) 0 0 1 (5)
Bilateral amputees, n (%)
TT + TT/AD 14 (9.7) 13 (11) 5 (25)
TT + TF/KD 2 (1.3) 0 1 (5)
Time from prosthetic fitting to LCI testing, 38 (3–418)† 38 (3–418)† 20 (3–135) 91 (4–231)
mean (range) wks

†Data for 64 amputees from Hässleholm-Kristianstad (not available for amputees from the other 3 centers)
TF, trans-femoral; KD, knee disarticulation; TT, trans-tibial; AD, ankle disarticulation

(no problem with any of the 5 dimensions), to -0.594 has been proposed as a standard for correlation in validity
(extreme problems with all 5 dimensions). In this study studies [21].
the Swedish version of the EQ-5D was used. The EQ-5D is
widely used, has shown to be reliable and valid in the Discriminative validity was evaluated by comparing the
Swedish general population, and is easy to complete [16]. LCI scores among amputees in different age groups and in
unilateral and bilateral amputees. We hypothesized that
Timed "Up-and-Go" Test younger amputees would have better LCI scores than
In the Timed "Up-and-Go" (TUG) test the participant is older amputees and that unilateral trans-tibial amputees
asked to, as fast as possible, rise from a chair, walk three would have better scores than bilateral trans-tibial
meters with his/her ordinary walking aid, turn around, amputees. We also analyzed the LCI with regard to
walk back and sit down again in the chair and the result is whether the scores would differ significantly according to
measured in seconds. The TUG test is easy to use in clini- gender as one previous study reported better scores
cal settings and it has been shown to be valid and reliable among men than among women [22]. For comparison of
in testing of function in an elderly population [17]. LCI scores among amputees in different age groups and in
men and women data from all 144 amputees were ana-
Evaluation of validity lyzed with Kruskal-Wallis test and the Mann-Whitney test,
We examined the completeness of item responses, the dis- respectively. For comparing LCI scores in unilateral and
tribution of the scores, and the extent of ceiling and floor bilateral trans-tibial amputees, data from 123 amputees
effects in the results from all 144 participants. We assessed were analyzed with the Mann-Whitney test.
construct validity of the LCI by testing a number of prede-
fined hypotheses regarding its relationship with other Evaluation of reliability
measures of function and health (convergent validity) and Internal consistency
its ability to discriminate among groups expected to differ Internal consistency measures the homogeneity in a scale
in locomotor capabilities (discriminative validity) [18]. and the items should be at least moderately correlated
The number of participants included in the different anal- with each other. Internal consistency was determined
yses is shown in Figure 1. using Cronbach alpha coefficient and the 95% confidence
intervals (CI) were calculated using the bootstrap method.
Convergent validity [19] was determined by comparing Values between 0.70 and 0.95 have been proposed to
the LCI results with the TUG test and EQ-5D results in 2 indicate good internal consistency [21]. Internal consist-
subgroups of amputees. We hypothesized that better LCI ency reliability of the LCI was assessed using the responses
scores would have moderate or strong correlation (> 0.5) from all 144 participants.
[20] with better TUG values, and that the LCI would cor-
relate at least moderately with the EQ-5D index in a posi- Test-retest reliability
tive direction (i.e., better function with the prosthesis Test-retest reliability was evaluated in the same subgroup
would correlate with better EQ-5D index). The correlation of 20 amputees that provided data for the validity analysis
between the LCI scores and the TUG test results and the using the EQ-5D. The participants completed the LCI on
EQ-5D scores were calculated with the Spearman correla- two occasions with a mean interval of 11 (range 7–14)
tion coefficient (r). A correlation coefficient of at least 0.7 days. The test-retest LCI scores were analyzed with the

Page 4 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2009, 7:44 [Link]

intraclass correlation coefficient (ICC) using the two-way The mean total score was 28.5 (SD 12.5, median 33), the
random and absolute agreement criteria. The ICC (1,1) mean basic score was 17.1 (SD 5.5, median 21) and the
and 95% CI were calculated for the total LCI as well as for mean advanced score 11.3 (SD 7.8, median 12).
the basic and advanced subscales for the unilateral and
bilateral amputees. For the ICC (range 0 to 1) a value of Convergent validity
0.70 has been considered as acceptable reliability [19]. In the subgroup that performed the TUG test, the mean
LCI was 29.6 (range 2–42) and the mean TUG result was
Ceiling and floor effects 34.2 (range 9–92) seconds. The correlation between the
Reliability and validity of an instrument may be influ- LCI and the TUG was strong (r = -0.75, 95% CI -0.89–
enced by the presence of high ceiling and/or floor effects. 0.56, p < 0.001). The mean EQ-5D index was 0.63 (SD
A ceiling or floor effect is considered present if more than 0.3; range -0.18–1.0). The correlation between the LCI
15% of the respondents achieved the highest or lowest and EQ-5D index was strong (r = 0.84, 95% CI 0.58–0.95,
possible score [21,23]. p < 0.001).

Statistical analysis Discriminative validity

The LCI scores (total, basic and advanced) were calculated The mean LCI score for the amputees in younger age
and presented as means, medians and standard devia- groups was significantly better than that for amputees in
tions. All statistical tests were 2-sided and a p-value of older age groups (Table 3). The mean LCI score for unilat-
0.05 was considered to indicate statistical significance. eral amputees was 32.5 (SD 9.7) and for bilateral
Data were analyzed with SPSS version 14.0 (SPSS Inc., amputees was 14.9 (SD 12.5) (p < 0.001). The mean total
Chicago, USA) and STATA 10.0 (StataCorp, College Sta- score for women was 27.2 (SD 11.8, median 30) and for
tion, Texas, USA). men 29.2 (SD 12.9, median 33), the difference was not
statistically significant (p = 0.2), but LCI scores of 36 or
Results higher were more common among men than women (39
Score distribution men compared to 16 women) (Figure 2).
All the 144 participants answered all items. Basic item 1
"rising from a chair" and item 2 "walk indoors" had the Internal consistency
highest mean scores (2.9 and 2.7 respectively) and the Internal consistency was high; Cronbach alpha for the
worst scores were registered for advanced items 5 and 6 total LCI was 0.95 (95% CI 0.94–0.96), for basic activities
"getting up and down a stair without a handrail", both was 0.93 (95% CI 0.91–0.95), and for advanced activities
scoring a mean of 1.2 (Table 2). was 0.94 (95% CI 0.92–0.95). The item-total correlations
ranged from 0.42 (item basic 1) to 0.85 (item advanced 3)

Table 2: The locomotor capabilities index (LCI) item scores (n = 144)

% %
Items* Mean (SD) Median Highest score Lowest score

Basic activities
1. Get up from a chair 2.9 (0.5) 3 91 0.7
2. Walk indoors 2.7 (0.6) 3 82 1.4
3. Walk outside 2.3 (1.1) 3 63 13
4. Go up stairs, handrail 2.4 (0.9) 3 65 9.0
5. Go down stairs 2.4 (0.9) 3 65 9.0
6. Step up sidewalk curb 2.2 (1.1) 3 63 16
7. Step down sidewalk curb 2.2 (1.1) 3 63 16
LCI – Basic 17.1 (5.5) 21 54 0.7
Advanced activities
1. Pick up object from floor 2.0 (1.3) 3 63 26
2. Get up from floor 1.6 (1.3) 1 42 26
3. Walk outside 1.8 (1.3) 2 44 26
4. Outside in bad weather 1.6 (1.3) 2 42 35
5. Go up stairs without handrail 1.2 (1.2) 1 26 40
6. Go down stairs without handrail 1.2 (1.2) 1 26 40
7. Walk while carrying an object 1.7 (1.4) 3 51 38
LCI – Advanced 11.3 (7.8) 12 23 12
LCI – Total 28.5 (12.5) 33 23 0.7

*Each item has 4 score levels ranging from 0 (worst) to 3 (best); each LCI score (basic, advanced and total) is the sum of item scores

Page 5 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2009, 7:44 [Link]

Test-retest reliability man correlation coefficient 0.62) and with the Rivermead
In the whole test-retest sample, the ICC for the total LCI Mobility Index (Spearman coefficient 0.75), which is sim-
was 0.91, for basic LCI was 0.88, and for advanced LCI ilar to the correlation shown with the TUG test in our
was 0.92, and all 95% confidence intervals were above study, and assessment of reliability showed high internal
0.70 (Table 4). The mean difference in the LCI scores consistency (Cronbach Alpha 0.95 for the total LCI and
between the two testing times was -1.6 for the total LCI exceeding 0.90 for both subscales) and a high test-retest
and -0.8 for the basic and advanced LCI (all differences agreement (ICC = 0.80), which also is similar to the relia-
were statistically non-significant). Among the 14 unilat- bility results in our study. However, a very high Cronbach
eral amputees the ICC for the total LCI was 0.83, for the alpha may indicate possible item redundancy.
basic LCI was 0.66 and for the advanced was 0.92 (Table
4). For the 6 bilateral amputees the ICC for the total LCI In our study, the mean LCI score was 28.5 in a population
was 0.90, for the basic 0.93 and for the advanced 0.59 and of 144 amputees with a mean age of 74 years. In a report
the mean score difference between the two testing times from the developers of the LCI, a younger population
was -0.3, 0, and -0.3, respectively. (mean age 63 years) of 211 trans-tibial and 122 trans-fem-
oral amputees had a mean LCI of 31.6 and 29.2, respec-
Ceiling and floor effects tively [24]. Franchignoni et al. reported a mean LCI of 41
Of the 144 participants, 43 amputees had scores of 40 or at the end of a rehabilitation program among 50 unilat-
higher and 33 (23%) had a maximum possible score (ceil- eral amputees with a median age of 51 years [25]. When
ing effect). High scores were more common among men comparing the LCI results in different studies the charac-
than women. Only 1 amputee (0.7%) had a worst possi- teristics of the study populations should be taken into
ble score and 12 had scores below 8 (Table 2). consideration. The LCI items that scored highest in our
study were "getting up from a chair" and "walking
Discussion indoors" and the lowest score was found for the item con-
This study shows that the Swedish version of the LCI has cerning climbing and descending a stair without a hand-
good validity. The predefined validity hypotheses were rail, findings similar to those in previous studies [25,26].
confirmed with good ability to discriminate among
groups expected to differ in their locomotor capabilities The Swedish version could discriminate between unilat-
and high correlations between the LCI and the TUG test eral and bilateral amputees and between younger and
and between the LCI and the EQ-5D. The reliability tests older amputees regarding degree of independence in per-
showed good internal consistency and the test-retest reli- forming locomotor activities. These findings support
ability in a small subsample was acceptable. other studies that have demonstrated the usefulness of
LCI in detecting differences in mobility [24].
The measurement properties of the Swedish LCI are simi-
lar to those reported for the original English version. In our study men had ceiling LCI scores more often than
Assessment of validity of the original English version of women but there were no statistically significant differ-
the LCI [13] showed a significant correlation with the ences in the mean scores. In a study that analyzed predic-
Functional Independence Measurement (FIM) test (Spear- tors of good function after major lower limb amputation,
Hermodsson et al. found male sex to be a statistically sig-
nificant predictor and that men were three times more
likely than women to achieve good function [22].

In our study a high correlation of -0.75 between LCI and

TUG was found. A correlation of -0.64 was reported by
Miller et al. who studied 55 amputees [27]. The TUG test
is an objective test compared to the subjective nature of
the LCI. The TUG test also shows how the patient's safety
thinking works in a stressed situation; for example,
whether they take the time to lock the wheels on the walk-
ing frame. Falls are common among amputees and one
cause of decreasing function is fall injuries [28].

Our results showed a strong correlation between the LCI

Figure 2 of the LCI scores (N = 144)
Distribution and self-perceived health measured with the EQ-5D.
Distribution of the LCI scores (N = 144). Walking is a fundamental human ability and seems to be
strongly correlated to health. Using the Nottingham

Page 6 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2009, 7:44 [Link]

Table 3: The locomotor capabilities index (LCI) in different age amputees in our study [30]. One possible explanation is
groups (n = 144) that the amputees in our study were participating in an
Age group Number of amputees LCI after-rehabilitation training program.
yr n (%) mean (SD)*
The test-retest reliability of the total LCI, measured with
40–59 11 (7.6) 39 (4) the ICC in the whole test-retest sample, was comparable
60–69 29 (20.1) 32 (12) to that previously reported in other studies. Miller et al.
70–79 54 (37.5) 27 (12) [27] compared the LCI with two other self-report scales
80+ 50 (34.7) 25 (13)
among 55 unilateral amputees (72% below-knee and
28% above knee), and reported that the ICC for the LCI
* p = 0.001
was 0.88, for the Hougton scale was 0.85, and for the
Prosthetic Evaluation Questionnaire was 0.77. One limi-
Health Profile, Pell et al. found that amputees with vascu- tation in our study is the small test-retest sample size of 20
lar disease reported significantly greater problems with amputees in the evaluation of test-retest reliability.
mobility, social isolation, lethargy, pain, sleep and emo- Recently, a research group stated that "no criteria have
tional disturbance than controls, but mobility was the been defined for the required sample size of studies
only significant independent factor in a regression analy- assessing measurement properties" and considered "a
sis [29]. The authors stated "the overall quality of life of sample size of at least 50 patients adequate for the assess-
amputees is likely to be enhanced by concentrating reha- ment of the agreement parameter, based on a general
bilitation efforts on improving mobility". We believe that guideline by Altman [31]" and an ICC of 0.70 as mini-
wheelchair mobility should be part of the rehabilitation mum standard for reliability [21].
program for amputees and we consider amputees to be
"functional prosthetic users" even if they use a wheelchair Different standards for acceptable ICC values have been
for most of the day but are provided with prosthesis and proposed and a common recommendation is that meas-
are independent in transfers and can walk a few steps. ures intended for clinical use should have ICC exceeding
0.90 whereas for research purposes ICC of 0.70 has been
Despite the strong correlation with the LCI, the EQ-5D is considered acceptable [19]. Although the ICC values for
a measure of health-related quality of life and does not the whole test-retest sample in our study were close to
specifically measure an amputee's function with the pros- 0.90, the 95% confidence intervals were lower but still
thesis. In addition, the ability of the EQ-5D to detect above 0.70 even for the two subscales. The inclusion of
change in amputees' function over time has not been eval- bilateral amputees in the test-retest sample may be con-
uated. In a study that used the EQ-5D in amputees with sidered problematic because it may increase the variability
diabetes and foot ulcers in our region (including patients of the reliability coefficient and therefore may inflate the
similar to the amputees in our study), the authors found reliability [32]. In the subsample of unilateral amputees
that patients who had undergone major amputation had the ICC values were lower particularly for the basic LCI.
worse EQ-5D index than patients who achieved primary Although the ICC values for the unilateral amputees were
healing and those who had undergone minor amputa- near or above levels considered acceptable, they were
tion. The 26 amputees in that study had a mean EQ-5D based on a small sample size and subsequently had wide
index of 0.31 compared with 0.63 found among the confidence intervals. A study with a larger sample of uni-

Table 4: Test-retest reliability

Time 1 Time 2
LCI mean (SD) mean (SD) Mean difference ICC
(95% CI) (95% CI)

All amputees (n = 20)

Total LCI 24.5 (13) 26.1 (14) -1.6 (-4.3–1.1) 0.91 (0.79–0.96)
Basic 15.4 (7) 16.2 (7) -0.8 (-2.4–0.8) 0.88 (0.72–0.95)
Advanced 9.2 (7) 10.0 (7) -0.8 (-2.2–0.6) 0.92 (0.80–0.97)
Unilateral amputees (n = 14)
Total LCI 29.9 (11) 32.1 (9) -2.1 (-5.6–1.3) 0.83 (0.56–0.94)*
Basic 18.1 (5) 19.3 (4) -1.1 (-3.3–1.0) 0.66 (0.24–0.87)*
Advanced 11.8 (7) 12.8 (7) -1.0 (-2.5–0.5) 0.92 (0.78–0.97)

*The results are influenced by one outlier, for the other 13 amputees ICC basic is 0.78 (0.43–0.93) and total 0.89 (0.70–0.97).
ICC, intraclass correlation coefficient; CI, confidence interval

Page 7 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2009, 7:44 [Link]

lateral trans-tibial amputees would be needed to further limb amputations. The LCI could be a useful tool in this
assess test-retest reliability and to confirm that the test- context.
retest reliability is adequate for clinical use.
Conclusion
Ceiling effects with the use of LCI have been reported pre- The Swedish version of the LCI demonstrated good valid-
viously; one study reported a best possible score in 46% ity and internal consistency in adult amputees. Test-retest
of 50 amputees (mean age 51 years) [25], and another in reliability in a small subsample appears to be acceptable.
40% of 329 amputees (mean age 60 years) [27]. In our The ceiling effect was high, which may imply that it would
study, the same pattern was observed, despite the high age be most useful in assessing amputees with low to moder-
of the participants amputated because of peripheral arte- ate functional abilities.
rial disease. To address the problem of the ceiling effect
the LCI-5 has been designed, with item response 3 "yes, Competing interests
alone" changed to 3 "yes, alone with ambulation aids" The authors declare that they have no competing interests.
and 4 "yes, alone without ambulation aids" [25]. The high
ceiling effect may have contributed to the high value for Authors' contributions
internal consistency. BL, HIA, AJ and IA conceived of and designed the study.
BL, AJ and IA analyzed and interpreted the data. AJ, BL and
In clinical practice there is increasing need to evaluate the IA performed the statistical analysis. BL and AJ drafted the
methods used in rehabilitation because of a greater paper and IA critically revised it for important intellectual
emphasis on patient safety and a growing interest in content. All authors read and gave approval of the final
health economics. Whatever the purpose of their use, the manuscript.
tests must show not only high reliability and validity but
also be easy to use in a clinical setting. The LCI appears to Additional material
meet those requirements with its ease of use in daily prac-
tice.
Additional file 1
Swedish version of LCI. The translated version of the Locomotor Capa-
Amputees with a low level of function may not know bilities Index
whether or not they can perform the activities inquired Click here for file
about in the questionnaire. Elderly amputees may, for [[Link]
safety reasons, have stopped performing certain activities [Link]]
when they are alone, and some may always use their
wheelchair when outdoors. We have sometimes found it
valuable that the amputees are given the opportunity to
try to perform some of the activities in the presence of the Acknowledgements
test administrator. For instance, for the item "getting up This research was supported by Hässleholm Hospital. The authors thank
Anna Larsson, Sara Hedén, Kim McLearnon and Stephan Mc Learnon for
after falling" the amputee could be allowed to sit on a
their help in the translation procedures, the rehabilitation teams in Gothen-
mattress on the floor and try to rise up and stand with the
burg and Stockholm for help with collecting data, and Biostatisticians Jonas
help of a chair and "walking while carrying an object" Ranstam and Aleksandra Turkiewicz at the Swedish National Competence
could be exemplified as walking 10 meters while carrying Centre for Musculoskeletal Disorders, Department of Orthopedics, Lund
a glass of water. Due to the highly functional nature of the University Hospital, Lund, Sweden.
items in the LCI, these tests may be useful, for instance
when rehabilitation goals are defined together with the References
amputees [13]. However, care should be taken not to 1. Heikkinen M, Saarinen J, Suominen VP, Virkkunen J, Salenius J: Lower
limb amputations: differences between the genders and
change the focus of the test from being a self-administered long-term survival. Prosthet Orthot Int 2007, 31:277-286.
test to an observed test. Franchignoni et al. suggested 2. Kald A, Carlsson R, Nilsson E: Major amputation in a defined
guidelines for item scoring (e.g. carrying an object) as a population: incidence, mortality and results of treatment. Br
J Surg 1989, 76:308-310.
possible improvement of the LCI [25]. 3. Laaperi T, Pohjolainen T, Alaranta H, Karkkainen M: Lower-limb
amputations. Ann Chir Gynaecol 1993, 82:183-187.
4. Eneroth M, Persson BM: Amputation for occlusive arterial dis-
The methods used for lower limb amputation and reha- ease. A prospective multicentre study of 177 amputees. Int
bilitation following amputation may differ between units, Orthop 1992, 16:383-387.
even when treating similar patients, and the differences 5. Fletcher DD, Andrews KL, Butters MA, Jacobsen SJ, Rowland CM,
Hallett JW Jr: Rehabilitation of the geriatric vascular amputee
may involve chosen levels of amputation, edema treat- patient: a population-based study. Arch Phys Med Rehabil 2001,
ment, and time chosen for prosthetic fitting [2,6,22]. 82:776-779.
Finding the best practice would require a standardized 6. Johannesson A, Larsson GU, Oberg T: From major amputation
to prosthetic outcome: a prospective study of 190 patients in
protocol for measuring rehabilitation progress after lower a defined population. Prosthet Orthot Int 2004, 28:9-21.

Page 8 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2009, 7:44 [Link]

7. Davies B, Datta D: Mobility outcome following unilateral lower

limb amputation. Prosthet Orthot Int 2003, 27:186-190.
8. Gallagher P, MacLachlan M: Psychological adjustment and cop-
ing in adults with prosthetic limbs. Behav Med 1999, 25:117-124.
9. Tisi PV, Callam MJ: Type of incision for below knee amputation.
Cochrane Database Syst Rev 2004:CD003749.
10. Back-Pettersson S, Bjorkelund C: Care of elderly lower limb
amputees, as described in medical and nursing records.
Scand J Caring Sci 2005, 19:337-343.
11. Grise MC, Gauthier-Gagnon C, Martineau GG: Prosthetic profile
of people with lower extremity amputation: conception and
design of a follow-up questionnaire. Arch Phys Med Rehabil 1993,
74:862-870.
12. Gauthier-Gagnon C, Grise MC: Prosthetic profile of the
amputee questionnaire: validity and reliability. Arch Phys Med
Rehabil 1994, 75:1309-1314.
13. Gauthier-Gagnon C, Grise MC: Tools to measure outcome of
people with lower limb amputation; Update on the PPA and
LCI. J Prosthet Orthot 2006, 18:61-69.
14. Beaton DE, Bombardier C, Guillemin F, Ferraz MB: Guidelines for
the process of cross-cultural adaptation of self-report meas-
ures. Spine 2000, 25:3186-3191.
15. Dolan P, Roberts J: Modelling valuations for Eq-5d health
states: an alternative model using differences in valuations.
Med Care 2002, 40:442-446.
16. Burstrom K, Johannesson M, Diderichsen F: Swedish population
health-related quality of life results using the EQ-5D. Qual Life
Res 2001, 10:621-635.
17. Podsiadlo D, Richardson S: The timed "Up & Go": a test of basic
functional mobility for frail elderly persons. J Am Geriatr Soc
1991, 39:142-148.
18. Kirshner B, Guyatt G: A methodological framework for assess-
ing health indices. J Chronic Dis 1985, 38:27-36.
19. Streiner DL, Norman GR: Health Measurement Scales. A practical guide
to their development and use 3rd edition. Oxford (NY): Oxford Univer-
sity Press; 2003.
20. Cohen J: Statistical power analysis for the behavioral sciences 2nd edition.
Hillsdale, NJ: Erlbaum Associates; 1988.
21. Terwee CB, Bot SD, de Boer MR, Windt DA van der, Knol DL,
Dekker J, Bouter LM, de Vet HC: Quality criteria were proposed
for measurement properties of health status questionnaires.
J Clin Epidemiol 2007, 60:34-42.
22. Hermodsson Y, Ekdahl C, Persson BM: Outcome after trans-tibial
amputation for vascular disease. A follow-up after eight
years. Scand J Caring Sci 1998, 12:73-80.
23. de Groot IB, Favejee MM, Reijman M, Verhaar JA, Terwee CB: The
Dutch version of the Knee Injury and Osteoarthritis Out-
come Score: a validation study. Health Qual Life Outcomes 2008,
6:16.
24. Gauthier-Gagnon C, Grise MC, Potvin D: Enabling factors related
to prosthetic use by people with transtibial and transfemoral
amputation. Arch Phys Med Rehabil 1999, 80:706-713.
25. Franchignoni F, Orlandini D, Ferriero G, Moscato TA: Reliability,
validity, and responsiveness of the locomotor capabilities
index in adults with lower-limb amputation undergoing pros-
thetic training. Arch Phys Med Rehabil 2004, 85:743-748.
26. Franchignoni F, Tesio L, Orlandini D: Mobility scales for lower
limb-prosthetic patient: the locomotor capabilities index.
Arch Phys Med Rehabil 2002, 83:582-583.
27. Miller WC, Deathe AB, Speechley M: Lower extremity prosthetic
mobility: a comparison of 3 self-report scales. Arch Phys Med Publish with Bio Med Central and every
Rehabil 2001, 82:1432-1440.
28. Pauley T, Devlin M, Heslin K: Falls sustained during inpatient scientist can read your work free of charge
rehabilitation after lower limb amputation: prevalence and "BioMed Central will be the most significant development for
predictors. Am J Phys Med Rehabil 2006, 85:521-532. disseminating the results of biomedical researc h in our lifetime."
29. Pell JP, Donnan PT, Fowkes FG, Ruckley CV: Quality of life follow-
ing lower limb amputation for peripheral arterial disease. Sir Paul Nurse, Cancer Research UK
Eur J Vasc Surg 1993, 7:448-451. Your research papers will be:
30. Ragnarson TG, Apelqvist J: Health-related quality of life in
patients with diabetes mellitus and foot ulcers. J Diabetes Com- available free of charge to the entire biomedical community
plications 2000, 14:235-241. peer reviewed and published immediately upon acceptance
31. Altman DG: Practical statistics for medical research London: Chapman
and Hall; 1991. cited in PubMed and archived on PubMed Central
32. de Vet HC, Terwee CB, Knol DL, Bouter LM: When to use agree- yours — you keep the copyright
ment versus reliability measures. J Clin Epidemiol 2006,
59:1033-1039. Submit your manuscript here: BioMedcentral
[Link]

Page 9 of 9
(page number not for citation purposes)
1432

Lower Extremity Prosthetic Mobility: A Comparison

of 3 Self-Report Scales
William C. Miller, PhD, MSc, A. Barry Deathe, MD, Mark Speechley, PhD
ABSTRACT. Miller WC, Deathe AB, Speechley M. Lower cine and the American Academy of Physical Medicine and
extremity prosthetic mobility: a comparison of 3 self-report Rehabilitation
scales. Arch Phys Med Rehabil 2001;82:1432-40.
EHABILITATION PRACTITIONERS and researchers
Objective: To assess and compare the reliability and validity
of the Houghton Scale, the Prosthetic Profile of the Amputee R need measures that can distinguish among levels of dis-
ability, predict prognosis, help plan patient care, and indicate
Locomotor Capabilities Index (PPA-LCI), and the Prosthetic
Evaluation Questionnaire (PEQ) mobility subscale, 3 disease- change in functional status occurring as a result of intervention.
specific self-report measures of functional mobility for lower The demand for accountability comes not only from funding
extremity prosthetic mobility. sources but also from the recipients of treatment. Many factors
Design: Four-week test-retest: 1 sample for reliability anal- complicate the issue of measurement in rehabilitation. Measur-
yses, 1 sample for validity analyses. ing disability, for instance, is prone to wide disagreement about
Setting: University-affiliated outpatient amputee clinic, in what should be measured and what is the best method to
Ontario, Canada. capture the information.1-3 In addition, there is the tendency
Participants: Two outpatient amputee samples (sample 1 among researchers to create new measures rather than to refine
[n ⫽ 55], for reliability analysis; sample 2 [n ⫽ 329], for existing instruments.4 Fifteen years ago, Keith2 lamented the
validity analysis). state of measurement and evaluation in rehabilitation, which
Interventions: Not applicable. continues today,3 despite the fact that assessment of function
Main Outcome Measures: Test-retest of reliability and has become more sophisticated.
convergent validity of the 3 scales. Convergent validity and Measuring the outcomes of rehabilitation of persons who
discriminative ability were also assessed after setting a priori have had a lower extremity amputation is a good example of
hypotheses for 2 scales of walking performance, balance con- the measurement dilemma faced in other areas of rehabilita-
fidence, and other indicators of ambulatory ability. tion. Successful outcome among amputees is multifaceted and
Results: The reliability of the PPA-LCI (intraclass correla- can vary, depending on the goals of the individual. Although no
tion coefficient [ICC] ⫽ .88) was slightly higher than the consensus exists on which measures should be used, most
Houghton Scale (ICC ⫽ .85) and the PEQ mobility subscale investigators advocate that a variety of outcomes relating to
(ICC ⫽ .77). The PPA-LCI was prone to high ceiling effects quality of life, including measurement of functional limitation,
(40%) that would limit its ability to detect improvement. Evi- should be considered.5
dence for convergent validity, when compared with the Although community reintegration and generally improved
2-Minute Walk Test, Timed Up and Go, and the Activity- mobility are the ultimate goals of amputee rehabilitation pro-
Specific Balance Confidence Scale, was supported as hypoth- grams, prosthetic mobility is often among the primary objec-
esized in all the scales. Each of the scales was able to discrim- tives of such programs. Various approaches have been used to
inate between different groups for amputation cause, walking assess prosthetic mobility. They include performance tests,6,7
distance, mobility device use, and automatism, with each hav- categories of ambulation,8-11 self-reported capability, and pros-
ing varying strength related to relative precision. The Hough- thetic use over a variety of ambulation activities.12-17 The
ton Scale was the only scale able to distinguish between am- optimal approach to assessment would be to use a combination
putation levels. of objective and subjective assessments to capture the many
Conclusions: Reliability and validity of all the scales are dimensions and qualities that contribute to a rehabilitation
acceptable for group level comparison. None of the scales had outcome. However, when faced with a busy clinical practice,
clearly superior psychometric properties compared with the using a battery of assessments to capture all the various dimen-
others. Further research is required to assess responsiveness. sions is often not feasible. There are strengths and limitations
Key Words: Amputees; Artificial limbs; Leg; Psychomet- associated with each approach. One advantage of self-report
rics; Rehabilitation; Reproducibility of results. measures appealing to both research and clinical applications,
© 2001 by the American Congress of Rehabilitation Medi- especially in clinics responsible for large geographic regions, is
that assessment of prosthetic mobility can occur by mail or
telephone. One criticism of outcome measures for the amputee
population is that many instruments have no known published
From the Faculty of Medicine, School of Rehabilitation Sciences, University of
psychometric properties.
British Columbia, Vancouver, BC (Miller); and Departments of Physical Medicine Three self-report scales of prosthetic mobility for the ampu-
and Rehabilitation (Deathe) and Epidemiology and Biostatistics (Speechley), Univer- tee population have recently received attention in the literature.
sity of Western Ontario, London, Ont, Canada. These include the Houghton Scale, the Prosthesis Evaluation
Accepted in revised form November 20, 2000.
No commercial party having a direct financial interest in the results of the research
Questionnaire (PEQ), and the Prosthetic Profile of the Amputee
supporting this article has or will confer a benefit upon the author(s) or upon any Locomotor Capabilities Index (PPA-LCI). The authors of the
organization with which the author(s) is/are associated. PPA-LCI and the PEQ have reported psychometric proper-
Reprint requests to William Miller, PhD, Faculty of Medicine, Schl of Rehabili- ties.12,17 The Houghton Scale, however, has face validity but no
tation Sciences, T325-2211 Wesbrook Mall, University of British Columbia, Van-
couver, BC V6T 2B5, Canada, e-mail: bcmiller@[Link].
known published psychometric information is available.12,17
0003-9993/01/8210-6257$35.00/0 Comparative psychometric information makes the task of se-
doi:10.1053/apmr.2001.25987 lecting a scale easier. However, interpreting psychometric in-

Arch Phys Med Rehabil Vol 82, October 2001

 COMPARISON OF 3 PROSTHETIC MOBILITY SCALES, Miller 1433

formation can be difficult because of such issues as different for the study was 77% of the total eligible clinic population.
inclusion-exclusion criteria for selecting testing samples. Fur- The subjects from sample 2 had been discharged from pros-
thermore independent replication of existing measures in- thetic rehabilitation for at least 12 months. Subjects present in
creases clinicians’ confidence in the soundness of subjective both data sets (n ⫽ 36) were excluded from analyses of
measures.4 sample 2.
The objectives of the present study were to provide the first
published estimates of the psychometric properties for the Procedure
Houghton Scale, and to conduct an independent study and
After giving consent, sample 1 subjects completed the self-
replication of the reliability and validity of the PPA-LCI and
report questionnaires in the waiting room of the outpatient
the PEQ mobility subscale. We also considered the compre-
clinic. On completion of their regularly scheduled follow-up
hensiveness, credibility, and use of the 3 scales. Independent
appointment with the amputee rehabilitation team, the team
head-to-head comparisons, such as those made in the present
physiatrist clinically assessed subjects’ medical and prosthetic
study, provide clinicians with additional evidence of a scale’s
stability. If a subject was deemed stable, he/she then completed
value when they select an instrument for clinical or research
the 2-Minute Walk and the Timed Up and Go (TUG) Tests in
applications.
a quiet, isolated hallway adjacent to the outpatient clinic area.
METHODS Both tests were conducted in accordance with the standardized
protocol for these scales. Four weeks later, each subject re-
ceived a package containing the self-report questionnaires and
Subjects
a self-addressed, stamped return envelope. Two questions were
The present study used data from 2 different samples. Data appended to the questionnaire to inquire about changes in
from the first sample were collected primarily to assess the medical and prosthetic stability over the test-retest interval.
reliability of the 3 self-report scales, whereas data from the The information collected from sample 2 was used to aug-
second sample were used to assess validity. Both samples were ment the assessment of validity of the 3 self-report scales of
assembled by means of a consecutive sampling strategy from interest. For this sample, subjects were sent a letter introducing
the St Mary’s Hospital Regional Amputee Program, a univer- the study along with an accompanying questionnaire. Data
sity-affiliated outpatient amputee clinic that serves the region collection was closed after all potential respondents received 3
of southwestern Ontario, Canada. The clinic has a catchment of written reminders (1 of which included a second questionnaire)
approximately 1.4 million people and covers a large rural and and a final follow-up phone call to request participation. The
urban area. At this clinic all individuals with a lower extremity protocol was approved by the university’s Review Board for
amputation who are expected to benefit from a prosthetic limb Health Sciences Research.
are offered or provided inpatient prosthetic rehabilitation and
then followed-up by the amputee rehabilitation team at 3, 6, Measurement
and 12 months after discharge from rehabilitation and yearly
thereafter. Individuals considered eligible for either sample Mobility scales. The Houghton Scale is a 6-item measure
included all persons who had either a unilateral below knee of prosthetic use performance that has been used as an outcome
(BK) or above knee (AK) amputation for any reason and in at least 3 published studies over the past 12 years.14-16 This
successfully completed their inpatient prosthetic training at measure assesses the amount of time the prosthesis is used, the
least 6 months before the start of the study. Only individuals manner in which it is used, whether a mobility device is used
who were at or over the age of 20 years and could understand when ambulating outside, and the perception of stability when
English were included. Further inclusion criteria are discussed walking over a variety of terrain. The measure provides a
in the sample descriptions later. 4-point ordinal response option for the first 3 questions whereas
the last 3 response options are binary (yes/no). The responses
Sample 1 are summed to provide a score ranging from 0 (poor perfor-
mance) to 12 (maximum performance). Scores above 8 indicate
For sample 1, we recruited 60 subjects from the amputee a successful prosthetic ambulator.14 Although the Houghton
outpatient clinic between May and September 1999. A total of Scale has face validity, we found no published information
76 individuals were potentially eligible, however, 10 declined regarding its development or psychometric properties.
to participate and 6 were deemed ineligible by the attending The LCI is an 11-item subscale of the larger PPA.12,13 The
physician because they had medical or prosthetic problems that PPA-LCI items cover a range of transfer and ambulation ac-
could cause functional instability. Because the primary goal of tivities, such as getting up from a chair or the floor, climbing
sample 1 was to examine test-retest reliability, we attempted to stairs and curbs, and walking on a variety of surfaces. Scores
minimize any change in mobility status over the retest interval. range from 0 to 33 based on a 4-point Likert scale, which
Therefore, only those subjects who were considered prosthet- assesses whether the subject is: able to perform both basic and
ically and medically stable were included. The final sample of advanced activities “by myself,” requires standby or hands on
55 persons (92%) who provided us with full information for assistance, or is completely “unable” to perform the activity.
test-retest analyses exceeds the minimal number (42) of sub- The summary score is derived by adding up the numeric value
jects required to determine if a suitably strong reliability coef- recorded for each item. High scores indicate more capability.
ficient ( p ⫽ .60 –.85) exists18 when alpha is .05 and beta is .20. The internal consistency (Cronbach’s ␣ ⫽ .95), 4-week test-
We oversampled to compensate for an anticipated 75% re- retest reliability (intraclass correlation coefficient [ICC] ⫽ 0.8),
sponse rate. and construct validity have been previously reported to be
strong.12,13
Sample 2 The PEQ is composed of 9 subscales assessing emotional
Data from a second sample of 329 subjects, recruited as part and social health, overall well-being, prosthetic function, and
of a separate follow-up survey conducted in the fall of 1998 of mobility over the past 4 weeks. For the purpose of the present
all patients from the same outpatient clinic, were also included study, we used only those items that addressed mobility. A total
in the present study.19 The response rate for subjects eligible of 9 items address ambulation and 4 items address transfer

Arch Phys Med Rehabil Vol 82, October 2001

1434 COMPARISON OF 3 PROSTHETIC MOBILITY SCALES, Miller

activities while using the prosthesis. To make the content more possible. The coefficient of variation (CV) was calculated by
similar to the other measures, we combined the 2 subscales. dividing the SD by the mean and multiplying by 100. The CV
The 13 items cover similar areas to those of the PPA-LCI, is a unit-free measure of the amount of relative variation that
including walking up and down stairs and hills, sitting down permits a comparison of different scales.28
and getting up from a chair or toilet, and walking on various Reliability was determined by examining internal consis-
surfaces. The internal consistency (Cronbach’s ␣ ⫽ .89 for tency and test-retest values for sample 1. We calculated Cron-
ambulation, ⫽ .47 for transfers), 4-week test-retest reliability bach’s alpha for each scale to assess internal consistency. Item
(ICC ⫽ .90, .73, respectively), and adequate construct validity consistency scaling was assessed by observing the change in
has been reported.17 The original scale uses a 100-mm visual the Cronbach’s alpha of the scale in the presence and absence
analog scale (VAS) anchored at one end by “cannot do it” and of each item. ICCs were derived for the summary scores by
at the other end by “no problem.” For the purposes of the
using 1-way analysis of variance (ANOVA) to assess test-retest
present study, the response scale was altered to a numeric scale
ranging from 0 to 10. Our decision to alter the scale were based reliability. Additionally, we calculated test-retest of each item
on pilot study feedback from a small group (n ⫽ 8) of volun- by using ICC for the PEQ mobility subscale. We used Ken-
teers from the inpatient prosthetic rehabilitation program who dall’s index to determine agreement for the ordinal level items
indicated that the VAS format was unfamiliar and not easily of the PPA-LCI and the first 3 items of the Houghton Scale,
understood. Similar observations have been reported by Guyatt and we used Cohen’s kappa to assess item reliability for the last
et al,20 who found subjects had fewer problems and required 3 items of the Houghton because of their dichotomous nature.
less training, by using a numerically based Likert scale as The data from both samples were used to assess criterion
opposed to a VAS. The mean is calculated to provide a sum- validity. Pearson’s product-moment correlation (r) was used to
mary score. measure associations between scales and other variables. The a
Instruments to assess convergent validity. To assess con- priori hypotheses posed for convergent validity included sta-
vergent validity of the 3 self-report mobility scales, we com- tistically significant and strong positive correlation for each
pared each with the 2-Minute Walk Test, the TUG, and the self-report measure with the 2-Minute Walk Test and the ABC
Activities-Specific Balance Confidence (ABC) Scale. The Scale, and a statistically significant strong negative correlation
2-Minute Walk Test is a performance measure of the distance with the TUG.
an individual is able to walk at his/her “usual” pace. This By using our selections on clinical experience and results
measure, which is used at several local clinics to determine the from previous studies, we set hypotheses to determine if each
progress of prosthetic performance, is a shortened version of primary scale could distinguish between the 6 clinically rele-
the original 12-Minute Walk Test.21 To provide clinicians with vant factors. We expected the scales to distinguish between BK
a test that took less time to complete, 6- and 12-minute versions and AK, and between vascular and other amputees. Three
were developed and tested.22 The 2-Minute Walk Test was indicators of mobility, the reliance on a walking device, inabil-
found to correlate highly with the 12-minute (r ⫽ .86) and ity to walk more than 1 block, and the need to concentrate on
6-minute (r ⫽ .89) versions.22 stepping while walking, were all expected to show a statistical
The TUG is a performance measure of many of the compo- ( p ⬍ .05) difference. Because these factors are dichotomous,
nents of basic mobility including balance, transfers, walking, we used independent t tests to compare mean levels of the 3
and turning while walking.23,24 The time it takes for an indi- scales.
vidual to stand from a sitting position, walk a 3-meter distance, Effect size was calculated by dividing the mean difference
turn, walk back to the chair, and sit down is recorded. Addi- between groups by the pooled SD to provide an estimate of the
tionally, the observer assesses the subject’s balance when the precision of each measure. For the remaining factor, the esti-
subject stands, turns, and then sits down again. The TUG was mated number of hours the prosthesis was worn each week, the
found to have excellent test-retest (r ⫽ .93) and interrater (r ⫽ variables were grouped based on the quartile distribution. We
.96) reliability and evidence of convergent or divergent validity used ANOVA to estimate the mean between group differences
among a sample of lower extremity amputees.25 for each scale, with a higher F statistic indicating a greater
The ABC Scale is a 16-item self-report measure of the amount of information relative to the error variance. The pres-
perceived balance confidence an individual has while complet- ence of a larger F value suggests a greater amount of informa-
ing specific activities such as walking around the house, as- tion relative to the error variance. The ratio of F statistics, by
cending and descending stairs, getting in and out of the car, and using the largest F value as the numerator, was computed to
bending over to pick up a slipper from the floor.26 The internal indicate the proportional efficiency of the instruments.29
consistency (␣ ⫽ .96), 2-week test-retest reliability (r ⫽ .92), To determine if 1 of the scales was clearly superior, we
and convergent and discriminant validity of this measure are calculated a summary score based on the unweighted rank-
strong among the elderly.26,27 order sum of components considered important within each of
Variables to assess discriminant validity. We identified the descriptive, reliability, and validity statistics. The descrip-
several factors to determine the ability of each scale to distin- tive components included the values for the CV and ceiling
guish between clinically important groups. These factors in- effect. The test-retest and internal consistency statistics were
cluded: (1) amputation level, (2) amputation cause, (3) pros- summed within the reliability dimension, and all the tests
thetic wearing time, (4) mobility device use (cane, walker, except those differentiating mobility device use and wearing
wheelchair), (5) reported walking distance, and (6) the ability time were considered within the validity dimension. Mobility
to walk without having to think about each step. device use and wearing time were excluded because they are
items included on the Houghton Scale. Scales were given a
Analyses score of 3 if they performed the best within a particular dimen-
Descriptive values for score distributions such as the mean, sion, 2 and then 1 for the scale with the least strength within the
median, standard deviation (SD), and the minimum and max- component. In the event of a tie between scales, the awarded
imum values for each measure were computed for each scale. scores were summed and then divided by the number of tests
We also calculated the ceiling and floor effects representing the involved in the tie. All data management and analyses were
percentage of subjects achieving the highest and lowest score calculated by using SPSS, version 8 for Windows.a

Arch Phys Med Rehabil Vol 82, October 2001

 COMPARISON OF 3 PROSTHETIC MOBILITY SCALES, Miller 1435

Table 1: Sample Demographics and Amputation-Related measure. The CV was relatively comparable across scales;
and Mobility Characteristics however, the Houghton Scale appeared to present the most
Sample 1 Sample 2 consistent value.
(n ⫽ 60) (n ⫽ 329)
Reliability and Internal Consistency
Sociodemographic Characteristics
Age (yr) 58.4 ⫾ 15.5 59.9 ⫾ 16.7 Estimates of reliability (table 3) showed that internal con-
Male (%) 73 74 sistency of the first administration scores was highest for the
Amputation Characteristics PEQ mobility subscale, followed by the PPA-LCI. The internal
Cause (%) consistency was lowest for the Houghton Scale (Cronbach’s
Vascular 55 53 ␣ ⫽ .68) and highest for the PEQ mobility subscale (Cron-
Nonvascular 45 47 bach’s ␣ ⫽ .95). Scaling procedures revealed that the alphas
Level (%) for each scale did not improve with the exclusion of any
BK 72 74 individual item.
AK 28 26 All test-retest reliability estimates exceeded 0.7, which is the
Years since amputation 7.1 ⫾ 3.4 16.0 ⫾ 15.4 suggested minimum standard for group comparisons.30 Total
Wearing time (hr/wk) 82.8 ⫾ 35.0 86.6 ⫾ 31.8 score test-retest reliability was strongest for the PPA-LCI,
Mobility Characteristics followed by the Houghton Scale; however, no scale was clearly
2-Minute Walk (m) 99.2 ⫾ 15.1 superior, as shown by the overlapping confidence intervals. A
TUG (s) 19.3 ⫾ 15.1 broad range of values was observed for item test-retest (table
3). Intraitem values greater than .75 are considered accept-
NOTE. Values are mean ⫾ SD. able.30 Only 1 item from the PPA-LCI achieved this standard in
the present study.

RESULTS Convergent Validity

Table 4 contains the Pearson’s product-moment correlation
Descriptive Properties among the 3 scales of interest and the convergent variables.
Comparing samples 1 and 2. The descriptive statistics for Each of the scales was related to the convergent variables in the
the sociodemographic and amputation-related characteristics anticipated direction. All 3 scales correlated at the r ⱖ 0.5 level
for both samples are shown in table 1. The samples were with the 2-Minute Walk Test, TUG, and the ABC Scale.
similar in age, gender, and amputation cause and level charac- Intercorrelation among the 3 prosthetic mobility scales ex-
teristics. Sample 2 had a mean number of years since amputa- ceeded the r ⫽ 0.5 level with the strongest coefficients being
tion of 2.3 times longer than sample 1 and also had a slightly observed between the PEQ mobility subscale and PPA-LCI.
longer mean prosthesis wearing time. Table 2 contains the
descriptive properties of the 3 self-report scales for the 2 Discriminant Validity
samples. Although minimal difference was found for the mean All 3 scales distinguished between the different groups with
score between the samples for the Houghton Scale, we found 1 exception. The Houghton Scale was the only instrument to
greater variation in the mean scores for the PEQ mobility differentiate statistically ( p ⬍ .05) between AK and BK am-
subscale and PPA-LCI between samples, as shown by the CV. putees, with a small effect size of .29, whereas the other scales
Despite slight variation in the scores, the results from tables 1 failed to detect a difference between amputation level (table 5).
and 2 suggest that sample 1 was representative of the larger Otherwise, the PEQ mobility subscale had better precision for
clinic sample. automatic walking and walking distance, whereas the PPA-LCI
Comparing the 3 scales. The score distribution, disper- was better for amputation cause and mobility device use. We
sion, and measures of central tendency are shown in table 2. All had expected the large effect size that we found for the Hough-
the scales had a nonsymmetric distribution and were skewed to ton Scale when distinguishing between individuals who used a
the left as seen by observing the floor and ceiling effects. The mobility device and those who did not because 1 of the Hough-
PPA-LCI had a large ceiling effect because more than one third ton items inquires about mobility device when ambulating
of the subjects obtained the highest score possible on the outside. Amputation level aside, the effect size values for the

Table 2: Descriptive Statistics and Features of Score Distributions for the Houghton Scale, PEQ Mobility Subscale, and PPA-LCI

Houghton Scale PEQ Mobility Subscale PPA-LCI

(6 items) (12 items) (11 items)
Sample Sample Sample
1 2 1 2 1 2

Mean 8.7 8.8 7.8 7.2 28.5 26.9

Median 9.0 9.0 8.5 8.1 31 31
Observed (min/max) 3–12 0–12 1–10 0–10 1–33 0–33
SD 2.3 2.0 2.2 2.6 6.1 8.2
CV (%)* 26.4 22.7 28.2 36.1 21.4 30.5
% Floor 0 0.3 0.6 0.3 1.7 0.3
% Ceiling 12.9 6.0 8.1 10.0 36.7 40.4

NOTE. Sample 1: n ⫽ 60; sample 2: n ⫽ 329.

Abbreviations: min, minimum; max, maximum.
* CV ⫽ (SD/Mean) ⫻ 100%.

Arch Phys Med Rehabil Vol 82, October 2001

1436 COMPARISON OF 3 PROSTHETIC MOBILITY SCALES, Miller

Table 3: Reliability Estimates for the Houghton Scale, PEQ Mobility Subscale, and the PPA-LCI (n ⴝ 55)

Houghton PEQ Mobility Subscale PPA-LCI

Internal consistency .68 .95 .89

Total score test-retest .85 .77 .88
95% CI .74–.90 .62–.85 .81–.93
SEM* .93 1.10 2.70
Item test-retest† .15–.72 .37–.73 .44–.77

* Standard error of measurement (SEM) ⫽ SD ⫻ 公(1 ⫺ reliability).

†
Chance-corrected kappa used for the Houghton Scale and PPA-LCI ICC used for the PEQ mobility subscale.

PEQ mobility subscale and PPA-LCI were all larger than 0.8, and PEQ mobility subscale were developed by using a frame-
suggesting a good ability to detect differences given similar work to inform item selection.17,32 In addition to selecting items
sample sizes.31 from the published literature, clinicians, researchers, and pa-
With regard to prosthetic wearing time, all 3 scales had tients participated in item generation,30,32,34-36 and statistical
statistically different F values between groups (table 6). The data reduction techniques were used to derive the final tool.13,17
Houghton Scale had the largest F value, which suggested that The techniques used in the construction of the Houghton Scale
it provided more information relative to the error variance.29 are not clear, because no published information about its de-
The PEQ mobility subscale was the only scale that found velopment is available. Its content appears to consider indica-
statistically significant differences between each of the groups. tors of prosthetic mobility that are more universal than the
Table 7 provides a summary of the overall comparison of the other 2 scales. Evidence of this is seen in the Houghton Scale’s
scales. The best overall score was recorded for the PPA-LCI Cronbach alpha, which has considerably less internal consis-
(23.5), followed closely by the PEQ mobility subscale. The tency than the other 2 scales. The PPA-LCI and PEQ mobility
scores were similar when we made comparisons across the subscale both had alphas greater than 0.9, which is considered
descriptive statistics, but the PPA-LCI had slightly larger excellent.37 Feinstein,38 however, believes that internal consis-
scores when we considered reliability and validity statistics. tency is not necessary for clinimetric tools because items are
often deliberately selected to contribute distinctive information
DISCUSSION regardless of the influence of homogeneity on the tool.
Mobility is the most obvious functional limitation for indi- Bombardier and Tugwell32 suggested that the contextual
viduals with lower limb amputations, yet the measurement of framework of the wording of the questions is important to its
mobility remains relatively underdeveloped for this population. credibility. The Houghton Scale appears to measure perfor-
Although several different formats of mobility outcome tools mance or “did do” issues, as opposed to the other 2 scales,
are available, few have been standardized or developed and which evaluate capability or “could do” issues. This may
tested with the amputee population in mind. To be useful, a explain why the former was able to discriminate between BK
measure should be easy to use, comprehensive, credible, reli- and AK amputees (table 5). Although those who have an AK
able, valid, and responsive.32,33 amputation believe that they can perform tasks such as walking
on uneven ground they may choose not to because of the
Credibility and Comprehensiveness increased energy demands or decreased balance confidence.
The goal of a disease-specific measure is to address issues Some debate exists about whether capability or performance
that are highly relevant to the population of interest. Credibility is more meaningful,32,38 but both are likely important to con-
and comprehensiveness are intimately linked to the test devel- sider,39 and selection should be based on the goal of the
opment—specifically, item selection and inclusion. Knowing clinician or researcher. If the goal is to measure change in
how the items were selected and understanding whether the disability related to treatment, then capability would be fa-
items are complete enough to capture the construct being vored. Performance measures, on the other hand, are a better
described is helpful when selecting a scale.32 Starting with a assessment of what the person does in his/her environment.
conceptual foundation that can guide item selection is essential Research suggests that estimates of capability are 20% higher
to ensure that all relevant dimensions of the construct are than performance among children39; however, given the limited
included. This necessary step is occasionally overlooked by amount of research on this subject among the amputee popu-
researchers in the area of rehabilitation.3 Both the PPA-LCI lation, further study is required to determine if differences

Table 4: Convergent Validity of the Houghton, PEQ Mobility Subscale, and PPA-LCI

Houghton Scale PEQ Mobility Subscale PPA-LCI

Sample Sample Sample
1 2 1 2 1 2

2-Minute Walk Test .64 .50 .64

TUG ⫺.60 ⫺.50 ⫺.64
ABC Scale .67 .63 .82 .85 .82 .81
PPA-LCI .60 .59 .77 .83
PEQ mobility subscale .59 .55

NOTE. All Pearson’s product-moment correlation coefficients are significant (p ⱕ .01). Blanks with 2-Minute Walk Test and TUG are because
no measurement was available. Sample 1: n ⫽ 60; sample 2: n ⫽ 329.

Arch Phys Med Rehabil Vol 82, October 2001

 COMPARISON OF 3 PROSTHETIC MOBILITY SCALES, Miller 1437

Table 5: Discriminant Validity of the Houghton Scale, PEQ Mobility Subscale, and PPA-LCI for All Groups (n ⴝ 329)

Houghton PEQ Mobility Subscale PPA-LCI

n Mean ES* Mean ES* Mean ES*

Amputation level
BK 243 8.9 .29 7.3 .11† 26.1 .13†
AK 86 8.3 7.0 27.2
Amputation cause
Vascular 153 8.1 .63 6.1 .81 22.7 .95
Nonvascular 176 9.4 8.2 30.5
Mobility device
No device used 204 9.7 1.62‡ 8.5 1.57 30.9 1.66
Device used 76 6.4 4.4 17.3
Walking distance
Unlimited 194 9.4 .54 8.9 1.08 31.3 .92
⬍1 block 135 8.3 6.1 23.8
Automatic walking
Yes 205 9.4 .78 8.5 1.32 30.6 .59
No 124 7.8 5.1 25.8

* Effect Size (ES) ⫽ mean difference between groups/SD; values of 0.2 ⫽ small, 0.5 ⫽ medium, 0.8 ⫽ large.
†
Means between groups not significantly different at p ⬍ .05.
‡
Mobility device use is an item of the Houghton Scale.

exist, and what their magnitude may be. The difference, if any essential to the use of an outcome tool. In the present study, we
exists, may suggest potential for rehabilitation.38 Therefore, used a 4-week interval to obtain estimates of test-retest reli-
rather than providing similar information, the Houghton Scale ability of the 3 prosthetic mobility scales. This interval was
may provide information different from either the PPA-LCI or similar to that used by the authors of the PPA and PEQ.12,17 We
PEQ mobility subscale. also purposely selected individuals who were considered med-
The level of measurement or number of response categories ically and prosthetically stable to ensure that we could maxi-
can influence the quality of a measure. Examining the distri- mize the reliability estimate. Debate continues regarding what
bution of the scores provides insight into the importance of this constitutes acceptable reliability. A variety of values, from 0.6
factor. As an example, all 3 scales had asymmetric distribu- to 0.9, have been reported to indicate minimal reliability when
tions, with the PPA-LCI having a very high ceiling effect ICC values are used.18,30,36,40,41 Contributing to the difficulty of
(40%) in both samples. More confidence is placed in scales that defining an acceptable standard is whether one is comparing
have a more symmetric distribution of values.37 One disadvan- groups or individuals. Higher coefficients (.85 or better) are
tage of scales with high ceiling effects is that they limit indi- recommended if the goal of the tool is to interpret data at the
viduals’ ability to improve, or they are incapable of capturing individual level.36,41 The test-retest reliability for all 3 scales
change among higher functioning individuals. Because the exceeded .75, considered minimally acceptable for group com-
PPA-LCI has such a large ceiling effect, this scale may be more parison.36
appropriate for a particular segment of the amputee population The test-retest estimate we found for the PPA-LCI was
such as the older individual with multiple comorbidities. The stronger (.88 vs .80) than that reported by the authors,12
distribution for the PEQ mobility subscale is more symmetric, whereas the value for the PEQ mobility subscale was weaker
perhaps because it has more response options, in our case than the original estimate.17 Several factors could explain the
ranging from 0 to 10. A larger response scale provides a greater difference in reliability estimates from the original published
variation and potentially a better measure of change.32 results. Reasons range from the type of analyses used to dif-
ferences related to the composition of the sample population,
Reliability all of which will either influence the scale’s true variance or its
Ensuring that the same measurement can be consistently error variance.36 Although all 3 studies used the ICC to provide
obtained when the status of the subject remains unchanged is an estimate of reliability, the type of ICC calculated can influ-

Table 6: Means and F Statistics for Different Groups for Each Measure (n ⴝ 329)

n Houghton Scale PEQ Mobility Subscale PPA-LCI

Wearing time* quartiles (hr/wk)

0–69.9 70 6.7 5.6 27.7*
70–97.9 80 9.0* 6.4 30.1*
98–111.9 81 9.5* 7.7 36.7*
ⱖ12 98 9.6* 8.7 39.2*
F statistic 46.5 30.2 26.1
Significance .000 .000 .000
F ratio† 1.0 1.5 1.8

NOTE. Wearing time is an item on the Houghton Scale.

* Means between groups not significantly different at p ⬍ .05.
†
Numerator for ratio was the largest F value (Houghton).

Arch Phys Med Rehabil Vol 82, October 2001

1438 COMPARISON OF 3 PROSTHETIC MOBILITY SCALES, Miller

Table 7: Rank-Order Score Comparison of the 3 Mobility Scales

Dimensions* Houghton Scale PEQ Mobility Subscale PPA-LCI

Descriptive 4.0 4.0 4.0

Reliability 3.0 4.0 5.0
Validity 11.5 14.0 14.5
Total score 28.5 22.0 23.5

* See Methods section to determine which components were considered within each dimension. Ranking scores: 3 ⫽ best, 1 ⫽ least.
Calculation of scores for ties: score1 ⫹ score2/2. Assumes all factors are weighted equally in terms of importance.

ence the amount of error variance. We chose to use a conser- Validity

vative estimate by using the 1-way ANOVA method because it Validity is assessed from the accumulation of evidence from
adjusts for possible learning effect between test periods.36,42 a variety of sources over time.47 Establishing validity by using
The error related to the different testing times is eliminated by empirical evidence provides support that the test scores are
using the 2-way ANOVA.43 It is not clear what type of ICC measuring what they purport to measure.48 In the present study,
was calculated in the other studies. validity was assessed by several different techniques. First, we
Other ways to improve reliability are related to enhancing observed the ability of each measure to converge with other test
the true variance. To enhance the true variance, one can reduce measures with respect to the direction and strength of a priori
the tools ceiling or floor effect. To achieve a more symmetric hypotheses.36 In all cases the anticipated direction of the rela-
distribution in the case of a high ceiling effect, easy items can tionships was observed. Although we found a range of corre-
be discarded or more difficult items can be added. Alterna- lation coefficients, all of them were statistically significant at a
tively, by modifying the response categories to provide a 5% alpha level. The PPA-LCI consistently had the largest
greater range of difficulty and by changing their descriptors one coefficients across all variables.
can influence the amount of true variance. An example of this Next, we examined the ability of each measure to distinguish
is the difference in the number of response categories among between variables considered important to prosthetic mobility
the 3 scales. Although the Houghton Scale and the PPA-LCI from a clinical perspective. The Houghton Scale was the only
are limited to 4 categories each, our version of the PEQ measure that was able to distinguish statistically between all
mobility subscale ranged from 0 to 10. Increasing the number the groups, whereas the other 2 scales failed to detect a differ-
of response categories should increase the opportunity to cap- ence only between the different amputation levels. Several
ture true variance. Logic suggests that error variance may reasons may account for this finding. First, the Houghton Scale
increase. Therefore, the estimates for the Houghton Scale and is a performance measure and, as anticipated from clinical
PPA-LCI should be higher than for a scale with more response observations, a performance difference exists between AK and
categories. However, we anticipated that our reliability esti- BK amputees. Although no difference may exist in the per-
mate would be larger than that of the original PEQ mobility ceived capability between individuals who have amputations at
subscale because Legro et al17 used a VAS format, which the different levels, the actual performance may be different
potentially has a greater chance for test-retest variability. between the amputation levels. Second, this finding could be
Although it is generally accepted that reliability improves the result of chance, which is plausible given that the mean
when a scale has more items,1,36 another reason why the reli- difference between scores on the Houghton Scale is much
ability estimate from the present study was lower may be smaller than one would expect. Finally, according to data from
because we combined the ambulation and transfer subscales of the present study, it may be possible that clinical impression is
the PEQ into a mobility subscale. However, post hoc analyses not supported statistically by using these scales to differentiate
showed that the ICCs were similar (.77) after partitioning the between BK and AK amputation levels.
mobility items We also calculated effect size to estimate the relative preci-
Sampling influences the error variance because a more het- sion of each scale to distinguish between groups. The larger the
erogeneous test population will improve the degree of error effect size, the greater the power of the test to detect differ-
variance.36 Overall, we had a smaller sample but a better ences and, consequently, the smaller the sample size required
response rate (90% vs 81% PEQ vs 84% PPA). Review of the for studies.31 Cohen31 suggested that effect size values of 0.8 or
sample composition from the different studies reveals that higher are large, 0.5 is medium, and 0.2 is small. With this in
variation exists related to age, gender, amputation cause, and mind, the Houghton Scale and PPA-LCI had medium to large
level. To improve generalizability of the reliability and validity effect size values and the PEQ mobility subscale had large
results, samples should follow the distributions of the popula- effect size values for all group differences except amputation
tion of interest as closely as possible. level. No single tool was consistently better at distinguishing
The reliability sample of the present study closely follows between groups. For instance, the PEQ mobility subscale had
the age, gender, and amputation level distribution of the larger higher precision when assessing walking distance and the abil-
clinic population. It appears that we oversampled individuals ity to walk without thinking, whereas the PPA-LCI had the
who had their amputations related to dysvascular causes and highest precision for distinguishing between amputation cause
those who had a recent amputation. It is unclear how represen- and use of a mobility device.
tative the other samples were of their respective regional pop-
ulations. In the case of the PEQ, the sample consisted primarily
of younger men whose amputations were for reasons of Future Research
trauma.17 When examining incident cases, we found that trau- Because we modified the PEQ in 2 ways for the purpose of
matic injury represents only 10% of all amputees, whereas the present study (combining the transfer and ambulation sub-
amputations for vascular pathology range from 50% to 95% of scales and then changing the response format from VAS to a
incident cases.44-46 10-point numeric format), the learning and abstraction it re-

Arch Phys Med Rehabil Vol 82, October 2001

 COMPARISON OF 3 PROSTHETIC MOBILITY SCALES, Miller 1439

quired may have posed special difficulty for older individuals 14. Houghton AD, Taylor PR, Thurlow S, Rootes E, McColl I. Suc-
who were at risk for cerebral infarcts from vascular insuffi- cess rates for rehabilitation of vascular amputees: implications for
ciency. Further testing is required to compare the PEQ’s use, preoperative assessment and amputation level. Br J Surg 1992;79:
reliability, and responsiveness by using different response op- 753-5.
tions. 15. Houghton A, Allen A, Luff R, McColl I. Rehabilitation after lower
We calculated summary scores for each scale. This approach limb amputation: a comparative of above-knee, through-knee and
assumes equality between the different levels of the response Gritti-Stokes amputations. Br J Surg 1989;76:622-4.
categories. Additional research, by using such techniques as 16. Leung EC, Rush PJ, Devlin M. Predicting prosthetic rehabilitation
outcome in lower limb amputee patients with the Functional
Rasch analysis, should be conducted to test this assumption.
Independence Measure. Arch Phys Med Rehabil 1996;77:605-8.
The responsiveness of the scales needs to be determined by 17. Legro MW, Reiber GD, Smith DG, del Aguila M, Larsen J, Boone
examining the minimally important difference that the scales D. Prosthesis Evaluation Questionnaire for persons with lower
are able to detect. We did not assess responsiveness, but have limb amputations: assessing prothesis-related quality of life. Arch
provided independent evidence of the reliability, which is a Phys Med Rehabil 1998;79:931-8.
necessary step to investigating responsiveness. After all, a tool 18. Donner A, Eliasziw M. Sample size requirements for reliability
that is not stable when change does not occur will not reliably studies. Stat Med 1987;6:441-8.
indicate change when change has occurred.4,42 Additional re- 19. Miller WC. Falling, fear of falling and balance confidence among
search is required to determine the responsiveness of the 3 individuals with a lower extremity amputation: prevalence and
scales. consequence [dissertation]. London (Ont): Univ Western Ontario;
2000.
CONCLUSION 20. Guyatt GH, Townsend M, Berman LB, Keller JL. A comparison
The Houghton Scale provides a quick measure of prosthetic of likert and visual analogue scales for measuring change in
function. J Chron Dis 1987;40:1129-33.
mobility performance, whereas the other 2 scales assess mo-
21. McGavin CR, Gupta SP, McHardy GJ. Twelve-minute walking
bility capability. The psychometrics of the 3 scales are rela- test for assessing disability in chronic bronchitis. Br Med J 1976;
tively similar, but the overall summary score suggests that the 1:822-3.
PPA-LCI provides slightly better properties. The high ceiling 22. Butland RJ, Pang J, Gross ER, Woodcock AA, Geddes DM. Two-,
effect of the PPA-LCI suggests that this scale may be more six-, and 12-minute walking test in respiratory disease. Br Med J
beneficial when assessing low-end prosthetic mobility. 1982;284:1607-8.
23. Mathias S, Nayak USL, Isaacs B. Balance in elderly patients: the
Acknowledgments: The authors acknowledge Cindy Lee and “Get-up and Go” test. Arch Phys Med Rehabil 1986;67:387-9.
Rubini Pathy who, as physical therapy students, assisted with subject 24. Podsiadlo E, Richardson S. The Timed “Up & Go”: a test of basic
recruitment, data collection, and management. We also extend appre- functional mobility for frail elderly persons. J Am Geriatr Soc
ciation to Marie-Claude Grisé for providing background information 1991;39:142-8.
on the PPA and the Prosthetics Research Study Group for providing 25. Schoppen T, Boonstra A, Groothoff JW, de Vries J, Göeken LN,
information on the PEQ. Eisma WH. The timed “Up and Go” test: reliability and validity in
persons with unilateral lower limb amputation. Arch Phys Med
References Rehabil 1999;80:825-8.
1. Feinstein AR, Josephy BR, Wells CK. Scientific and clinical 26. Powell LE, Myers AM. The Activities-specific Balance Confi-
problems in indexes of functional disability. Ann Intern Med dence (ABC) scale. J Gerontol A Biol Sci Med Sci 1995;50:M28-
1986;105:413-20. 34.
2. Keith RA. Functional assessment measures in medical rehabilita- 27. Myers AM, Fletcher PC, Myers AH, Sherk W. Discriminative and
tion: current status. Arch Phys Med Rehabil 1984;65:74-8. evaluative properties of the Activities-specific Balance Confi-
3. Keith RA. Conceptual basis of outcome measures. Am J Phys dence (ABC) scale. J Gerontol A Biol Sci Med Sci 1998;53:
Med Rehabil 1995;74:73-80. M287-94.
4. Williams IJ, Naylor CD. How should health status measures be 28. Portney LG, Watkins MP. Foundations of clinical research: ap-
assessed? Cautionary notes on procrustean frameworks. J Clin plications to practice. Stamford (CT): Appleton & Lange; 1993.
Epidemiol 1992;45:1347-51. 29. Prieto L, Alonso J, Ferrer M, Antó JM. Are results of the SF-36
5. Kent R, Fyfe N. Effectiveness of rehabilitation following ampu- health survey and the Nottingham Health Profile Similar? A
tation. Clin Rehabil 1999;13:43-50. comparison in COPD patients. J Clin Epidemiol 1997;50:463-73.
6. Collin C, Wade DT, Cochrane GM. Functional outcome of lower 30. Nunnally JC, Berntein IH. Psychometric theory. New York:
limb amputees with peripheral vascular disease. Clin Rehabil McGraw-Hill; 1994.
1992;6:13-21. 31. Cohen J. Statistical power analysis for the behavioral sciences.
7. Siriwardena GJ, Bertrand PV. Factors influencing rehabilitation of Hillsdale (NJ): Lawrence Erlbaum; 1988.
arteriosclerotic lower limb amputees. J Rehabil Res 1991;28:35- 32. Bombardier C, Tugwell P. Methodological considerations in func-
44. tional assessment. J Rheumatol 1987;14 Suppl 15:6-10.
8. Kegel B, Carpenter ML, Burgess EM. Functional capabilities of 33. Kirshner B, Guyett G. A methodological framework for assessing
lower extremity amputees. Arch Phys Med Rehabil 1978;59:110- health indices. J Chron Dis 1985;38:27-36.
20. 34. Grisé MC, Gauthier-Gagnon C, Martineau GG. Prosthetic profile
9. Narang IC, Mathur BP, Sigh P, Jape VS. Functional capabilities of of people with lower extremity amputation: conception and design
lower limb amputees. Prosthet Orthot Int 1984;8:43-51. of a follow-up questionnaire. Arch Phys Med Rehabil 1993;74:
10. Pohjolainen T, Alaranta T, Karkkainen M. Prosthetic use and 862-70.
functional and social outcome following lower limb amputation. 35. McDowell I, Newell C. Measuring health: a guide to rating scales
Prosthet Orthot Int 1990;14:75-9. and questionnaires. 2nd ed. New York: Oxford Univ Pr; 1996.
11. Wolf E, Lilling M, Ferber I, Marcus J. Prosthetic rehabilitation of 36. Streiner DL, Norman GR. Health measurement scales. A practical
elderly bilateral amputees. Int J Rehabil Res 1989;12:271-8. guide to their development and use. New York: Oxford Univ Pr;
12. Gauthier-Gagnon C, Grisé MC. Prosthetic Profile of the Amputee 1995.
Questionnaire: validity and reliability. Arch Phys Med Rehabil 37. Wade DT. Measurement in neurological rehabilitation. New York:
1994;75:1309-14. Oxford Univ Pr; 1992.
13. Gauthier-Gagnon C, Grisé MC, Lepage Y. The Locomotor Capa- 38. Feinstein A. Clinimetrics. New Haven (CT): Yale Univ Pr; 1987.
bilities Index: content validity. J Rehabil Outcome Meas 1998;2: 39. Young NL, Williams JI, Yoshida KK, Bombardier C, Wright JG.
40-6. The context of measuring disability: does it matter whether capa-

Arch Phys Med Rehabil Vol 82, October 2001

1440 COMPARISON OF 3 PROSTHETIC MOBILITY SCALES, Miller

bility or performance is measured? J Clin Epidemiol 1996;49: 44. Ebskov LB. Level of lower limb amputation in relation to etiol-
1097-101. ogy: an epidemiological study. Prosthet Orthot Int 1992;16:163-7.
40. Fleiss JL. The design and analysis of clinical experiments. New 45. Gregory-Dean A. Amputations: statistics and trends. Ann R Coll
York: Wiley; 1986. Surg Engl 1991;73:137-42.
41. McHorney CA, Tarlov AR. Individual-patient monitoring in clin- 46. Prohjolainen T, Alaranta T. Lower limb amputations in Southern
ical practice: are available health status surveys adequate? Qual Finland 1984-1985. Prosthet Orthot Int 1988;12:9-18.
Life Res 1995;4:293-307. 47. Anastasi A. Psychological testing. 6th ed. New York: Macmillian;
42. Beaton DE, Hogg-Johnson S, Bombardier C. Evaluating changes 1988.
in health status: reliability and responsiveness of five generic 48. Messick S. Validity of psychological assessment: validation of
inferences from persons’ responses and performances as scientific
health status measures in workers with musculoskeletal disorders.
inquiry into score meaning. Am Psychologist 1995;50:741-9.
J Clin Epidemiol 1997;50:79-93.
43. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing Supplier
rater reliability. Psychol Bull 1979;19:3-11. a. SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

Arch Phys Med Rehabil Vol 82, October 2001

Accepted Manuscript

Two-Minute versus 6-Minute Walk Distances during 6-Minute Walk Test in

Neuromuscular Disease: Is the 2-Minute Walk Test an Effective Alternative to a 6-
Minute Walk Test?

J.W. Witherspoon, R. Vasavada, R.H.H. Logaraj, M. Waite, J. Collins, C. Shieh, K.

Meilleur, C. Bönnemann, M. Jain
PII: S1090-3798(18)30105-3
DOI: [Link]
Reference: YEJPN 2486

To appear in: European Journal of Paediatric Neurology

Received Date: 5 April 2018

Revised Date: 1 October 2018
Accepted Date: 2 October 2018

Please cite this article as: Witherspoon J, Vasavada R, Logaraj RH, Waite M, Collins J, Shieh C,
Meilleur K, Bönnemann C, Jain M, Two-Minute versus 6-Minute Walk Distances during 6-Minute Walk
Test in Neuromuscular Disease: Is the 2-Minute Walk Test an Effective Alternative to a 6-Minute Walk
Test?, European Journal of Paediatric Neurology (2018), doi: [Link]

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Two-Minute versus 6-Minute Walk Distances during 6-Minute Walk Test in

Neuromuscular Disease: Is the 2-Minute Walk Test an Effective Alternative to a 6-Minute
Walk Test?
Witherspoon, JWa; Vasavada, Rb; Logaraj RHb, H; Waite, Mb; Collins Jc, Shieh Cd, Meilleur Ka,
Bönnemann, Ce; Jain, Mb

PT
a
National Institute of Nursing Research/Tissue Injury Branch/Neuromuscular Symptoms Unit,

RI
National Institutes of Health; Bethesda, MD, 20814; USA
b
Clinical Research Center, Rehabilitation Medicine Department, National Institutes of Health;

SC
Bethesda, MD, 20814; USA
c
Department of Rehabilitation Science, George Mason University; Fairfax, VA, 22030; USA
d

U
Office of Research on Women’s Health, National Institutes of Health; Bethesda, MD, 20814;
AN
USA
e
Neuromuscular and Neurogenetic Disorders of Childhood Section, Neurogenetics Branch,
M

National Institutes of Health; Bethesda, MD, 20814; USA

D
TE

[Link]@[Link], ruhivasavada@[Link], rlogaraj@[Link],

EP

[Link]@[Link], jcolli22@[Link], [Link]@[Link], meilleurk@[Link],

[Link]@[Link], Mjain@[Link]
C
AC

Corresponding Author:
Jessica Witherspoon
[Link]@[Link]
C: (302) 236 4030
 ACCEPTED MANUSCRIPT

Abstract

Functional tests such as Motor Function Measure-32 (MFM-32), supine to stand, ascend/descend

stairs permit the assessment of task-specific motor function in neuromuscular disease (NMD).

The 6-minute walk test (6MWT), though functional, is primarily used to assess endurance and

PT
disease progression in children with neuromuscular disorders. Barriers to 6MWT administration,

RI
in this population, can include reduced attention span due to age and inability to tolerate test

length due to weakness. We propose task-specific functional deficits are related to endurance.

SC
Additionally, the 2-minute walk test (2MWT) could effectively replace the 6MWT in this

population. Seventy-seven participants, ages 5-18, with a variety of neuromuscular disorders

U
performed the 6MWT, timed functional tests (TFT), and the MFM-32. Correlation and paired t-
AN
test analyses were used to compare the distance walked in the first 2 minutes (2MWD) to the

distance walked in the entire 6 minutes (6MWD) and to the functional outcome measures above.
M

The 2MWD strongly correlated with 6MWD and the other outcome measures. Paired t-test
D

analysis also showed that the 2MWD did not differ from the distance walked in the last two
TE

minutes of the 6MWT. Although equivalence testing could not reject the claim that this

difference exceeded the upper practical limit of 9.5 m, it only showed a modest overestimation of the
EP

4-6MWD compared with the 2MWD. Together, our results support the ability of the 2MWD to

predict the 6MWD, specifically in the pediatric neuromuscular disease population.

C
AC

Keywords: 2-minute walk test, 6-minute walk test, neuromuscular disorders, pediatrics
 ACCEPTED MANUSCRIPT

1 1. Introduction

2 The 6-minute walk test was originally developed to assess walking capability, disease

3 progression, and treatment efficacy [1] in adults with cardiopulmonary conditions [2]. It is now

PT
4 also used in individuals with neuromuscular disorders including Duchenne muscular dystrophy

5 (DMD), Kennedy’s disease, inclusion body myositis, and spinal muscular atrophy and in other

RI
6 neurological disorders such as multiple sclerosis [3-7]. Although previous work supports the va-

SC
7 lidity and reliability of the 6MWT in the neuromuscular disease population, the length of this test

8 limits the ability of participants to complete the assessment. Thirty percent of adolescents with

U
9 neuromuscular disease (NMD) were shown to have behavioral challenges about three times
AN
10 greater than children without NMD [8]. Therefore, behavioral concerns (i.e., personality disor-

11 ders, social problems, attention deficits, anxiety, and/or depression) [8], reduced ambulatory abil-
M

12 ity, and cognitive impairment, especially when fatigued, are of particular concern in the pediatric

13 population. Thus, the 6MWT may not serve as an effective measure for monitoring disease pro-
D

14 gression and functional ability in the pediatric population [9, 10]. However, the 2MWT is of
TE

15 shorter duration, objective, easily administered, and provides a standardized evaluation of func-

16 tional capacity in populations with reduced ambulatory capability, including adults with neuro-
EP

17 muscular disease [11-16].

C

18 The 2-minute walk test (2MWT, independent of 6MWT) has been used as an endurance
AC

19 measure in the aged population and in individuals with lower extremity amputations, cystic fi-

20 brosis, traumatic brain injury, and neurological disorders [11-18]. In support, a strong correlation

21 between the distance walked in the 2MWT and 6MWT was observed in subjects with sporadic

22 inclusion body myositis [11] and a range of neuromuscular diseases [1]. Studies also showed

23 strong predictability of 6MWT performance based on 2MWT results and vice versa [1, 11].

1
 ACCEPTED MANUSCRIPT

24 Kosak and Smith [19] compared the validity of the 2, 6, and 12-minute walk tests in adults re-

25 covering from a stroke and showed the distance walked in 2 minutes is highly correlated

26 (r=0.997) with distance walked in 6 (r=0.997) and 12 minutes (r=0.995). We propose the results

27 of this study will support the 2MWT as an effective alternative to the 6MWT in pediatric popula-

PT
28 tions affected by neuromuscular disease, without compromising reliability.

RI
29 Other clinical endpoints often used in NMD to assess motor function and functional abil-

SC
30 ity include Motor Function Measure-32 (MFM-32) and Timed Functional Tests (TFTs) [20]. The

31 MFM-32 is a valid and reliable assessment of motor function and disease progression. It contains

U
32 three motor function domains including supine-to-stand (domain 1, D1), axial and proximal
AN
33 (domain 2, D2), and distal (domain 3, D3) [21]. This assessment is complemented by TFTs [21].

34 TFTs such as 10m walk/run, supine-to-stand, and stair ascent/descent are task-specific tests that
M

35 have been shown to correlate with 6MWD in DMD [22]. Previous work, in Inclusion Body My-

36 ositis (IBM), has shown the 2MWT is highly correlated with the 6MWT and other functional
D

37 tests including stair ascent/descent thus making the 2MWT a potential alternative to the 6MWT.
TE

38 Similar to other NMDs, some of the patients with IBM were unable to finish the 6MWT due to

39 fatigue. As such, the authors propose the 2MWT as a better alternative given it is “less fatiguing
EP

40 and better tolerated.” [11]

C

41 The primary objective of this study is to establish the utility of the 2MWT based on the
AC

42 correlation between 2-minute walk distance (2MWD, the first 2 minutes of the longer 6-minute

43 walk) and the 6-minute walk distance (6MWD) in children, with various neuromuscular disor-

44 ders, who performed the 6MWT. We use the term “distance” in both cases (2MWD, 6MWD) to

45 highlight the difference between our study testing and previous reports, in which the 2MWT and

46 the 6MWT were administered separately. The secondary objective is to compare the 2MWD and

2
 ACCEPTED MANUSCRIPT

47 6MWD to motor performance tests including the timed floor-to-stand test, timed ascent and de-

48 scent of stairs, and the Motor Function Measure 32 (MFM 32).

49 2. Materials and Methods

PT
50 2.1 Subjects

RI
51 A total of 77 participants ages 5 to 18 enrolled in two natural history protocols at the Na-

52 tional Institutes of Health between 2010 and 2015 after obtaining informed consent and/or assent

SC
53 previously approved by the Institutional Review Board. A variety of neuromuscular disorders

54 were represented in this sample, including collagen VI-related dystrophy (COL6-RD), laminin

55
U
alpha 2-related dystrophy (LAMA2-RD), limb-girdle muscular dystrophy (LGMD), and RYR1-
AN
56 related myopathies (RYR1-RM), among other neuromuscular diagnoses (Table 1). All enrolled
M

57 ambulatory participants were able to walk 10 meters (m) without the use of assistive devices

58 and/or ankle-foot orthoses. Subjects were evaluated using a battery of functional tests, including
D

59 the 6MWT, timed functional tests, and the MFM-32.

TE

60 2.2 6MWT

61 The 6MWT was administered as per the American Thoracic Society (ATS) guidelines
EP

62 [23], with the exception of pulse oximetry and use of a 30m corridor. The modified ATS guide-
C

63 lines for rest were also not used. Upon arrival to the testing site, participants rested for 5 minutes.
AC

64 Blood pressure and heart rate were measured pre-exercise, immediately post-exercise, and after a

65 5-minute recovery. Participants were instructed to walk “as quickly and safely as possible” in a

66 50m corridor for 6 minutes. Participants did not receive external motivation. The use of orthotics

67 and assistive devices was not permitted. Participants were also given small objects to drop at the

68 end of each minute as a marker of distance walked in that minute. If participants required rest

3
 ACCEPTED MANUSCRIPT

69 during the course of the test, they were permitted to do so without sitting or leaning against ob-

70 jects. A clinician trailed slightly behind the participants to ensure safety and avoid a pacing bias.

71 If the protocol required participants to return for subsequent visits, only the first administration

72 of the test was used for analysis.

PT
73 2.3 Motor Function Measure 32 (MFM-32)

RI
74 The MFM-32 is a 32-item test that quantifies functional capabilities in individuals with

SC
75 neuromuscular disorders [24]. This motor assessment has been validated in the congenital mus-

76 cular dystrophy, congenital myopathy, and limb girdle muscular dystrophy populations [21]. The

U
77 32 items are scored on a 4-point Likert scale and are further divided into subcategories including:
AN
78 D1 (13 items) standing and transfers, D2 (12 items) axial and proximal motor function, and D3

79 (7 items) distal motor function. Standing and transferring function was further analyzed in rela-
M

80 tion to the 6MWT. This subcategory includes hopping on one foot, running 10m, walking on
D

81 heels, etc.
TE

82 2.4 Timed Functional Tests

83 Timed and graded functional tests (TFT) were performed, including the floor-to-stand
EP

84 test, ascending 4 steps, and descending 4 steps. The TFT have been used as outcome measures in
C

85 several neuromuscular disease clinical trials [25-27]. Taken together, these assessments capture
AC

86 functional ability as does the 6MWT [22].

87 2.4.1 Floor to Stand

88 Participants were positioned in supine on a floor mat. They were instructed to stand up as

89 quickly as possible without using external support. If external support was needed, a

4
 ACCEPTED MANUSCRIPT

90 sturdy table or chair was used. The timer was started as soon as the test administrator

91 said, “go” and stopped once the participant was in a standing position.

92 2.4.2 Ascend & Descend 4 Steps

PT
93 The ascend/descend step test was administered on a 4-step staircase with handrails. Par-

94 ticipants were instructed to climb the steps as safely and quickly as they could once the

RI
95 test administrator said, “go.” Once the participants reached the top, they were asked to

SC
96 stop, at which point the timer was also stopped. Participants were then asked to descend

97 the steps as safely and quickly as they could. The timer for the second part of the test be-

U
98 gan when the test administrator said, “go” and stopped once the participant reached the
AN
99 bottom.

100 2.5 Statistical Analysis

M

101 Descriptive statistics are given in the form of mean (±SD) or N (%). Regression of
D

102 6MWD on 2MWD was used to define the predicted 6MWD, with R-squared measured to deter-
TE

103 mine goodness of fit. The Shapiro-Wilk test of Normality was used to assess 2MWD, 6MWD,

104 and final 2 minutes’ distance (4-6 MWD) data for evidence of non-Normality. Pearson correla-
EP

105 tion was calculated for all pairs of timed distances and TFTs. Spearman correlation of timed dis-

106 tances and MFM-32 D1 and total score were also calculated. Equivalence testing with the two
C

107 one-sided tests procedure (TOST) assessed comparability of 2MWD and 4-6MWD. The equiva-
AC

108 lence margin was set to 9.5m, as this is one third of the minimal clinical important difference

109 (MCID) established by McDonald et al. for the 6MWT (22). While their study used a 30m

110 corridor and we used a 50m corridor, it has been shown that six-minute walk distance is not af-

111 fected by straight courses ranging between 15m to 50m [23]. Bland-Altman plots were also used

5
 ACCEPTED MANUSCRIPT

112 to assess agreement between the 2MWD and 4-6MWD. Statistical significance was set as p <

113 0.05. All analyses were conducted in Stata 14 [28].

114 3. Results

PT
115 Seventy-seven participants completed the 6MWT. Of these, 73 participants had a meas-

116 ured 4-6MWD. Sixty-six subjects completed the floor-to-stand test, 72 subjects completed the

RI
117 timed ascent and descent of stairs, and 69 subjects completed the MFM-32 (Table 2).

SC
118 Mean distances walked were comparable for the initial two minutes at 149.8 m (±40.3)

119 compared to the final two minutes 149.2m (±42.2). The mean 6MWD was 442.1 m (±121.6).

U
120 The scatterplot and regression line of 6MWD against 2MWD are given in Figure 1. The 95%
AN
121 confidence interval for the slope of the regression was [2.77, 3.10]. The mean difference between

122 the predicted 6MWD from the regression and that obtained by multiplying 2MWD by three (an
M

123 assumption that speed was the same throughout the 6 minutes) was 7.2m and was statistically
D

124 significant (p < 0.0001), but was smaller than the margin of equivalence of 9.5m.
TE

125 Four subjects were excluded from analyses comparing the 2MWD with the 4-6 MWD as

126 intermediate distances were not recorded. With these four excluded, mean 2MWD was slightly
EP

127 higher at 152.4 m (±39.3). Among the 73 subjects with both measurements, the 2MWD demon-

128 strated a strong correlation with the 4-6 MWD (r=0.896, p<0.0001) (Figure 2). The 2MWD and
C

129 4-6 MWD were not significantly different (p = 0.07) with a mean difference of 5.07m and 95%
AC

130 CI [-0.52, 10.65]. Equivalence testing rejected the lower margin of equivalence (p < 0.0001) but

131 failed to reject the claim that 2MWD was at least 9.5m greater than 4-6 MWD (p = 0.0591). The

132 Bland-Altman plot for 2MWD and 4-6MWD showed strong agreement between both measure-

133 ments, with 97% of the data falling within the limits of agreement (see supplemental figure 1).

6
 ACCEPTED MANUSCRIPT

134 The distance walked in both 2 minutes and 6 minutes showed moderately strong inverse

135 correlations with all four TFT tests (-0.7 < r < -0.6). For each TFT test, both distances had com-

136 parable correlation coefficients and all correlations were significant at p <0.0001 (Table 3). Ad-

137 ditionally, strong correlations existed between both the 2MWD and the 6MWD with the D1 cat-

PT
138 egory of the MFM-32 as well as MFM-32 total score (Table 3).

RI
139 4. Discussion

SC
140 The 6MWT is used as a measure of endurance and function in pediatric populations with

141 reduced ambulatory ability, resulting from neuromuscular diseases. The 2-minute walk test

U
142 (2MWT, independent of 6MWT) has been used in the aged population, individuals with lower
AN
143 extremity amputations, cystic fibrosis, traumatic brain injury, and neurological disorders as a

144 measure of endurance [11-18]. It is also believed to serve as a potential alternative to the 6MWT
M

145 in study populations that have issues with behavior, attention span, and/or the inability to tolerate
D

146 test length due to weakness [8-10]. The findings of this study support the potential of a 2MWT as
TE

147 an effective alternative to the 6MWT as a measure of function and performance in children with

148 neuromuscular diseases. Our study demonstrated a strong correlation between the 2MWD and
EP

149 the 6MWD. This result was supported by equivalence testing confirming that 2MWD may mod-

150 estly overestimate 4-6 minute distance. Regression of 6MWD on 2MWD provided further sup-
C

151 port for approximating 6MWD with 2MWD multiplied by a factor of 3. The regression equation
AC

152 was 6MWD = 2.93*(2MWD) + 2.64, and the value 3 was within the 95% CI for the slope. These

153 results consistently support the 2MWT as an alternative to the 6MWT for this population.

154 In addition to demonstrating the validity of the 2MWT as an endurance measure, we as-

155 sessed the validity of the 2MWT as a functional measure. Both the 2MWD and the 6MWD were

156 inversely correlated with the time taken to go from floor-to-stand, ascend 4 steps, and descend 4

7
 ACCEPTED MANUSCRIPT

157 steps. The correlation coefficients observed for both the 2 and 6-minute walk distances were very

158 similar, further confirming the validity of the 2MWT as a performance measure. The 2MWD and

159 6MWD also showed a strong correlation with the D1 MFM-32 score. The D1 score was specifi-

160 cally assessed versus the other two domains of the MFM-32, because it measures standing and

PT
161 transfer capability in these individuals. These activities most closely resemble the type of func-

RI
162 tion assessed in the 2MWT and 6MWT. This strong correlation provides further evidence for the

163 use of the 2MWT as a functional measure.

SC
164 One of the major concerns with the 2MWT is that children may walk at a faster pace giv-

165 en the shorter time frame. However, Alfano et al. showed 2MWT walking velocity was highly

166
U
correlated with 6MWT velocity in IBM, but were significantly different. Given the velocities
AN
167 were highly correlated, the authors were able to use linear regression to obtain an equation that

168 allowed for the conversion of 2MWT performance to 6MWT [11]. Based on these results, a line-
M

169 ar regression model can be used to resolve the concern of faster walking speeds in a 2MWT.
D

170 Since we used the first two minutes of the 6MWD and did not conduct a separate 2MWT, we
TE

171 compared the distance traveled and walking speed at different time points to confirm that our

172 subjects maintained an even pace throughout the test. Table 4 shows walking velocities at each
EP

173 minute interval. Among participants with 2MWD and 4-6 MWD data, they ambulated a mean

174 distance of 152m in the first two minutes and 149m in the last 2 minutes of the test, walking at
C

175 an average velocity of 1.27 m/s and 1.24 m/s for the first two and last two minutes respectively.
AC

176 While the distance walked during these time frames correlated very strongly at .896, the velocity

177 decreased by 2.3% in the last 2 minutes from the first 2 minutes (table 4). Andersen et al. [1] re-

178 ports a decrease in the walking speed from the first to the 6th minute by 1.4% in healthy adult

179 controls. Equivalence testing between the first two minutes and last two minutes revealed that

8
 ACCEPTED MANUSCRIPT

180 differences in these distances were not statistically significant and did not reach the equivalence

181 margin. Since this equivalence margin was derived from a 6MWT MCID and not a 2MWT

182 MCID, we wanted to further assess this relationship using a Bland-Altman plot. Bland Altman

183 suggests two measures, 2-minute and 6-minute walking velocities, are considered in agreement if

PT
184 95% of the data fall within ±2s, where s is the standard deviation [29]. In our plot, there were

RI
185 only 3 data points that fell outside the upper and lower limits.

186 4.1 Limitations

SC
187 While these results support the use of the 2MWT as a substitute for the 6MWT, there is

188 concern that comparing the initial 2MWD to the total 6MWD could produce a pacing effect. Par-

189
U
ticipants may have selected a faster walking speed had they been asked to walk for 2 minutes as
AN
190 opposed to 6 minutes. However, this concern has been addressed by use of a linear regression

191 model given the high correlation between 2 and 6 minute speeds [11]. Another option is a fol-
M

192 low-up study that compares the 2MWT and 6MWT in NMD. The tests would be administered on
D

193 different days to provide adequate recovery time to avoid the effects of fatigue.
TE

194 The lack of MCID for this specific patient population is another limitation. This study

195 used the MCID from the DMD population which may not apply to all neuromuscular disorders
EP

196 of childhood equally. Thus, having MCIDs specific to some of the conditions in this study would

197 be of benefit. Furthermore, the MCID we used to compare the 2MWD and 4-6MWD was ex-
C

198 trapolated from the 6MWT MCID. A follow-up study might explore the MCID specific to a
AC

199 2MWT in our sample population.

200 5. Conclusion

201 Our findings suggest that the 2MWT may replace the 6MWT as a measure of both endur-

202 ance and functional ability in a pediatric population with neuromuscular disease given the high

9
 ACCEPTED MANUSCRIPT

203 correlation between 2MWD and 6MWD. The shorter duration of the 2MWT is conducive to in-

204 dividuals with muscle weakness and fatigue. Therefore, the 2MWT could potentially be adminis-

205 tered to a larger percentage of individuals who are functionally ambulatory but have difficulty

206 walking for 6 consecutive minutes. If we are able to obtain similar results in a 2MWT as in the

PT
207 6MWT, then participants may not require frequent rest periods as permitted by the ATS or expe-

RI
208 rience frequent bouts of fatigue-related weakness. Additionally, both the 6MWD and the

209 2MWD, in our study, showed strong correlations with the functional measures, including MFM-

SC
210 32 and TFTs, valid measures shown to be demonstrative functional ability in NMD. In conclu-

211 sion, the aforementioned functional measures can be used to assess functional ability in NMD.

212
U
The 2MWT has similar utility to the 6MWT, thereby serving as an optimal measure for tracking
AN
213 patients with progressive NMD over time, yet it requires less time thus less fatiguing to the pa-

214 tient.
M

215 Acknowledgements:
D
TE

216 We thank the patients for their commitment of time and effort in participating in this re-

217 search. We are grateful to Carmel Nichols and Ching-Yi Shieh for their statistical analyses. We
EP

218 thank the Congenital Muscular Dystrophy Clinical Outcome Measures (CMD COM) Team:

219 Leslie Nelson, Bonnie Hodsdon, Rebecca Parks, Roxanna Bendixen, Michelle McGuire, Tina
C

220 Duong, Michelle Hsia, KC Keller, Donovan Lott, Irene C. Chrismer, Melody M. Linton, Jeff El-
AC

221 liott, Jahannaz Dastgir, Linda Hynan, Elizabeth Hartnett, Gilberto M. Averion, James Collins,

222 Eunice Kim, Angela Kokkinis, Veronica Hinton, Diana Bharucha, Carole Vuillerot, Anne

223 Rutkowski. We also thank Joan Austin.

224 Funding:

10
 ACCEPTED MANUSCRIPT

225 This work was supported by the NIH intramural research funds from the Clinical Center;

226 the National Institute of Neurological Disorders and Stroke (NCT1568658); and the National In-

227 stitute of Nursing Research (NCT02362425).

PT
RI
U SC
AN
M
D
TE
C EP
AC

11
 ACCEPTED MANUSCRIPT

References

[1] Andersen LK, Knak KL, Witting N, Vissing J. Two- and 6-minute walk tests assess walking

capability equally in neuromuscular diseases. Neurology 2016;86:442-5.

[2] Butland RJ, Pang J, Gross ER, Woodcock AA, Geddes DM. Two-, six-, and 12-minute

PT
walking tests in respiratory disease. Br Med J (Clin Res Ed) 1982;284:1607-8.

RI
[3] Kierkegaard M, Tollback A. Reliability and feasibility of the six minute walk test in

subjects with myotonic dystrophy. Neuromuscul Disord 2007;17:943-9.

SC
[4] Mazzone E, Bianco F, Main M, et al. Six minute walk test in type III spinal muscular

U
atrophy: a 12month longitudinal study. Neuromuscul Disord 2013;23:624-8.
AN
[5] McDonald CM, Henricson EK, Han JJ, et al. The 6-minute walk test as a new outcome

measure in Duchenne muscular dystrophy. Muscle Nerve 2010;41:500-10.

M

[6] Montes J, McDermott MP, Martens WB, et al. Six-Minute Walk Test demonstrates
D

motor fatigue in spinal muscular atrophy. Neurology 2010;74:833-8.

TE

[7] Wetzel JL, Fry DK, Pfalzer LA. Six-minute walk test for persons with mild or moderate

disability from multiple sclerosis: performance and explanatory factors. Physiother Can
EP

2011;63:166-80.

[8] Darke J, Bushby K, Le Couteur A, McConachie H. Survey of behaviour problems in

C

children with neuromuscular diseases. Eur J Paediatr Neurol 2006;10:129-34.

AC

[9] Bohannon RW, Bubela D, Magasi S, et al. Comparison of walking performance over the

first 2 minutes and the full 6 minutes of the Six-Minute Walk Test. BMC Res Notes

2014;7:269.

12
 ACCEPTED MANUSCRIPT

[10] Vill K, Ille L, Schroeder SA, Blaschek A, Muller-Felber W. Six-minute walk test versus two-

minute walk test in children with Duchenne muscular dystrophy: Is more time more

information? Eur J Paediatr Neurol 2015;19:640-6.

PT
[11] Alfano LN, Lowes LP, Dvorchik I, et al. The 2-min walk test is sufficient for evaluating

walking abilities in sporadic inclusion body myositis. Neuromuscul Disord 2014;24:222-

RI
6.

SC
[12] Brooks D, Davis AM, Naglie G. Validity of 3 physical performance measures in inpatient

geriatric rehabilitation. Arch Phys Med Rehabil 2006;87:105-10.

U
[13] Brooks D, Hunter JP, Parsons J, Livsey E, Quirt J, Devlin M. Reliability of the two-minute
AN
walk test in individuals with transtibial amputation. Arch Phys Med Rehabil

2002;83:1562-5.
M

[14] Brooks D, Parsons J, Hunter JP, Devlin M, Walker J. The 2-minute walk test as a measure
D

of functional improvement in persons with lower limb amputation. Arch Phys Med
TE

Rehabil 2001;82:1478-83.

[15] Brooks D, Parsons J, Tran D, et al. The two-minute walk test as a measure of functional
EP

capacity in cardiac surgery patients. Arch Phys Med Rehabil 2004;85:1525-30.

[16] Rossier P, Wade DT. Validity and reliability comparison of 4 mobility measures in
C

patients presenting with neurologic impairment. Arch Phys Med Rehabil 2001;82:9-13.
AC

[17] Katz-Leurer M, Rotem H, Keren O, Meyer S. The relationship between step variability,

muscle strength and functional walking performance in children with post-traumatic

brain injury. Gait Posture 2009;29:154-7.

13
 ACCEPTED MANUSCRIPT

[18] Upton CJ, Tyrrell JC, Hiller EJ. Two minute walking distance in cystic fibrosis. Arch Dis

Child 1988;63:1444-8.

[19] Kosak M, Smith T. Comparison of the 2-, 6-, and 12-minute walk tests in patients with

PT
stroke. J Rehabil Res Dev 2005;42:103-7.

[20] Pereira AC, Ribeiro MG, Araujo AP. Timed motor function tests capacity in healthy

RI
children. Arch Dis Child 2016;101:147-51.

SC
[21] Berard C, Payan C, Hodgkinson I, Fermanian J, Group MFMCS. A motor function measure

for neuromuscular diseases. Construction and validation study. Neuromuscular

U
disorders : NMD 2005;15:463-70.
AN
[22] McDonald CM, Henricson EK, Abresch RT, et al. The 6-minute walk test and other clinical

endpoints in duchenne muscular dystrophy: reliability, concurrent validity, and minimal

M

clinically important differences from a multicenter study. Muscle Nerve 2013;48:357-

D

68.
TE

[23] Laboratories ATSCoPSfCPF. ATS statement: guidelines for the six-minute walk test. Am J

Respir Crit Care Med 2002;166:111-7.

EP

[24] Berard C, Payan C, Fermanian J, Girardot F, Groupe d'Etude MFM. [A motor function

measurement scale for neuromuscular diseases - description and validation study]. Rev
C

Neurol (Paris) 2006;162:485-93.

AC

[25] Beenakker EA, Fock JM, Van Tol MJ, et al. Intermittent prednisone therapy in Duchenne

muscular dystrophy: a randomized controlled trial. Arch Neurol 2005;62:128-32.

14
 ACCEPTED MANUSCRIPT

[26] Griggs RC, Moxley RT, 3rd, Mendell JR, et al. Prednisone in Duchenne dystrophy. A

randomized, controlled trial defining the time course and dose response. Clinical

Investigation of Duchenne Dystrophy Group. Arch Neurol 1991;48:383-8.

PT
[27] Pradhan S, Ghosh D, Srivastava NK, et al. Prednisolone in Duchenne muscular dystrophy

with imminent loss of ambulation. J Neurol 2006;253:1309-16.

RI
[28] StataCorp, Stata statistical software. 2015, StataCorp LP: College Sation, TX.

SC
[29] Giavarina D. Understanding Bland Altman analysis. Biochemia medica 2015;25:141-51.

U
AN
M
D
TE
C EP
AC

15
 ACCEPTED MANUSCRIPT

FIGURES

PT
6MWD vs. 2MWD

RI
875

700 y = 2.9343x + 2.6396

6MWD (m)

SC
525

350

U
175
AN
0
0 60 120 180 240 300
2MWD (m)
M

Figure 1: 6MWD vs. 2MWD with Trend Line

This model displays each subject’s 2MWD (x-axis) and 6MWD (y-axis). The regression equa-
tion estimates the relationship between the 2MWD and 6MWD. The regression equation has a
D

slope of 2.93 and a y-intercept of 2.64.

TE

4-6 MWD vs. 2MWD

EP

250

188
4-6 MWD (m)

125
AC

63

r=0.896, p<0.01
0
0 60 120 180 240 300
2MWD (m)

Figure 2: 4-6 MWD vs. 2MWD

16
 ACCEPTED MANUSCRIPT

Correlation between distances walked in the first two minutes and last two minutes of the
6MWT.
TABLES

Table 1: Demographic Data

Study participants are grouped by diagnosis. Gender breakdown, age range, and mean age are
presented. Several additional diagnoses are represented in the “other” category, including giant

PT
axonal neuropathy, LMNA-related myopathy, ACTA1 myopathy, and SEPN1 congenital muscu-
lar dystrophy.

RI
Gender Total Mean Age (years) ± SD

Male Female

SC
RYR1-RM 10 9 19 9.7 ± 2.59

COL6-RD 9 8 17 10.7 ± 3.48

LAMA2-RD 0 3 3
U 8.1 ± 4.77
AN
LGMD 5 2 7 8.1 ± 2.65
M

Other 20 11 31 10.5 ± 2.64

Total 44 33 77 10.1 + 2.93

D

Table 2: Mean values from Outcome Measures

TE

Activity (unit) n Mean ± SD

EP

2MWD (m) 77 149.8 ± 40.3

6MWD (m) 77 442.1 ± 121.6

C

4-6 MWD (m) 73 149 ± 4.93

AC

Floor-to-Stand Time (s) 66 6.62 ± 4.54

Ascend 4 Steps Time (s) 72 4.35 ± 4.40

Descend 4 Steps Time (s) 72 3.05 ± 2.35

D1 MFM Standing & Transfers

67 69.4 ± 22.7
Score (%)

17
 ACCEPTED MANUSCRIPT

MFM Total Score (%) 67 82.3 ± 14.2

Table 3: Correlation Analysis of Functional Outcome Measures

n Correlation Coefficient (r)

PT
2MWD & Supine to Stand Time 66 -0.624

RI
6MWD & Supine to Stand Time 66 -0.620

2MWD & Ascend 4 Steps Time 72 -0.602

SC
6MWD & Ascend 4 Steps Tine 72 -0.611

2MWD & Descend 4 Steps Time 72 -0.603

6MWD & Descend 4 Steps time

U 72 -0.603
AN
2MWD & D1 MFM Standing &
67 0.773
Transfers Score
M

6MWD & D1 MFM Standing &

67 0.777
D

Transfers Score
TE

2MWD & MFM Total Score 67 0.650

6MWD & MFM Total Score 67 0.660

EP

Table 4: 6MWT velocities at each minute interval. P-values are from pairwise comparisons of
each minute interval to the first minute.
C

Minute Velocity (m/s) Std. Dev. p-value

AC

1 1.32 0.36 -

2 1.23 0.34 0.0005

3 1.24 0.35 0.0013

4 1.24 0.34 0.0003

5 1.26 0.37 0.0408

18
 ACCEPTED MANUSCRIPT

6 1.24 0.35 0.0105

SUPPLEMENTAL FILES
Supplemental file 1: 4-6 MWD and 2MWD Bland-Altman Plot

PT
RI
U SC
AN
M
D
TE

This model displays the average 2MWD (2MWD + 4-6MWD/2) on the x-axis and the difference
(2MWD - 4-6MWD) on the y-axis. The top and bottom dotted lines depict the upper and lower
limits respectively. The mean difference, 3.01 m, is shown with a solid line.
C EP
AC

19
Ann Vasc Dis Vol.5, No.2; 2012; pp151–156
©2012 Annals of Vascular Diseases doi: 10.3400/[Link].11.00098
Original Article

Diabetes-Related Lower Extremities Amputations in Saudi

Arabia: The Magnitude of the Problem
Hasan Ali Alzahrani, MD, FRCS (Glasg)

Background: Data on diabetes-related lower extremities amputations in the Kingdom of Saudi Arabia
(KSA) and perhaps in all of Middle East and North African (MENA) countries are limited, in view of
the absence of national registries.
Methods: This review aims to challenge media figures by review of data in the local database of the author,
available published data, as well as by analysis of recent annual reports of the Saudi Ministry of Health
to estimate the magnitude of the problem. Different methods of analysis are used based on the number
of beds, operations and admissions in KSA to generate approximate figures of the annual expected
numbers of amputations in KSA and MENA countries. The 2010 International Diabetes Federation IDF
comparative prevalence rate of 16.8% was used to standardize the analysis methods.
Results: Findings of 2 previous studies and 3 analytical methods led me to the prediction that about 325
amputations are likely to occur annually in Jeddah compared to 741 in Riyadh and 3970 in KSA. When
we applied the results of KSA to those of MENA countries, 44208 amputations were predicted annually.
Conclusion: Half a million diabetes-related amputations of the lower extremities are likely to occur in
KSA and MENA countries over the coming decade. National registries are urgently needed.

Keywords: diabetes mellitus, ulceration, amputation, ischemia

Introduction (Agence-France-Presse) news agency that the number of

diabetes-linked amputations is rising quickly and begin-

M ore than 50% of lower limb amputations are per-

formed on diabetic patients.1) Diabetic foot disor-
ders/ulcers (DFD/DFU) are one of the common chronic
ning to occur at younger ages. He added that some 90
diabetic amputations are performed monthly, i.e. three
per day were performed in the KSA capital of Riyadh.6)
complications of diabetes, which may lead to major and This information was published by many newspapers and
minor limb loss.2–5) These amputations are usually as- media sites on the web, sometimes without mentioning
sociated with an extended hospital stay, high mortality Al-Gannass saying that he “could not provide a figure for
and low rehabilitation rates.2,4,5) such cases.” This piece of information was continuously
On the 10th of March 2009, Dr Abdulaziz Al-Gannass, quoted by public reporters in media.
a foot and ankle surgeon at the National Guard King In view of the absence of national registry and official
Abdulaziz Medical City in Riyadh, KSA, told the AFP data for diabetes-related amputations in KSA and perhaps
in most of the Middle East and North African Countries
Department of Surgery, Faculty of Medicine FOM, King Abdu- (MENA), it may be the time to challenge the above quoted
laziz University (KAU), Jeddah, the Kingdom of Saudi Arabia figures in view of the most recent Saudi official data as
Received: December 12, 2011; Accepted: January 5, 2012 reported by the Saudi Ministry of Health and Saudi Cen-
Corresponding author: Hasan Ali Alzahrani, MD, FRCS (Glasg). tral Department of Statistics & Information,7,8) as well as
Department of Surgery, Faculty of Medicine FOM, King Abdulaziz reviewing the author’s local data-base4,5) and available
University (KAU), P.O. Box 80215, Jeddah 21589, the Kingdom of
Saudi Arabia
data published in the literature9–14) aiming to estimate the
Tel: +966-2-6408209, Fax: +966-2-6298871 magnitude of the problem and forecast the future not only
E-mail: haaz59@[Link] in KSA but perhaps in other similar countries with high

Annals of Vascular Diseases Vol.5, No.2 (2012) 151

Alzahrani H.

Table 1  Population, population at risk of having Diabetes Mellitus (DM) and number of diabetic patients
Description Jeddah Riyadh KSA MENA
Overall population ( × 103) (8) 3456 5254 27137 574115
Population at risk (Aged above 20 years) (15) 2073 3152 15186.8 344469
Number of diabetics (based on 16.8 comparative prevalence rate) (15) 348.0 529.0 2551.0 32032

KSA: Kingdom of Saudi Arabia; MENA: Middle East and North African

Table 2  Assumptive prevalence rates of amputations based on estimated numbers of diabetic patients as reported in media6)
Descriptions Jeddah Riyadh KSA MENA
Estimated number of diabetic patients at risk ( × 103) 348.0 529.0 2551 32035
Assumptive number of amputations based on 0.2% risk for diabetics 696.0 1080 5102 64070

KSA: Kingdom of Saudi Arabia; MENA: Middle East and North African

steadily increasing prevalence rates of diabetes particu- Assumption

larly Arab10) and MENA countries.15,16) Such data may be
used as an indirect indicator or marker for measuring the Based on the figure quoted by Al-Gannass6) of KSA, a
effectiveness of prevention and management provided to total of at least 1080 diabetes-related amputations (i.e. 90/
diabetics suffering from foot ulceration. This may also month) are performed annually in Riyadh, the capital of
excite policy makers to launch national registries for KSA, which has a population of 5.25 million. This means
diabetes-related amputations in KSA and MENA coun- an incidence rate of 205 amputations per million, 2.0 per
tries, which are confronting an impending health crisis. 10 000 population, or 1 amputation per 4878 of the total
population of Riyadh (1:4878, approximate incidence rate
Methods of 0.02% in the overall population). This rate of amputation
increases when only the diabetic population in adults is
Aim considered. The IDF 2010 comparative prevalence rate of
16.8% of diabetes in KSA general population was used in
To estimate the likely annual occurrences of diabetes- this study for the sake of standardization of source of Arab
related lower extremities amputations and address the and MENA country estimations Table 1. The IDF estimated
need for national registries in KSA and MENA countries. a total of 15.186 million as population at risk of having DM
(representing 60% of population above 20 years).15) This
Design percentage is almost similar to the Saudi MOH report.7)
The approximate numbers in the population at risk,
In view of the absence of a Saudi national registry for numbers of diabetics and estimated amputation rates
amputations, the author tried to verify if the previously based on the above assumption of amputation rate in
quoted figure in media was correct and generate approxi- capital of KSA are shown in Table 2 in Riyadh, Jed-
mate figures for the prevalence of diabetic amputations. dah, KSA, and MENA countries. If Riyadh has more
Different calculation formulae were used, based on the than half a million of diabetic patients for whom 1080
most recent information published by our group4,5) in amputations were performed, this means simply one
Jeddah, which is the second largest city in KSA, aiming amputation will be performed on every 490 diabetic
to predict the total number of amputations performed patients (approximately 1:500 ratio) i.e. a prevalence rate
annually, in all of KSA. Such information is needed as a of 0.2% in Riyadh diabetics’ population. Based on this
base-line indicator for effectiveness of health care services assumptive rate, one can predict that Jeddah will record
provided. It will also help in future health plans and for 690 amputations/year, KSA 5,102 and MENA countries
forecasting the future magnitude of diabetic lower extremi- 64070 amputations/year, Table 2.
ties amputations not only in KSA but also in all Arab and These above figures will be challenged below using
MENA countries, which share high prevalence rates of Jeddah’s data to reach an estimate of the annual numbers
diabetes, and common cultural and social backgrounds. of amputations expected in KSA and MENA countries.

152 Annals of Vascular Diseases Vol.5, No.2 (2012)

 Amputations among Saudi Diabetics

Table 3  Estimated incidence rates of amputations based on KAUH five years study4)
Descriptions KAUH Jeddah Riyadh KSA
Based on bed numbers Number of beds 650 4416 11635 44099
(Percentage to KSA capacity) * (1.26%) (10%) (26.4%) (100%)
Amputation: beds ratio 1:14 1:14 1:14 1:14
Number of amputation 45 306.0 650 3053
Based on operation numbers Number of Operations ** 9342 89385 184852 929403
(Percentage to KSA) (1.0%) (9.6%) (19.8%) (100%)
Amputation: operations ratio 1:208 1:208 1:208 1:208
Number of amputation 45 431.0 890 4476
Based on admission numbers Number of admissions 41227 166601 425283 3042465
(Percentage to KSA) (1.4%) (5.8%) (14.0%) (100%)
Amputation: Admissions ratio 1:916 1:916 1:916 1:916
Number of amputation 45 182.0 464 3321
Estimated average numbers of amputations 45 306 668 3616
KSA: Kingdom of Saudi Arabia; KAUH: King Abdulaziz University Hospital
* Number of beds = MOH beds + Estimate of other sectors [Other sectors = (MOH × 40%)] in Jeddah & Riyadh
** Number of operations = MOH operations + Estimate of other sectors [Other sectors = (MOH × 52.2%)] in Jeddah & Riyadh

Data Sources Estimation Methods

Data from the following sources were reviewed and an- King Abdulaziz University Hospital KAUH, Jeddah
alyzed aiming to estimate the magnitude of the problem: Study4)
• Reported prevalence rates of diabetes by IDF15) KAUH is a tertiary care general hospital and has
• Two previously published studies by the author’s all specialties and sub-specialties. It is the only public
group in Jeddah4,5) university hospital in Jeddah city, with a bed capacity of
• Annual report of Saudi Ministry of health 20097) 650, during the period of our studies (reached 702 lately),
• Saudi Central Department of Statistics & Information which represents 1.26% of the whole KSA bed capacity
CDSI 2010 national census8) 55932 beds in all sectors. In 2009,7) the total number of
surgical operations performed in KAUH was 9342 which
Results represent 1% of the total number performed in the whole
KSA 929403. KAUH contributed to all hospitals’ admis-
Saudi Population sions in KSA (3.0 million admissions) by admitting 41227
patients i.e. 1.4% of the KSA’s admissions.
According to the most recent CDSI 2010 national In a retrospective study over a five year period (January
census,8) the population in KSA reached 27136977 (in- 2004–December 2008), we reported 222 diabetes-related
cluding 8429401 expatriates). The Saudi CDSI report lower limb amputations performed in KAUH. This means
also indicated a 3.2% growth rate compared to 2004’s that approximately 45 amputations were performed on
census.8) It also showed that Riyadh’s population is around diabetic patients per annum or 1 amputation to every
5 million (5254560) compared to 3.5 million (3456259) in 14 beds in KAUH (1:14), or 1 amputation in every 208
Jeddah which is the second largest city in KSA.8) Forty surgical operations performed (1:208) or 1 amputation in
percent of the population is below 20 years of age,7,15) every 916 patients admitted to KAUH (1:916).
Table 1. Based on the above figures, comparisons will In Table 3, the previous rates are applied to Jeddah,
be done with studies done in Jeddah and correlate it to Riyadh and the whole KSA aiming to predict the number
Riyadh’s incidence rate of 0.2% risk of amputation in of amputations in Jeddah, Riyadh and the whole Kingdom.
diabetics Table 2. The figures obtained are 306, 650 and 3053 amputa-
tions respectively based on the amputation: beds ratio

Annals of Vascular Diseases Vol.5, No.2 (2012) 153

Alzahrani H.

Table 4  Estimated incidence rates of amputations based on statistics from 3 major hospitals in Jeddah in year 20085)
Descriptions KAUH Jeddah Riyadh KSA
Based on bed numbers Number of beds 2052 4416 11635 44099
(Percentage to KSA capacity) * (4.6%) (10.0%) (26.4%) (100%)
Amputation: beds ratio 1:16 1:14 1:14 1:14
Number of amputation 128 275 725 2750
Based on operation numbers Number of Operations ** 24342 89385 184852 929403
(Percentage to KSA) (1.0%) (9.6%) (19.8%) (100%)
Amputation: operations ratio 1:190 1:190 1:190 1:190
Number of amputation 128 470 972 4887
Based on admission numbers Number of admissions 72937 166601 425283 3042465
(Percentage to KSA) (1.4%) (5.8%) (14.0%) (100%)
Amputation: Admissions ratio 1:569 1:569 1:569 1:569
Number of amputation 128 292 746 5,339
Estimated average numbers of amputations 128 345 814 4325
KSA: Kingdom of Saudi Arabia; KAUH: King Abdulaziz University Hospital
* Number of beds = MOH beds + Estimate of other sectors [Other sectors = (MOH x 40%)] in Jeddah & Riyadh
** Number of operations = MOH operations + Estimate of other sectors [Other sectors = (MOH x 52.2%)] in Jeddah & Riyadh

at KAUH. These amputation numbers are higher based Discussion

on operation ratio 431, 890, and 4,476 but lower based
on admission ratio 182, 464, and 3321 respectively. The Diabetes is the global epidemic of the 21st century.
averages for these figures are 306 for Jeddah, 668 for Ri- The number of adults with diabetes in the world was
yadh and 3616 for the whole KSA. It is obvious, therefore, 135 million in 1995. At present, 366 million people are
that the current estimation of 668 is less than the quoted affected by diabetes worldwide and 80% of them live in
number of 1080 by 38.0%. low and middle-income countries.16) By 2030, the global
burden of diabetes is projected to increase to 552 million
Diabetes-related amputations in three major public people.16) Between 2010 and 2030, there will be a 69% in-
hospitals in Jeddah5) crease in the number of adults with diabetes in developing
In 2008, the amputations number performed on dia- countries and a 20% increase in developed countries.17)
betic patients in 3 major hospitals in Jeddah was 128. Data that has been accumulated over the last three
The 3 major hospitals had a total bed capacity of 2052 decades confirms an epidemic of type 2 diabetes in
(about 50% of Jeddah’s bed capacity) and admitted 72937 KSA.11) KSA is one of the top ten countries (number 6)
patients over one year period. Surgeons performed ap- for prevalence of diabetes and is expected to continue in
proximately 24342 operations in all disciplines. Calcula- the coming two decades with prevalence rate of 20.0%
tions were again done on the same line of previous study, in the age group 20–79 years.16) Other countries in the
Table 4. The averages of the three methods were 345 “top-ten” included Kuwait 21.1, Lebanon 20.2, Bahrain
amputations for Jeddah, 814 for Riyadh and 4325 for KSA. 19.9 and United Arab of Emirates 19.2 as reported by IDF
Finally, averages of three methods in both studies were recently.16) The published diabetes mellitus DM preva-
taken, and it was found that 325 amputations are estimated lence rates in KSA vary. These rates are still not well
to be performed annually in Jeddah, compared to 741 in known due to various methodologies used by research-
Riyadh and 3970 for the whole kingdom i.e. 2.5 amputa- ers.10–12,15,16) For instance Al-Nozha et al.12) conducted a
tions per 10000 population. Again, this figure is less than national survey and reported a prevalence rate of 23.7% in
the one quoted by 31%, Table 5. When applied to MENA the adult Saudi population above 30 years of age (26.2%
countries, 44208 amputations are predicted annually. in males versus 21.5% in females). In contrast, Elhadad
et al.11) reviewed all of studies and commented that the
last rate is somewhat inflated. The International Diabetes

154 Annals of Vascular Diseases Vol.5, No.2 (2012)

 Amputations among Saudi Diabetics

Table 5  Summary of estimated incidence rates of amputations based

on averages obtained by the three methods of analysis
Description Jeddah Riyadh KSA
Assumption * 696 1,080 5102
Method 1 averages ** 306 668 3616
Method 2 averages *** 345 814 4325
(Both studies averages) 325 741 3970
KSA: Kingdom of Saudi Arabia
* Table 2, ** Table 3, *** Table 4

Federation IDF in 2010 reported a comparative prevalence national registries for amputations among diabetics who
rate of 16.8% in the age group between 20–79 years.15) In represent 20% of their nations.
this study, the IDF prevalence rate of 16.8% of diabetes in It is estimated that 325 amputations are likely to be
the general population of KSA was used15) as it is more performed in Jeddah compared to 741 in Riyadh and 3970
conservative and was generated by an international body for the whole kingdom. These figures are less than the one
after a review of various published studies. It is closer to quoted in media by 31%. This means an incidence rate of
the period during which previous studies of the popula- 2.6 per 10000 people at risk. This rate is on the low side
tion of Jeddah were performed. but within the range published in industrialized coun-
Diabetes-related lower-extremity amputations rates are tries, which ranged from 0.06 to 3.83 per 10 (3) people
important indicators for the effectiveness of health care at risk.18) The figures generated in the current study are
provided to diabetic patients including prevention and for all amputations i.e. major and minor. According to
management of foot ulcers and to forecast the magnitude our previous studies, major amputations accounted for
of the problem. Unlike developed countries where data on 45%4)and 38%5) i.e. average of 41.5% are majors. Based
the incidence of amputations can be obtained relatively on the current data, it can be predicted that 1647 major
easily,18) these rates are unknown in most of developing amputations are performed annually in KSA.
countries, which record high prevalence of diabetes such Up to my knowledge, this is the first report which
as KSA and MENA countries. Amputationfigures require tried to predict the magnitude of the diabetes-related
very careful interpretation. Variations in the definition lower-extremity amputations rates in KSA and MENA
of amputation (i.e. major, above ankle; or minor, distal countries and to challenge the figures quoted in media.
to ankle), selection of the population, and the choice of However, it is still a “snapshot” on a complex health
numerator and denominator make it difficult to perform problem. However, there are significant limitations in
comparisons. In addition to disease severity, many as- this exercise. Firstly, it is based on the two studies of the
pects of the structure of care services affect this decision, author’s group. One of the two studies4) was a retrospec-
including access to primary care, quality of primary care, tive single center study, and the second,5) a multi-central
delays in referral, availability and quality of specialist study. However, it was not as comprehensive as the others
resources, and prevailing medical opinion. It follows that because it was conducted in only three major hospitals in
a high incidence of amputation can reflect a higher disease Jeddah. Secondly, amputation figures were generated by
prevalence, late referral, limited resources, or a particularly diverse methods, which included amputations per hospi-
interventionist approach by a specialist team.18) In KSA tal beds, per operations and per admissions. The second
and perhaps most of MENA countries cultural, social and reason may be a point of strength rather than weakness
individual patient factors play a major role on the ampu- as using multiple methods of calculations will reduce
tations’ decision. This makes cross comparison between the possibility of bias when only one variable is used.
different data more complicated. However, amputations’ Finally, whatever is done will not replace the accurate
rates will definitely help as a base-line indicator19) in the data collection in a prospective setup.
same country or similar countries that share racial and Intended future registries will hopefully establish, not
cultural backgrounds, and health care systems, such as only the overall future trends of amputation in general, but
those of MENA countries. The current study aims to raise also the high-low amputation ratio, i.e. the major: minor ratio,
the level of awareness in these countries about the need for which is important in proper evaluation of improvements

Annals of Vascular Diseases Vol.5, No.2 (2012) 155

Alzahrani H.

in care organization and multidisciplinary-centered proto- article/143. Accessed 17 November 2011.

cols.19,20) Similar to developed countries21) such registries 9) Qari FA, Akbar D. Diabetic foot: Presentation and
treatment. Saudi Med J 2000; 21: 443-6.
will give opportunities for cross-national learning to
10) Al-Wahbi AM. The diabetic foot. in the Arab world.
improve health system performance. It will also help in as- Saudi Med J 2006; 27: 147-53.
sessing the outcomes of newer modalities of treatment.22–24) 11) Elhadd TA, Al-Amoudi AA, Alzahrani AS. Epidemi-
This work shows the difficulty and shortcomings of ology, clinical and complications profile of diabetes
scattered data in the literature and will support the timely in Saudi Arabia: a review. Ann Saudi Med 2007; 27:
241-50.
need for accurate data-base sources. United Nations 12) Al-Nozha MM, Al-Matouq MA, Al-Mazrou YY, et
Resolution 61/225, has already adopted unanimously in al. Diabetes in Saudi Arabia. Saudi Med J 2004; 25:
2006, recognizes diabetes as a serious and costly disease 1603-10.
that poses a threat to individual well-being and economic 13) Al-Tawfiq JA, Johndrow JA. Presentation and outcome
of diabetic foot ulcers in Saudi Arabian patients. Adv
progress, especially in low- and middle-income coun- Skin Wound Care 2009; 22: 119-21.
tries.25) It also addressed the problems of data accuracy 14) Al-Mahroos F, Al-Roomi K. Diabetic neuropathy foot
in less-developed countries by stating “In view of absence ulceration, peripheral vascular disease and potential
of transparency, public health services, national registries, risk factors among patients with diabetes in Bahrain:
national diabetes programs and national well-designed A nationwide primary care diabetes clinic-based study.
Ann Saudi Med 2007; 27: 25-31.
community studies, most of the available published data 15) International Diabetes Federation (2010). [Link]
including those published by recognized international [Link]/content/mena-data. Accessed 17
bodies are not accurate”.25) November 2011.
In conclusion, half a million of MENA country citi- 16) International Diabetes Federation (2011). [Link]
[Link]/diabetesatlas/5e/the-global-burden. Accessed
zens are likely to have a minor or major diabetes-related 18 November 2011.
amputations according to the figures generated in the 17) Shaw JE, Sicree RA, Zimmet PZ. Global estimates of
current study of KSA. Most of the amputees will not be the prevalence of diabetes for 2010 and 2030. Diabetes
rehabilitated and may lose their jobs or incomes. In view Res Clin Pract 2010; 87: 4-14.
18) Jeffcoate WJ, van Houtum WH. Amputation as a
of this, national registries are urgently needed in KSA
marker of the quality of foot care in diabetes. Diabe-
and MENA countries. tologia 2004; 47: 2051-8.
19) Rogers LC, Bevilacqua NJ. Organized programs to
References prevent lower-extremity amputations. J Am Podiatr
Med Assoc 2010; 100: 101-4.
1) Al Zahrani HA, Saban SA, Merdad HT. Management 20) López-de-Andrés A, Martínez-Huedo MA, Carrasco-
of diabetic foot ulcer. Asian J Surg 1991; 14: 24-7. Garrido P, et al. Trends in lower-extremity amputa-
2) Al Zahrani HA, Ghandourah NM, Merdad HT. Limb tions in people with and without diabetes in Spain,
amputations in Western Saudi Arabia. Asian J Surg 2001–2008. Diabetes Care 2011; 34: 1570-6.
1992; 15: 119-22. 21) Squires DA. The U.S. health system in perspective:
3) Al Zahrani HA, Al Bar HM, Bahnassi A, et al. The a comparison of twelve industrialized nations. Issue
distribution of peripheral arterial disease in a defined Brief (Commonw Fund) 2011; 16: 1-14.
population of elderly high-risk Saudi patients. Int 22) Hayashi H, Yamada S, Kumada Y, et al. Immersing
Angiol 1997; 16: 123-8. feet in carbon dioxide-enriched water prevents expan-
4) Badri MM, Tashkandi WA, Nawawi A, et al. Extremi- sion and formation of ischemic ulcers after surgical
ties amputations over five years 2005–2009 in King revascularization in diabetic patients with critical limb
Abdulaziz University Hospital, Jeddah, Saudi Arabia. ischemia. Ann Vasc Dis 2008; 1: 111-7.
JKAU Med Sci 2011; 18: 13-25. 23) Nishida T, Ueno Y, Kimura T, et al. Early and long-
5) Tashkandi WA, Badri MM, Badawood SM, et al. term effects of the autologous peripheral stem cell
Lower limb amputation among diabetics in three major implantation for critical limb ischemia. Ann Vasc Dis
hospitals in Jeddah, Saudi Arabia. JKAU Med Sci. 2011; 4: 319-24.
2010; 18(1): 23-35. 24) Tara S, Miyamoto M, Takagi G, et al. Prediction of
6) Arabian Business (2009). [Link] limb salvage after therapeutic angiogenesis by autolo-
[Link]/riyadh-sees-90-diabetes-amputations- gous bone marrow cell implantation in patients with
[Link]. Accessed 17 November 2011. critical limb ischemia. Ann Vasc Dis 2011; 4: 24-31.
7) Saudi Ministry of Health (2010) [Link] 25) United Nations (2007). [Link]
sa/statistics/[Link]. Accessed 17 November 2011. journal/asp/[Link]?m=A/RES/61/2 Accessed on 19
8) Saudi Central Department of Statistics and Informa- November 2011.
tion (2010) [Link]

156 Annals of Vascular Diseases Vol.5, No.2 (2012)

J Chron Dis Vol. 38. No. 1. pp. 27-36, 1985 0021-9681:85$3.00-~0.00
Printed in Great Britain. All rights reserved Copyright I(~’1985 Pergamon Press Ltd

A METHODOLOGICAL FRAMEWORK FOR

ASSESSING HEALTH INDICES

BRAM KIRSHNER and GORDON GUYATT

McMaster University Health Sciences Centre. 1200 Main St West, Hamilton, Ontario,
Canada L8N 325

(Receirrrl in recised~brm 13 Auyust 1984)

Abstract-Tests or measures in clinical medicine or the social sciences can be used for three
purposes: discriminating between subjects, predicting either prognosis or the results of some other
test. and evaluating change over time. The choices made at each stage of constructing a quality of
life index will differ depending on the purpose of the instrument. We explore the implications of
index purpose for each stage of instrument development: selection of the item pool. item scaling.
item reduction. determination of reliability, of validity, and of responsiveness. At many of these
stages. not only are the requirements for discriminative. predictive. and evaluative instruments not
complementary. they are actually competing. Attention to instrument purpose will clarify the
choices both for those developing quality of life measures and for those selecting an appropriate
instrument for clinical studies.

INTRODUCTION

A HOST of quality of life indexes have recently been developed to measure complex
domains like emotional and social function, well-being, disability, and overall health
status. Although methodological standards for quality of life indexes exist [la], their
usefulness is limited by failure to clearly distinguish between the possible uses of these
instruments in clinical practice and in research. The result is confusion as to the best way
to construct and validate quality of life measures.
In response to this problem, we present in the following discussion a methodological
framework for the development and assessment of health status measures which
emphasizes the specific purpose for which the instrument will be used. Whilst this
framework can be applied to any measuring instrument, whether of hemoglobin, blood
pressure, or emotional function, we believe it will be most useful for clarifying issues
related to the construction and validation of quality of life measures.

THREE PURPOSES OF HEALTH STATUS MEASURES

The potential applications of health status measures can be divided into three broad
categories: discrimination, prediction and evaluation.

I. DiLscrimirzatice inde.u
A discriminitive index is used to distinguish between individuals or groups on an
underlying dimension when no external criterion or gold standard is available for
validating these measures. Intelligence tests, for example, are used to distinguish between
children’s learning abilities. The Minnesota Multiphasic Personality Inventory was
developed in order to distinguish those with emotional and psychological disorders from
the general population [S]. One of the most promising uses of discriminitive functional
status measures is in surveys which attempt to quantify the burden of illness across
different communities [6, 71.

2. Predictive [Link]
A predictive index is used to classify individuals into a set of predefined measurement
categories when a gold standard is available, either concurrently or prospectively, to
27
28 BRAM KIRSHNERand GORDON GUYAT-I

determine whether individuals have been classified correctly. This type of index is
generally used as a screening or diagnostic instrument to identify which specific
individuals have or will develop a target condition or outcome.
Given that a gold standard is available, one might ask why an investigator would go
through the trouble of developing a new measuring instrument which, at best, could only
be as good as one which already exists. The answer, hopefully, is that the index has
something to offer which the gold standard does not. The index may be less risky,
uncomfortable or costly, or applicable earlier in the course of disease. We may, for
example, use a simple functional status measure to determine which elderly people require
additional social services in the home, the gold standard being a detailed assessment of
need or the response to the intervention. Simple functional status measures have been
shown to be among the most accurate predictors of prognosis in cancer patients [S]. The
Denver Developmental Screening Test is designed to identify children who are likely to
have learning problems in the future [9].

3. Eduutice indexes
An evaluative index is used to measure the magnitude of longitudinal change in an
individual or group on the dimension of interest. The development of evaluative
instruments has provided the main focus for those interested in measurement of quality of
life. Such instruments are needed for quantitating the treatment benefit in clinical trials,
and for measuring quality adjusted life years in cost-utility analyses. A number of
investigators have developed indexes designed for general populations [2, 10P14]. More
recently, the need for disease specific quality of life questionnaires for use in clinical trials
has been recognized, and led to construction of instruments for studies in cancer [lS] and
in rheumatology [16, 171.

PURPOSE SPECIFIC INSTRUMENTS: IMPLICATIONS FOR INDEX DEVELOPMENT

In clinical practice, the introduction of new tests or measures often occurs without
scientists feeling the need, when assessing an instrument’s usefulness, to focus on the
specific purpose of the test. The reason is that the process of development of most clinical
tests, and the requirements for successful application, are identical for discrimination,
prediction and evaluation. For example, the measurement of cardiac output can be used to
tell us who has normal cardiac function and who doesn’t; who is likely to experience heart
failure in the future; and whether someone’s heart is working better or worse than it did in
the past. The same is true for most of the wide variety of laboratory or physiological tests
upon which clinicians rely. It is not true, however, when we look at instruments for
measuring quality of life. In fact, as we shall point out, the requirements for maximizing
one of the functions of discrimination, prediction, or evaluation may actually impede the
others. As a result, validating the index for one of these purposes does not necessarily
ensure that it can be used for the remaining ones.
The following steps are involved in constructing an index to measure quality of life:
-selection of the item pool
-item scaling
-item reduction
--determination of reliability
-determination of validity
determination of responsiveness
To illustrate the different methodological issues involved in developing a quality of life
index for each of these three purposes, we shall use an example pertaining to the
measurement of functional status (by which we mean the way a person feels, and how he or
she functions in day-to-day activities) in patients with disabling chronic cardiorespiratory
disease. We shall look at three investigators who wish to construct an index to measure
functional status in patients with chronic cardiorespiratory disease. The first, whom we
shall call D, (for discrimination) has the impression that functional status is related as
much to personality variables, such as motivation and outlook on life, as it is to
 A Methodological Framework for Assessing Health Indices 29

physiological dysfunction. To test this hypothesis requires a measure of functional status

that would discriminate between individuals; D has taken on the task of constructing this
instrument. The second investigator, P, (for prediction) wishes to predict mortality in his
patient population, and thinks that at least some aspects of functional status may be
powerful clues to prognosis. The third investigator, E, (for evaluation) wants to know if the
treatments being used improve patients’ functional status and needs to develop an
instrument which will evaluate the ability of an intervention to change the functional
status of patients with chronic cardiorespiratory disease. We shall examine the steps
involved in index construction for each of the three investigators in order to see how their
approaches might differ. A summary of the major issues in index construction for each
step is listed, according to the purpose of the instrument, in Table 1.

Diacrimmative criteria Predictive criteria Evaluative criteria

Item SelectlO” -tap important components of the -statistical associatmn with tap areas related to change in
domam criterion measure health status
--universal apphcablhty to -responsiveness to clinically
respondents significant change
-stability over time

Item scaling -short response sets which -response sets which maximize -response sets with sufficient
facilitate uniform mterpretation correlations with the crltermn gradations to register change
[Link]

Item reduction -internal scaling or consistency -power to predict YS respondent -responsiveness vs respondent
--comprehensiveness and burden burden
reductmn of random error vs
respondent burden

Rellabllity large and stable mtersubject -stable Inter and intrasubject -stable intrasubject variation:
variation: correlatmn between variation: chance corrected inslgmficant variatmn between
replicate measures agreement between rephcate rephcate measures
measures

Validity -cross-sectIonal construct validity: <riterion validity: agreement -longitudinal construct validity:
relationship between index and with crlterion measure relationship between changes in
external measures at a single Index and external measures over
point in time time

Responsiveness not relevant -not relevant power of the test to detect a

clinically important difference

SELECTION OF THE ITEM POOL

When constructing a questionnaire, the first task is to bring together a set of items
which might plausibly be included in the final instrument. These items can be selected in a
number of ways: the investigator may use personal judgement, tap the clinical experience
of his colleagues, consult the relevant literature, or ask patients with the disease how it
affects their lives. Regardless of which route is taken, the investigator must begin with
some criteria for inclusion of individual items in the initial pool.
Investigator D, whose goal is to discriminate between patients with cardiorespiratory
disease according to their functional status, will want to focus on features of daily life in
which disability related to underlying disease has a substantial impact. Items which are
important to patients, and which are stable over at least short periods of time will be
chosen. This latter criterion is necessary if the final index is going to be consistent in how
it distinguishes between subjects according to their functional status. Selection will also
focus on items that are performed by virtually all of the subjects at a particular functional
level. For example, ability to climb stairs would not be a desirable item if 50% of the
respondents never had occasion to negotiate a flight of stairs: these patients would be left
without a rating, and could not be compared with the rest of the population.
Investigator P would need only a single criterion for item selection: the relation between
a particular item and subsequent mortality. Unsure about which items fulfill this criterion,
he may begin with a large pool of varied items chosen because they are common or
important to patients. It is, however, items that can predict mortality that are being
sought, regardless of whether or not they are causally related to this outcome.
The third researcher, E, must apply a unique criterion: the likelihood that patient status
on a particular item will change with application of one of the available interventions. For
example, if the questionnaire was designed for use in trials of bronchodilators, emotional
aspects of functional status would have little place in the index because variability in
emotional functional would show up as noise which might obscure a treatment effect. On
the other hand. if an investigator wished to measure the effect of a comprehensive
rehabilitation program, including counselling and group therapy, on functional status,
emotional items would be important. The degree of dependency in living accommodations
(e.g. community, group home, or institution) which could be a crucial element of the
discriminative instrument, and may well be important to the predictive index, would be
rejected by investigator E, because therapy is unlikely to change the living arrangements
of such patients.
The procedure for achieving comprehensiveness is different when selecting an item pool
for an evaluative instrument than for either a descriminative or predictive tool. We
mentioned that one would want each item to be answered by most respondents who
complete a discriminative questionnaire; the same need not be true of an evaluative index.
To cite our previous example, while climbing stairs may be involved in the day-to-day
activities of only 50 ‘I0 of the population, it may be a crucial chore for those who live in a
two-storey home with bedroom and bathroom on the upper level. If an intervention made
it substantially easier for this subsample to climb stairs (even if it had no other major
effect), then it may be considered worthwhile. Evaluative instruments must, therefore,
include a strategy for measuring all clinically important treatment effects.

ITEM SCALING

By scaling, we mean the options available for patients in answering each question; that
is, the scale associated with each item. The most simple scale possible is one in which the
respondent is offered two choices for each item: he either does perform a particular
activity, or he doesn’t. Such a scale would look very attractive to D. First, this strategy of
response options is necessary if one wishes to use cumulative scaling. Second, different
patients may mean different things by each of, say. “very mild, mild, moderate, severe, or
extremely severe shortness of breath.” When one is trying to discriminate between people,
as D is, then this variability in interpretation is problematic. Such interpretive problems
are likely to be substantially less if the response options are, “no problem with shortness of
breath going up stairs,” vs “problems with shortness of breath going up stairs.”
In P’s case, the issue at hand is to identify item scales which will maximize correlations
with the criterion measure. In this instance, it is difficult to generalize as to the best scaling
procedure, since the number of response categories may affect the correlation differently
from item to item, or instrument to instrument.
The requirements for Investigator E’s index are quite different from those of D and P.
He must ensure that his instrument is responsive (sensitive to change), i.e. individual items
must show changes in score when clinically important improvement or deterioration
occurs. D’s two-option scale suffers badly from lack of responsiveness. If a patient has
difficulty with shortness of breath on climbing stairs, the only way an intervention will
change his score is if it eliminates shortness of breath altogether. However, if E’s index is
to be a success, it must register any clinically important decrease in distress. Up to a point
(and no one is yet sure of what that point is), increasing response options on a scale wil
increase item responsiveness. Therefore, E may choose 5, 7 or 9-point scales, or perhaps
opt for a visual analogue scale; the investigator will certainly stay clear of yes-no
questions. It should also be noted that E is not concerned with between-person differences,
but in measuring within-person change over time.
 A Methodological Framework for Assessing Health Indices 31

ITEM REDUCTION AND INTERNAL CONSISTENCY

An investigator may initially choose items from the total item pool for the first version
of the questionnaire based on item frequency and importance (as assessed by patients or
health providers), but the content of the definitive instrument must depend on item
performance in the setting in which the index will ultimately find its use. Therefore,
performance criteria must be established so that items which do not contribute to, or
actually detract from, the usefulness of the instrument can be deleted.
Investigator D will be satisfied with the questionnaire if it discriminates between people
according to their functional status. If the total questionnaire is to achieve this purpose,
the individual questions must have this discriminative ability. Therefore, questions to
which most or all of the respondents give similar or identical answers are of no use.
Idiosyncratic items in which patients who by other criteria have a low functional status
perform well and visa versa, must be excluded. Finally, the researcher must discover the
items in which the most of the between-person variance is accounted for by factors other
than cardiorespiratory functional status. For example, suppose it is found that while
people respond quite differently when asked about the difficulty they have climbing stairs,
the amount of difficulty they report bears little relation to their functional status judged by
other criteria. Perhaps their answer is related more to the grade of the stairs they
encounter, the speed at which they attempt to mount stairs, or what they generally find at
the top, than to the severity of their disease. If this was the case, the question should be
removed from a discriminative instrument intended to distinguish between the patients
according to functional status.
There are a number of ways of ensuring that items retained for the final discriminative
index conform to these standards. One very powerful method is to ensure that the final
instrument meets cumulative scaling criteria. Cumulative scales are those that order items
by the degree of severity of limitation they describe and for which only one pattern of
responses (across items) is associated with each scale level [lg].
For example, suppose a questionnaire has three items asking about ability to run a mile,
walk a mile, and walk a block. If this questionnaire fulfills cumulative scaling criteria, then
anyone who said they could run a mile would also report being able to walk both a mile
and a block. If only one affirmative answer was given, it would always be to the question
about walking a block, and a second positive response would always be to the query about
ability to walk a mile. Thus, in a cumulative scale the patients’ score, a single number, tells
you how they answered every question on the test. While often difficult to construct, and
for some areas (such as emotional function) virtually impossible, satisfaction of cumulative
scaling criteria solves a number of problems for discriminative instruments. It insures that
each question is in fact able to discriminate between individuals, that there are no
idiosyncratic items in the questionnaire, and that items bear a fixed relation to one
another. Further, satisfaction of cumulative scaling criteria eliminates the need for
weighting the importance of each question, and provides very strong evidence that the
between-person variability in responses is related to status on a uniform underlying
dimension.
Where cumulative scaling is not possible investigators can employ other statistical
techniques to achieve item reduction. Procedures designed to measure internal
consistency such as those by Kuder and Richardson 1191 and Cronbach [20] can be used
to select the group of items which will maximize the precision of the instrument to
measure a given construct. These internal consistency measures are based on the number
of items included in the index, as well as the set of correlations among them: hence, they
provide good estimates of measurement error attributable to inappropriate and/or
inadequate sampling of the content domain. Another advantage of these measures is that
they can be used to show the point at which the inclusion of additional items in the index
will not add substantially to the instrument’s precision. Investigator D, who wishes only to
be able to discriminate between people according to functional status, may well proceed
with item reduction by deleting items which fail to meet the requirements of either
cumulative scaling or internal consistency, since inclusion of these items would decrease the
32 BIIAM KIKSHNER and GORDON GUYATT

power of the index to discriminate between patients with cardiorespiratory disease

according to their functional status. In considering how many items to delete, investigator
D will want to take respondent burden (the time and effort required by the subject to
complete the questionnaire) into consideration.
While procedures for measuring internal consistency are often used to refine and test
evaluative indexes, one can question the rationale for doing so. Measures of internal
consistency such as the KR20 and Cronbach’s alpha are based on the assumptions that the
precision of the index will increase incrementally with the co-variance of the items and the
number of items included. While this is true of a discriminative index, it is not true of an
evaluative one. To begin with, items in an evaluative index need not be correlated at a
single point in time but rather, must be consistent in the way they measure change in
health status over two points in time. Second, maximizing the number of items in the
index, regardless of whether they are correlated, can increase the likelihood of including
items that are insensitive to efficacious treatment. Besides unnecessarily increasing the
respondent burden, any variability in items in which change in score is not related to
therapy will contribute to the instrument’s random error, and may thus obscure any
important treatment effects which do occur.
For E, it is imperative that his index is responsive to all clinically important differences
in functional status. This requires identification and deletion of unresponsive items. One
way in which E could proceed with this task is to administer the preliminary
questionnaire to a group of patients with chronic cardiorespiratory disease before and
after an intervention with known efficacy in improving functional status. Responsive items
could be retained for subsequent use and items whose score failed to change could be
expunged. E must also be careful to ensure that item responsiveness is not idiosyncratic to
a specific intervention or subpopulation, and as a result may wish (before deleting any
items) to repeat this experiment in more than one subgroup with more than one
intervention.
Investigator P needn’t be concerned with internal consistency, cumulative scaling, or
item responsiveness. All that is needed is patience, for the questionnaire must be
administered to a group of people who are then followed over time to see who dies and
who remains alive. Items that fail to discriminate between the two groups can be
discarded, and those with predictive ability retained. Further, to maximize the power of
the instrument, answers to each of the items can be regressed against mortality, and
weights provided for each item that will allow for the best prediction of ultimate outcome.

RELIABILITY

Reliability of a single application of a measurement instrument can be defined as the

ratio of the variance attributable to true differences among subjects to the total variance
121,221.
Reliability can be measured by serial administration of a test to a group of subjects
whose status on the dimension being examined is believed to be stable. The issue of
reliability, however, is different for each of the three investigators.
Investigator D is concerned with demonstrating that the difference observed between
individuals are reproducible over two or more points in time-that between-person
variation will remain constant. This can usually be assessed with correlation coefficients.
While these coefficients do not take account of systematic changes in the scores, such
change may not be important (for example, if everyone’s score rose by half a standard
deviation). This is because scores from descriptive indexes often do not have meaning in
themselves. Rather, they are usually interpreted in terms of referent values (e.g. the average
and standard deviation) or their place in a distribution. For example, it is common
practice to standardize the scores from intelligence tests on a given population and to
interpret individual scores in terms of percentiles.
In the case of predictive indexes, the scores or measurement categories are already
meaningful in that they relate to one or more target conditions. In this instance, reliability
pertains to agreement between replicate observations-whether the index classifies each
 A Methodological Framework for Assessing Health Indices 33

individual into the same category at time 1 and time 2. From investigator P’s viewpoint,
the usefulness of his index is based on its ability to give predictions which are more
reliable and accurate than those which would be obtained by chance alone. Therefore, he
concludes that reliability should be demonstrated with a chance-corrected measure of
agreement, which takes into account both systematic change and random error.
As far as reliability is concerned, investigator E is only interested in whether replicate
observations on each individual remain stable over time-that is, that magnitude of the
within-person variance is small. It is entirely possible that small changes in the within-
person variation may result in poor between-person reliability-this would not be of
concern to E, provided the degree of within-person variation remained insignificant from
both a statistical and clinical standpoint.
To illustrate this point, consider how the three investigators would react if the data
depicted in Table 2 represented the results obtained from test-retest procedures for each of

TABLE 2. TEST-RFTEST R~LMR~L~TY

Patients Time 1 Ttmc 2

I 15* 14
2 14 15
3 15 14
4 14 I5
5 IS 14
b 14 15
7 I.5 14
8 14 15

*score IS out of 20.

their instruments. Although within-person variation is small, the reliability of the

instrument is very poor. In fact, because the between-person variance is zero, the value of
the product moment correlation coefficient and the intraclass correlation coefficient
(which describes reliability as the ratio of between-subject variance to total variance, and
can be used as a parametric analogue of the chance-corrected measure of agreeement
Kappa) [23] both have values of 0.
The product moment correlation coefficient and the intraclass correlation coefficient
reflect the uselessness of the instrument depicted from the viewpoints of both investigators
D and P. D is trying to discriminate between subjects according to their present functional
status, P according to their risk of subsequent death. The index whose results are depicted
in Table 2 won’t be able to discriminate between patients at all, because there isn’t any
between-patient variability.
The results of the reliability study may not appear so bleak to E, who is not interested in
measuring between-person variation, but only within-person changes over time. Let us say
that E subsequently presents evidence that his index can accurately measure the degree of
intra-subject variation in functional status across different points in time when change
does occur. Because the data in Table 1 demonstrate small intra-subject variance at a time
when functional status is not changing, it is then likely that the index will be a useful
outcome measure for clinical trials.
The message here is that a descriptive index, in order to be useful, must have a large and
stable between-subject variation. Predictive indexes must meet an additional criterion:
there must not be any systematic change in subject scores over time. For evaluative
indexes, small within-subject variance in stable subjects and a large change in score when
functional status improves or deteriorates are the factors of importance.

VALIDITY
Criterion validity refers to the extent to which a measuring instrument produces the
same results as a gold standard, or criterion measure [24&26]. Investigator P can make use
of criterion validity, for mortality is a relatively easy outcome to measure. If his
questionnaire is indeed able to forecast mortality, its validity is established. Investigators
34 BKAM KIKSHN~I< and Gowm G~,YAIT

D and E have a more difficult task, for a gold standard to measure functional status in
cardiorespiratory patients is not available. Content validity refers to the completeness
with which an index covers the important areas of the domain which it is attempting to
represent [24426]. While we will not dwell on methods for demonstrating content validity,
it is worth mentioning that the domains that D and E are trying to sample are different. D
is attempting to sample all important, relatively stable aspects of functional status
common to most members of each functional class. while E’s domain is restricted to
salient activities and feelings which are subject to clinically important change with
treatment.
Construct validity is concerned with the extent to which a particular measure relates to
other measures in a manner which is consistent with theoretically derived hypotheses
concerning the concepts (or constructs) that are being measured [24-261. To demonstrate
construct validity. D might want to show that patients with poorer cardiac and
respiratory function by physiological measurement score lower in functional status; that
questionnaire score is related directly to exercise capacity; and that global ratings of
quality of life by patient, relative. and health worker bear a close relation to results of the
new index. All these measurements would be undertaken at a single point in time. and it is
the cross-sectional difference between subjects on each of these tests that would be
correlated.
Investigator E might want to use the same physiological, exercise. and global rating
methods to validate the index, but it would not be between subject differences at a single
point in time that would be examined. Rather. it would be important to show that
longitudinal within-subject changes in index scores with an intervention bore the expected
relation to changes in the other variables measured.
This point has been neglected in the health status measurement literature to date. To
cite a number of examples: Bush and colleagues [3. lo]. Parkerson and associates [Z], and
Spitzer rt ul. 1151, constructed quality of life measures for, respectively, the general
population, family practice populations, and cancer patients. Each investigative team
indicated that their instrument was intended for use as an outcome measure in clinical
trials. The first two groups demonstrated expected correlations between questionnaire
results and variables such as age, socio-economic status, and a number of chronic medical
conditions. Spitzer and colleagues provided evidence of the validity of their QL-Index by
demonstrating substantial correlations between a number of ratings of quality of life by
patients, health workers and relatives, and scores on their new questionnaire. They also
showed that the QL-Index could discriminate between groups of healthy people, those
with cancer or other chronic diseases, and the seriously ill. This data can appropriately
lead to the conclusion that the QL-Index has been validated for distinguishing between
groups of healthy people, those with chronic disease and the seriously ill: however, not one
of the three instruments has been validated for measuring within-person change in the
setting of a clinical trial. It could be argued that instruments that adequately discriminate
between people are likely to be valid for measuring within-person change over time as
well. While this may be true, it is not necessarily so.
Therefore, for the most convincing, or definitive. demonstration of the validity of an
evaluative instrument, its relation to other measures must be examined prospectively in a
setting in which change over time is measured.

RESPONSIVENESS

While demonstration of reliability and validity is sufficient for concluding that an

instrument is useful for descriptive or predictive purposes, we must also have information
about an evaluative instrument’s responsiveness before we can confidently use it as an
outcome measure in a clinical trial. This issue concerns the power of the index to detect a
difference when one is present -that is, the sample size required to observe a small,
medium, or large change or effect size in the population.
 A Methodological Framework for Assessing Health Indices 35

There are a number of strategies for assessing responsiveness. These include ensuring
that scores improve with application of a treatment of known efficacy, and use of the index
in a clinical trial followed by examination of change scores in those who by other criteria
improved or deteriorated.
Regardless of which method is chosen, the investigator must come to grips with the
effect of random and systematic error on the power of the test. When large sources of
measurement error are present, then beta error is increased and consequently the power of
the test is decreased. If the power of the test turns out to be too low, requiring an
unattainable N to observe a desired effect size at a given level of alpha, then the index is
not useful as an evaluative instrument.

SUMMARY

The tests we use in clinical practice and research have three basic purposes: to
discriminate between individuals along a continuum of health, illness or disability; to
predict outcome or prognosis; and to evaluate within-person change over time. While for
most testing procedures, the prerequisites for each role are complementary, in the case of
quality of life measures the requirements may not only be independent, but competing.
Therefore, those who wish to develop new instruments in this area should have their
specific goal clearly in mind and tailor approaches to item selection, scaling, reduction,
and assessment of reliability and validity to their primary purpose. Likewise, those who
wish to use an existing index in clinical practice or research should ensure that the
instrument has confirmed reliability, validity, and, if necessary, responsiveness, for the
specific use they are intending to make of it.

Ackno~,l~dgc~nz~nt-~This work was partially supported by the Medical Research Council of Canada and the
Ontario Ministry of Health.

REFERENCES
I. Sackett DL. Chambers LW. McPherson AS <>Ial: The development and application of indices of health:
general methods and a summary of results. Am J Pub1 Health 67: 423428, 1977
2. Parkerson GR. Gehlback SH, Wagoner EH et al: The Duke-UNC health profile. Med Care 19: 806-828,
1981
3. Kaplan RM. Bush JW. Barry CC: Health status: types of validity and the index of well being. Health Ser Res
I I: 47X 507. 1976
4. Bombardier C, Tugwell P: A methodological framework to develop and select indices for clinical trials:
statistical and judgmental approaches. J Rheumatol 9 (Suppl. 8): 169-172. 1982
5. Hathaway S. McKinley C: Minnesota Multiphasic Personality Inventory. New York Psychological
Corporation. 195 I
6. Maddocks CL. Kerasic RB: Planning Services for Older People: Translating National Objectives into
Effective Programs. Durham, N.C.: Center for the Study of Aging and Development. 1976
7. National Center for Health Statistics: Health Interview Survey Procedures. 1957-1974. Rockville, Md:
National Center for Health Statistxs, 1975 (Vital and health statistics. Series 1, No. 11) (DHEW publication
no. (HRA)75-1311)
8. Sorna GP, Hulmes EC. Petrovich Z: Lung cancer. In Cancer Treatment. Haskell CM (Ed). Philadelphia:
WB Saunders. 1980. p. 197
9. Frankenburg WK, Goldsmith AD, Camp BW: The revised DDST: its accuracy as a screening instrument. J
Paediatr 79: 98X-995. 1971
IO. Kaplan RM. Bush JW: Health-related quality of life measurement for evaluation research and policy
analysis. Health Psycho1 1: 61-80. 1982
Il. Spitrer VJO, Sackett DL, Sibley JC et al: The Burlington randomized trial of the nurse practitioner. N Engl
J Med 290: 25 l-256. 1974
12. Bergner M. Bobbit RA, Carter WB et crl: The sickness impact profile development and final revision of a
health status measure. Med Care 19: 787 %OS, 198 I
13. Ware JE. Brook RH. Davies-Avery A rr al: Conceptualization and measurement of health for adults in the
health insurance study: Model of Health and Methodology. Rand Corporation. May 1980, p. 1
14. Hunt SM. McKenna SP, McEwan J et al: A quantitative approach to perceived health status: a validation
study. J Epid Common Health 34: 281-286, 1980
15. Spitzer WO. Dobson AJ, Hall J et al: Measuring the quality of life of cancer patients. J Chron Dis
34: 585 597.1981
16. Meenan RF. Gertman PM, Masen JH: Measuring health status in arthritis: the arthritis impact
measurement scales. Arthritis Rheum 23: 146 152, 1980
36 BKAM KIRSHNER and GOKDONGWATT

17. Fries CF. Spitr TR. Kraines RG rt crl: Measurement of patient outcome in arthritis. Arthritis Rhenm
23: 137-145. 1980
18. Guttman L: A basis for scaling qualitative data. Am So&l Rev 9: 1399150, 1944
19. Ruder GF, Richardson MW: The theory and estimation of test reliability. Psychometrika 2: 151 160, 1937
20. Cronbach LJ: Co-efficient alpha in the internal structure of tests. Psychometrika 16: 297 334. 1951
21. Wmer BJ: Statistical Principles in Experimental Design. 2nd edn. Kogakuska: McGraw-Htll 1971.
pp. 283- 287
22. Guilferd JP: Fundamental Statistics in Psychology and Education. 4th edn. Toronto. 1965. pp. 438 442
23. Kramer MS. Feinstetn AR: Clinical biostatistics LIV: the biostatistics of concordance. Clin Pharm Ther
26: 111-123. 1981
24. Nunally JC: Psychometric Theory. New York: McGraw Hill. 1978
25. American Psychologtcal Associatton: Standards for Education and Psychological Tests. 1200 Seventeenth
Street N.W., Washington, D.C. 20036: American Psychological Associations Inc.
26. Carmines EG. Zcller RA: Reliability and Validity Assessment. Beverly Hills, Calif: Sage Publications. 1979
 䢇
Research Report

Predicting the Probability for Falls in

Community-Dwelling Older Adults
Using the Timed Up & Go Test
Background and Purpose. This study examined the sensitivity and
ўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўў

specificity of the Timed Up & Go Test (TUG) under single-task versus

dual-task conditions for identifying elderly individuals who are prone
to falling. Subjects. Fifteen older adults with no history of falls (mean
age⫽78 years, SD⫽6, range⫽65– 85) and 15 older adults with a history
of 2 or more falls in the previous 6 months (mean age⫽86.2 years,
SD⫽6, range⫽76 –95) participated. Methods. Time taken to complete
the TUG under 3 conditions (TUG, TUG with a subtraction task
[TUGcognitive], and TUG while carrying a full cup of water
[TUGmanual]) was measured. A multivariate analysis of variance and
discriminant function and logistic regression analyses were performed.
Results. The TUG was found to be a sensitive (sensitivity⫽87%) and
specific (specificity⫽87%) measure for identifying elderly individuals
who are prone to falls. For both groups of older adults, simultaneous
performance of an additional task increased the time taken to com-
plete the TUG, with the greatest effect in the older adults with a history
of falls. The TUG scores with or without an additional task (cognitive
or manual) were equivalent with respect to identifying fallers and
nonfallers. Conclusions and Discussion. The results suggest that the
TUG is a sensitive and specific measure for identifying community-
dwelling adults who are at risk for falls. The ability to predict falls is not
enhanced by adding a secondary task when performing the TUG.
[Shumway-Cook A, Brauer S, Woollacott M. Predicting the probability
for falls in community-dwelling older adults using the Timed Up & Go
Test. Phys Ther. 2000;80:896 –903.]

Key Words: Balance, Fall prevention.

ўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўў

Anne Shumway-Cook

Sandy Brauer

Marjorie Woollacott

896 Physical Therapy . Volume 80 . Number 9 . September 2000

 ўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўўў
T
he Timed Up & Go Test (TUG) is a test of turning).1 Thus, one goal of our research was to deter-
balance that is commonly used to examine mine the sensitivity and specificity of the TUG for
functional mobility in community-dwelling, frail predicting falls in community-dwelling older adults.
older adults (aged 70 – 84 years).1 The test
requires a subject to stand up, walk 3 m (10 ft), turn, Recent research has suggested that assessment of bal-
walk back, and sit down. Time taken to complete the test ance under multi-task conditions may be a more sensi-
is strongly correlated to level of functional mobility.1 tive indicator of balance problems and falls than assess-
Older adults who are able to complete the task in less ment of balance in a single-task context.2–5 Researchers
than 20 seconds have been shown to be independent in using laboratory tests of balance under dual-task condi-
transfer tasks involved in activities of daily living, have tions have found that the ability to maintain stability can
high scores on the Berg Balance Scale, and walk at gait be affected by performance of concurrent cognitive
speeds that should be sufficient for community mobility tasks,2– 4 and this effect is enhanced in older adults with
(0.5 m/s).1 In contrast, older adults requiring 30 sec- balance impairments and a recent history of falls.5
onds or longer to complete the task tend to be more
dependent in activities of daily living, require assistive Lundin-Olsson and colleagues6 investigated the effect of
devices for ambulation, and score lower on the Berg performing multiple tasks on balance, mobility, and falls
Balance Scale.1 in frail older adults who lived in an institutional setting.
Physical frailty is defined by severely impaired strength,
Although the TUG has been shown to be useful for mobility, balance, and endurance. They modified the
predicting level of functional mobility, its validity for TUG to add a manual task (TUGmanual) (ie, carrying a
identifying community-dwelling older adults who are at glass of water) and found that frail older adults who had
risk for falls is unknown. Functional mobility is a term used a time difference of greater than 4.5 seconds between
to reflect the balance and gait maneuvers used in the TUGmanual and the TUG were more prone to falls
everyday life (eg, getting in and out of a chair, walking, during the following 6 months. They concluded that the

A Shumway-Cook, PT, PhD, is Associate Professor, Division of Physical Therapy, University of Washington, Box 356490, Seattle, WA 98195 (USA)
(ashumway@[Link]). Address all correspondence to Dr Shumway-Cook.

S Brauer, PT, PhD, is Postdoctoral Fellow, Department of Exercise and Movement Science, University of Oregon, Eugene, Ore.

M Woollacott, PhD, is Professor and Chair, Department of Exercise and Movement Science, and Member, Institute of Neuroscience, University
of Oregon.

All authors provided concept/research design, writing, data collection and analysis, fund procurement, subjects, facilities and equipment, and
institutional liaisons. The authors acknowledge Robin High, Statistical Consultant, for his statistical expertise.

The study was approved by the Institutional Review Board of Northwest Hospital, Seattle, Wash.

This investigation was supported by a grant from the Foundation for Physical Therapy to Dr Shumway-Cook and by National
Institute on Aging grant AG-05317 to Dr Woollacott.

This article was submitted May 12, 1999, and was accepted May 30, 2000.

Physical Therapy . Volume 80 . Number 9 . September 2000 Shumway-Cook et al . 897

difference in time between the TUG and TUGmanual is Method
useful for identifying institutionalized elderly people
who are prone to falls.6 Whether performance on the Subjects
TUG in a dual-task condition is a useful measure for Thirty community-dwelling older adults living in the
identifying community-dwelling older adults who are greater Seattle area were enrolled in the study after
prone to falls is not known. Thus, another goal of our giving informed consent. The participants were volun-
study was to determine whether, in community-dwelling teers recruited from subjects involved in previous aging
older adults, the TUG performed under dual-task con- studies that were carried out by the first author. The
ditions was a more sensitive and specific predictor of falls participants were 15 older adults with no history of falls
than the TUG measure alone. (mean age⫽78 years, SD⫽6, range⫽65– 85) and 15
older adults with a history of 2 or more falls in the
In a previous study using a simultaneous task paradigm, previous 6 months (mean age⫽86.2 years, SD⫽6,
we studied the effects of 2 types of secondary tasks (a range⫽76 –95). The inclusion criteria were that the
spatial orientation task versus a language task) on pos- subjects had to be aged 65 years or older, living inde-
tural control during stance under 2 surface conditions pendently in the community, able to walk 9.1 m (30 ft)
(firm versus compliant).5 We compared the effects of with or without an assistive device but without the
these tasks on balance between a group of young adults assistance of another person, able to follow simple
(aged 24 – 44 years) and 2 groups of older adults instructions, and able to carry a cup by a handle.
(ie, those with a history of falls [aged 65– 86 years] and
those without a history of falls [aged 65–94 years]).5 In Criteria for inclusion in the group of older adults with a
young adults, neither secondary task affected stability in history of falls included a self-report of 2 or more falls
stance. In contrast, in the older nonfallers, the effect of within the past 6 months. A fall was defined as any event
a secondary task was dependent on the difficulty of the that led to an unplanned, unexpected contact with a
postural task. In the less challenging postural condition supporting surface. We excluded falls resulting from
(standing on a firm surface), neither cognitive task unavoidable environmental hazards such as a chair
affected balance; however, there was a significant collapsing. In addition, we excluded people who had
increase in postural sway when cognitive tasks were only 1 fall within 6 months in order to maximize the
performed in the more challenging postural condition possibility of selecting a sample of older adults with
(standing on a compliant foam surface). Finally, the recurrent fall problems. Further criteria for the older
older adults with balance problems and a history of adults in the faller category included an absence of
recurrent falls swayed more when performing either known neurological or musculoskeletal diagnosis that
secondary task even in the less challenging postural could account for possible imbalance and falls, such as
condition. Results from that study suggest that the effect cerebrovascular accident, Parkinson disease, cardiac
of a secondary task on postural control was dependent problems, transient ischemic attacks, or lower-extremity
on the balance abilities of the subject, the difficulty of joint replacements.
the balance task, and the type of secondary task being
performed.5 The effects of different types of secondary Procedure
tasks on functional mobility have not been determined. After informed consent was obtained, subjects com-
Certain types of secondary tasks performed in conjunc- pleted a health status questionnaire providing informa-
tion with the TUG may be more sensitive predictors of tion on age, residential status, marital status, medical
falls than others. Therefore, another goal of this study history, current coexisting medical conditions, self-
was to compare the sensitivity and specificity of 2 condi- reported history of imbalance, type of assistive device
tions of TUG performance (cognitive versus manual) in used for ambulation, and use of prescription medica-
identifying community-dwelling older adults who are at tions. This information was used to characterize the
risk for falls. demographics and health status of subjects participating
in the study. In order to verify our classification of the 2
Our hypotheses were (1) that, although the TUG itself groups based on balance abilities, balance was measured
would be a sensitive predictor of falls, the dual-task TUG using 1 self-report instrument and 2 performance-based
would be a more sensitive predictor of falls than the measures. The Activities-specific Balance Confidence
TUG alone and (2) that the TUG with the addition of a Scale (ABC) is a self-report measure of balance that
cognitive task (TUGcognitive) would be a more specific people can use to rate their perceived confidence
and sensitive predictor of falls than the TUGmanual. related to balance when performing common activities
of daily living.7 We asked our subjects to use the ABC to
rate their degree of confidence (1⫽no confidence to
10⫽full confidence) in performing 10 basic activities of
daily living and instrumental activities of daily living

898 . Shumway-Cook et al Physical Therapy . Volume 80 . Number 9 . September 2000

 ўўўўўўўўўўўўўўўўўўўўўўўўўўў
without fear of loss of balance. The result was a score Table 1.
Demographics of Older Adults Without a History of Falls (n⫽15) and
ranging from 10 to 100. In a previous study examining
Older Adults With a History of Falls (n⫽15)
the psychometric properties of the ABC on 102 older
adults living in the community, the ABC test-retest
Nonfallers Fallers
correlation (r) was .92 (P⬍.001), indicating good test-
retest reliability; the Cronbach alpha was .96, indicating Agea (y)
high internal consistency.7 The test did not discriminate X 78.4 86.2
fall status among community-dwelling older adults.7 SD 5.8 6.4
Range 65–85 76–95

Subjects then underwent a 45-minute performance- Sex

Female 10 (67%) 8 (53%)
based evaluation of balance and mobility function. Bal-
ance was evaluated using the Berg Balance Scale, which Assistive device
None 15 (100%) 3 (20%)
rates performance from 0 (cannot perform) to 4 (nor- Cane 7 (47%)
mal performance) on 14 different tasks, including ability Walker 5 (33%)
to sit, stand, reach, lean over, turn and look over each No. of prescriptions
shoulder, turn in a complete circle, and step.8 The total 0–1 50% 42%
possible score on the Berg Balance Scale is 56, which is 2–3 50% 42%
supposed to indicate excellent balance. The Berg Bal- ⱖ4 16%
ance Scale has been shown to yield measurements with No. of comorbidities
excellent interrater and test-retest reliability (intraclass 0–1 82% 33%
correlation coefficient [ICC]⫽.98 and .97, respectively) 2–3 18% 59%
ⱖ4 8%
and good internal consistency (Cronbach alpha⫽.96).8
Scores obtained from the Berg Balance Scale been Self-paced gait speed (m/s)a 1.2 (0.1) 0.47 (0.2)
shown to be correlated with scores obtained from other Berg Balance Testa
tests of balance and mobility, including the Tinetti (range 0–56) 55.5 (1) 32.7 (7.5)
Mobility Index (r ⫽⫺.91) and the Get Up & Go Test Activities-specific Balance
(r ⫽⫺.76).9 Confidence Scalea
(range 0–100) 93.2 (7) 53.0 (17)

Mobility was evaluated by asking subjects to walk for 3 a

Significant difference (P⬍.05) between groups.
minutes at their preferred speed. Distance walked was
measured, and speed for self-paced gait was determined.
Subject demographics and clinical test results are sum- Data Analysis
marized in Table 1. Two raters measured performance by timing the TUG in
all 3 conditions. Interrater reliability as measured with
Experimental Protocol an ICC (3,3) was high, with ri⫽.98, .99, and .99 for the
Subjects were asked to complete 3 trials of the TUG TUG, TUGmanual, and TUGcognitive, respectively.10 As a
under 3 conditions: performance of the TUG alone, result, further analyses were performed on the data
performance of the TUG with the addition of a cognitive obtained from one rater only. Histograms and descrip-
task (TUGcognitive), and performance of the TUG with tive statistics were calculated to determine distributions,
the addition of an upper-extremity motor task detect outliers, and consider the need for transforma-
(TUGmanual). The tasks were presented in random tions. A logarithmic transformation was performed on
order. When performing the TUG, subjects were given time data due to nonnormality, the result of positively
verbal instructions to stand up from a chair, walk 3 m as skewed data.
quickly and as safely as possible, cross a line marked on
the floor, turn around, walk back, and sit down. Those A multivariate analysis of variance (MANOVA) was then
subjects who used an assistive device when walking in the performed to determine whether group differences
community were requested to use that device. In the existed on the transformed time measures. Scores were
TUGcognitive, subjects were asked to complete the test averaged over the 3 trials. In cases where the MANOVA
while counting backward by threes from a randomly demonstrated a difference (P⬍.05), a post hoc within-
selected number between 20 and 100. In the TUGmanual, subject analysis was performed. To determine the ability
subjects were asked to complete the test while carrying a of the TUG performed under the 3 conditions to predict
full cup of water. Subjects who used a walker for ambu- faller versus nonfaller group membership, cross-vali-
lation (n⫽5) were excluded from this condition. Sub- dated discriminant function analyses were performed on
jects were given one TUG practice trial to familiarize 5 variables: the 3 transformed timed measures and 2
themselves with the task. difference measures (TUGmanual⫺TUG and TUGcognitive⫺
TUG). A prior probability of classifying an older adult as
a faller was set at 35%, based on falls rate literature.11,12

Physical Therapy . Volume 80 . Number 9 . September 2000 Shumway-Cook et al . 899

Table 2.
Comparison of Time Taken (in Seconds) to Complete the TUG, TUGmanual, and TUGcognitive in Older Adults Without a History of Falls (n⫽15) and
Older Adults With a History of Falls (n⫽15)a

TUG TUGmanual TUGcognitive

Nonfallers Fallers Nonfallers Fallers Nonfallers Fallers

X 8.4 22.2 9.7 27.2 9.7 27.7

SD 1.7 9.3 1.6 11 2.3 11.6
Range 6.4–12.6 10.3–39.2 6.9–12.6 14–48.7 6.2–14.6 11–49.6
a
TUG⫽Timed Up & Go Test, TUGmanual⫽Timed Up & Go Test with addition of a manual task, TUGcognitive⫽Timed Up & Go Test with addition of a cognitive
task.

A logistic regression procedure was used to determine

the cutoff value for each timed test that maximized
sensitivity and specificity and predicted the probability of
fallers at .5 or greater. All of these analyses were per-
formed with SAS 6.12.13,* The Pearson product moment
correlation coefficient (r) was used in bivariate correla-
tions, and a 3-way analysis of variance was performed to
compare demographics across the 2 groups. These anal-
yses were performed using SPSS 8.0.14,†

Results
Results from the MANOVA showed that the older adults
with a history of falls were slower (F⫽22.97, P⬍.001) Figure 1.
than the older adults without a history of falls in all 3 Comparison of the time taken to complete the Timed Up & Go Test (TUG)
as a function of the assistive device used for ambulation.
conditions (TUG, TUGcognitive, and TUGmanual). Table 2
compares the time taken to complete the TUG in all 3
conditions for both groups of subjects.
F⫽21.9, P⬍.001). Figure 2 compares the performance of
Because there was a difference in age (t⫽⫺3.5, P⫽.002) individuals within each of the 2 groups on the 3 tasks.
between the older adults with a history of falls and the Table 3 displays the increase in time taken to complete
older adults without a history of falls, a MANCOVA was the TUG in the 2 secondary task conditions for each of
used to reanalyze the TUG data using age as a covariate. the 2 groups. In the older adults with a history of falls,
Using this analysis, there was still a difference in time the time taken to complete the TUGcognitive increased on
taken to complete the TUG between the 2 groups, average 25% compared with 16% in the older adults
suggesting that the differences found in older subjects without a history of falls. The TUGmanual increased 22%
were not due to age alone, but rather due to balance in the older adults with a history of falls compared with
status. In addition, the time taken to complete the TUG 15% in the older adus without a history of falls.
by the older adults with a history of falls was highly
correlated (r ⫽.95) with the type of assistive device used Discriminate analysis indicated that the 3 measures
for ambulation. The time taken to complete the TUG (TUG, TUGmanual, and TUGcognitive) were equivalent
with no device was 9.0 seconds (range⫽6.4 –13.4), the with respect to classifying fallers and nonfallers, suggest-
time taken to complete the TUG with a cane was 18.1 ing that all 3 tests were comparable for identifying
seconds (range⫽14.6 –22.3), and the time taken to com- community-dwelling older adults who are prone to falls.
plete the TUG with a front-wheeled walker was 33.8 In Table 4, we compare the sensitivity and specificity for
seconds (range⫽28.3–39.2). These results are illustrated each of the 3 tests in predicting falls in community-
in Figure 1. dwelling older adults. The TUG alone correctly classified
13/15 fallers (87% sensitivity) and 13/15 nonfallers
The addition of either a cognitive task or a manual task (87% specificity). Compared with the TUG under single-
increased the time taken to complete the TUG in both task conditions, use of the dual-task TUG (either cogni-
groups of older adults (older adults with a history of falls: tive or manual) to identify fall status decreased sensitivity
F⫽79.3, P⬍.001; older adults without a history of falls: to 80% while increasing specificity to 93%, thus main-
taining a comparable overall prediction rate (87%).
Using difference scores between the dual-task TUG and
* SAS Institute Inc, PO Box 8000, Cary, NC 27511. the single-task TUG to determine probability for falls
†
SPSS Inc, 444 N Michigan Ave, Chicago, IL 60611.

900 . Shumway-Cook et al Physical Therapy . Volume 80 . Number 9 . September 2000

 ўўўўўўўўўўўўўўўўўўўўўўўўўўў
Table 4.
Comparison of the Sensitivity and Specificity for Identifying Older
Adults With and Without a History of Fallsa

Sensitivity Specificity Overall

(% Fallers) (% Nonfallers) Prediction

TUG 13/15 (87%) 13/15 (87%) 26/30 (87%)

TUGmanual 12/15 (80%) 14/15 (93%) 26/30 (87%)
TUGcognitive 12/15 (80%) 14/15 (93%) 26/30 (87%)
TUGmanual⫺TUG 10/15 (67%) 13/15 (87%) 23/30 (76%)
TUGcognitive⫺TUG 7/15 (47%) 14/15 (93%) 21/30 (70%)
a
TUG⫽Timed Up & Go Test, TUGmanual⫽Timed Up & Go Test with addition
of a manual task, TUGcognitive⫽Timed Up & Go Test with addition of a
cognitive task.

prediction rate of 90%. On the TUGmanual, classification

of older adults as fallers using the time score of 14.5
seconds or longer resulted in a 90% correct prediction
rate. Finally, elderly subjects who completed the
TUGcognitive in 15 seconds or longer were classified as
fallers with an overall correct prediction rate of 87%.

The relationship between fall status and time taken to

perform the TUG in the 3 conditions is shown in Figure 3.
Figure 2.
Comparison of the performance of individuals within the group of older
In all 3 conditions, this relationship was nonlinear. For
adults without a history of falls and the group of older adults with a example, an individual who took 13 seconds to complete
history of falls on the 3 tasks (Timed Up & Go Test [TUG] alone, TUG the TUG had a 69% probability of being a faller. An
with addition of a cognitive task [TUGcognitive], and TUG with addition of individual who took 14 seconds to complete the TUG
a manual task [TUGmanual]). had an 83% probability of being a faller. This nonlinear-
ity could also be seen in the TUGmanual, where a 16-
Table 3. second score was associated with a 70% probability of
Difference in Time Taken to (in Seconds) Complete the TUGmanual and being a faller, whereas the probability rose to 81% with
TUGcognitive Compared With the TUG in Older Adults Without a a score of 17 seconds.
History of Falls (n⫽15) and Older Adults With a History of Falls
(n⫽15)a
Discussion and Conclusions
We investigated the sensitivity and specificity of the TUG
TUGcognitive TUGmanual
under both single and dual task conditions in identifying
Nonfallers Fallers Nonfallers Fallers fall prone older adults living in the community. Our
X 1.28 5.56 1.32 4.98 results indicate that the TUG itself is a sensitive and
SD 1.0 3.2 1.0 2.2 specific indicator of whether falls occur in community-
Range ⫺0.3–3.0 0.7–11.7 0.01–3.9 2.2–10.9 dwelling older adults. Thus, we believe the TUG is a
a
TUG⫽Timed Up & Go Test, TUGmanual⫽Timed Up & Go Test with addition
relatively simple screening test that takes only minutes to
of a manual task, TUGcognitive⫽Timed Up & Go Test with addition of a complete, and we contend that it appears to be a valid
cognitive task. method for screening for both level of functional mobil-
ity and risk for falls in community-dwelling elderly
people. Results from the discriminate analysis suggest
resulted in lower prediction rates than when the actual
that older adults who take longer than 14 seconds to
test scores were used.
complete the TUG have a high risk for falls. Our cutoff
value of 14 seconds is different from that of Podsiadlo
Logistic regression analysis was used to determine cutoff
and Richardson,1 who found that a cutoff value of
levels for the TUG, the TUGmanual, and the TUGcognitive.
greater than 30 seconds was best for predicting func-
Cutoff levels that maximized both sensitivity and speci-
tional dependence among older adults. The differences
ficity and had a predicted probability of .5 or larger were
in time values may reflect the differences in subjects
determined to identify the specific time score that best
used in the 2 studies. Podsiadlo and Richardson’s study
classified fallers and nonfallers. As shown in Table 5,
included older adults with a wide range of neurologic
older adults who took 13.5 seconds or longer to perform
pathologies. In contrast, we studied community-
the TUG were classified as fallers with an overall correct

Physical Therapy . Volume 80 . Number 9 . September 2000 Shumway-Cook et al . 901

Table 5.
Balance Test Cutoff Scores to Maximize Sensitivity, Specificity, and Predicted Probabilitya

Cutoff Sensitivity Specificity Overall Predicted

Score (s) (% Fallers) (% Nonfallers) Prediction Probability

TUG ⱖ13.5 80% 100% 90% .77

TUGmanual ⱖ14.5 86.7% 93.3% 90% .5
TUGcognitive ⱖ15 80% 93.3% 86.7% .5
a
TUG⫽Timed Up & Go Test, TUGmanual⫽Timed Up & Go Test with addition of a manual task, TUGcognitive⫽Timed Up & Go Test with addition of a cognitive
task.

remain useful as predictors of future falls in some

populations of elderly individuals.

Effect of Multiple Tasks on Mobility in Elderly People

Who Are Prone to Falls
Results from this study confirm that simultaneous per-
formance of a secondary task had a deleterious effect on
functional mobility. The addition of a secondary task
increased the time taken to complete the TUG by 22% to
25%. This effect was independent of the type of second-
ary task performed (either manual or cognitive). These
Figure 3. findings do not support our hypothesis that cognitive
The time taken to perform the Timed Up & Go Test (TUG), TUG with
tasks would affect mobility more than manual tasks.
addition of a manual task (TUGmanual), and TUG with addition of a
cognitive task (TUGcognitive) as a function of probability of predicting
elderly individuals who are prone to falls. Clinical Implications
Given the incidence and consequences of falls among
older adults, screening methods that identify elderly
dwelling, frail older adults, but we excluded older adults individuals who are prone to falls and who may benefit
with known neurologic diseases. from interventions designed to improve balance and
decrease falls and risk for falls are critical. Results from
Our finding that all 3 tests (TUG, TUGmanual, and our study suggest that the TUG is a simple screening test
TUGcognitive) were comparable in determining the like- that is a sensitive and specific measure of probability for
lihood of falling in community-dwelling older adults did falls among older adults.
not support our hypothesis that measurement of mobil-
ity under multi-task conditions would be a more sensitive References
1 Podsiadlo D, Richardson S. The timed “Up & Go”: a test of basic
indicator of likelihood for falls. In addition, difference
functional mobility for frail elderly persons. J Am Geriatr Soc. 1991;39:
scores (eg, TUGmanual⫺TUG) did not increase the abil- 142–148.
ity to identify community-dwelling older adults who are
2 Guerts ACH, Mulder TW, Nienhuis B, Rijken RAJ. Dual task assess-
prone to falls. This finding is not consistent with Lundin- ment of reorganization in persons with lower limb amputation. Arch
Olsson and colleagues’6 finding that comparing mobility Phys Med Rehabil. 1991;72:1059 –1064.
performance between single-task and dual-task condi-
3 LaJoie Y, Teasdale N, Bard C, Fleury M. Attentional demands for
tions was a more useful way of predicting future falls in static and dynamic equilibrium. Exp Brain Res. 1993;97:139 –144.
institutionalized elderly people. There are 2 dissimilari-
4 Lundin-Olsson L, Nyberg L, Gustafson Y. “Stops walking when
ties between these studies that could explain these talking” as a predictor of falls in elderly people. Lancet.
differences. First, Lundin-Olsson and colleagues studied 1997;349(9052):617.
a population of frail older adults living in an assisted-
5 Shumway-Cook A, Woollacott M, Kerns KA, Baldwin M. The effects of
living environment. We examined older adults who, two types of cognitive tasks on postural stability in older adults with and
despite having balance impairments, were living inde- without a history of falls. J Gerontol A Biol Sci Med Sci. 1997;52:M232–
pendently within the community. Second, in Lundin- M240.
Olsson and colleagues’ study, predictors of future falls 6 Lundin-Olsson L, Nyberg L, Gustafson Y. Attention, frailty, and falls:
were sought. In contrast, we were looking for measures the effect of a manual task on basic mobility. J Am Geriatr Soc.
that are useful for identifying individuals with a likeli- 1998;46:758 –761.
hood of falling. It may be that, although difference 7 Powell LE, Myers AM. The Activities-specific Balance Confidence
scores are not useful in a discriminate function, they may (ABC) Scale. J Gerontol A Biol Sci Med Sci. 1995:50:M28 –M34.

902 . Shumway-Cook et al Physical Therapy . Volume 80 . Number 9 . September 2000

 ўўўўўўўўўўўўўўўўўўўўўўўўўўў
8 Berg KO, Wood-Dauphinee SL, Williams JI, Gayton D. Measuring 12 Reinsch S, McRae P, Lachenbruch PA, Tobis JS. Attempts to prevent
balance in the elderly: preliminary development of an instrument. falls and injury: a prospective community study. The Gerontologist.
Physiotherapy Canada. 1989;41:304 –311. 1992;32:450 – 456.
9 Berg KO, Wood-Dauphinee SL, Williams JT, et al. Measuring balance 13 SAS/STAT Version 6.12. Cary, NC: SAS Institute Inc; 1996.
in the elderly: validation of an instrument. Can J Public Health.
14 SPSS for Windows: Advanced Statistics, Release 8.0. Chicago, Ill: SPSS
1992;83(suppl 2):S7–S11.
Inc; 1998.
10 Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater
reliability. Psychol Bull. 1979;86:420 – 428.
11 Lach HW, Reed AT, Arfken CL, et al. Falls in the elderly: reliability
of a classification system. J Am Geriatr Soc. 1991;39:197–202.

Physical Therapy . Volume 80 . Number 9 . September 2000 Shumway-Cook et al . 903

294 Book Reviews

the interactionist position on personality. Instead, a rather dated view of personality as ‘traits’ is adopted
rather uncritically throughout. The second, more important deficit concerns the epistemology of establishing
cause. The book is part of a series on Health Psychology, and there is no understanding of the importance
of epidemiology in trying to establish causes of disease. The authors are addressing an epidemiological
question: does personality increase the risk of disease? There is therefore a dearth of reference to the
epidemiological literature and the book as a whole seems light on empiricism and overburdened with
lengthy and verbose discussion.
The exception is a clear and well argued chapter by Stone and Costa who take apart the suggestion
that neuroticism is a risk factor for cardiovascular disease. By marshalling some of the well conducted
epidemiological studies in the area they illustrate that sound epidemiological evidence is needed to
inform debate in this area and is absent in its support for the notion that neuroticism is bad for health.
They conclude that neuroticism leads to increased reporting and consultation for episodes of ill-health.
For particular attack is an article by none other than H.S. Friedman which used meta-analytic methods
claiming that it supported the existence of a disease prone personality. Stone and Costa take this article
apart though Friedman still quotes this paper approvingly in his introductory chapter.
This editorial strategy may be a refreshingly frank encouragement of debate. The most ‘brilliant’
contribution to this book is a thorough hatchet job on the editor’s own work!

GLYN LEWIS

Hon. Senior Lecturer

Section of Epidemiology and General Practice
Institute of Psychiatry
De Crespigny Park
London, SE5 8AF

Health Measurement Scales. A Practical Guide to Their Development and Use. DAVID L STREINER
and GEOFFREYR NORMAN. Oxford: Oxford University Press, (HB) 1989. (PB) 1990. Price 515.00 (PB).
175 pp.

THE DEVELOPMENTof’ measurement scales, be they in self report or interview format, has expanded
enormously in the health and social sciences. As quality of life has assumed greater importance in
valuation than mere quantity of life, measurement has become accordingly more complex and more
subtle. This book is intended as an introduction to concepts of measurement in the health sciences. It
is a pragmatic guide to the construction of new instruments or the appraisal of existing forms. As the
authors state in their introduction, it is not intended as a textbook, but as a guide which emphasises
concepts underlying current approaches to measurement, rather than the complicated mathematics which
underpin them.
The book proceeds through the steps involved in the construction of new scales in a logical fashion
placing concepts of validity, reliability, generalizability and measurement of change in clear perspective.
The result is an intelligible guide to the principles of measurement and to my mind is superior to any
other format I have seen. It is well written and treats the reader as astute but not necessarily statistically
fluent. My only negative criticism would be of its rather superficial treatment of translation and the
difficulties of cross cultural measurement, perhaps a subject for a book in its own right.
For anyone attempting to measure health status, be it physical, psychological or social, or for those
who want to take a more critical view of the literature on measurement, I cannot recommend this small
book highly enough.

MICHAEL KING
Book Review Editor
Royal Free Hospita!
London NW3 2QG

Gays, Lesbians and Their Therapists. Edited by CHARLES [Link] York: W. W. Norton
& Company, 1991. Price $34.95. 274 pp.

[Link] two principal themes to this book on therapy with homosexual men and women. The first
is the distinctiveness of this client group and their therapists and the second concerns the emotional
imperfection of the therapist in any psychotherapeutic encounter. The introduction by the editor traces
both the history of discrimination against homosexuals (patients and therapists) and the gulf which
Health and Quality of Life Outcomes BioMed Central

Research Open Access

Correspondence between EQ-5D health state classifications and
EQ VAS scores
David K Whynes* for the TOMBOLA Group

Address: School of Economics, University of Nottingham, Nottingham NG7 2RD, UK

Email: David K Whynes* - [Link]@[Link]
* Corresponding author

Published: 7 November 2008 Received: 24 July 2008

Accepted: 7 November 2008
Health and Quality of Life Outcomes 2008, 6:94 doi:10.1186/1477-7525-6-94
This article is available from: [Link]
© 2008 Whynes and the TOMBOLA Group; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License ([Link]
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract
Background: The EQ-5D health-related quality of life instrument comprises a health state
classification followed by a health evaluation using a visual analogue scale (VAS). The EQ-5D has
been employed frequently in economic evaluations, yet the relationship between the two parts of
the instrument remains ill-understood. In this paper, we examine the correspondence between
VAS scores and health state classifications for a large sample, and identify variables which
contribute to determining the VAS scores independently of the health states as classified.
Methods: A UK trial of management of low-grade abnormalities detected on screening for cervical
pre-cancer (TOMBOLA) provided EQ-5D data for over 3,000 women. Information on distress and
multi-dimensional health locus of control had been collected using other instruments. A linear
regression model was fitted, with VAS score as the dependent variable. Independent variables
comprised EQ-5D health state classifications, distress, locus of control, and socio-demographic
characteristics. Equivalent EQ-5D and distress data, collected at twelve months, were available for
over 2,000 of the women, enabling us to predict changes in VAS score over time from changes in
EQ-5D classification and distress.
Results: In addition to EQ-5D health state classification, VAS score was influenced by the subject's
perceived locus of control, and by her age, educational attainment, ethnic origin and smoking
behaviour. Although the EQ-5D classification includes a distress dimension, the independent
measure of distress was an additional determinant of VAS score. Changes in VAS score over time
were explained by changes in both EQ-5D severities and distress. Women allocated to the
experimental management arm of the trial reported an increase in VAS score, independently of any
changes in health state and distress.
Conclusion: In this sample, EQ VAS scores were predictable from the EQ-5D health state
classification, although there also existed other group variables which contributed systematically
and independently towards determining such scores. These variables comprised psychological
disposition, socio-demographic factors such as age and education, clinically-important distress, and
the clinical intervention itself.
Trial registration: ISRCTN34841617

Page 1 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2008, 6:94 [Link]

Background tion has been paid to correspondence within the self-

The EQ-5D is a well-established and widely-used generic report instrument itself, to the relationship between the
instrument for assessing health-related quality of life [1]. individual's EQ VAS score and his or her EQ-5D classifica-
Designed as a self-completion questionnaire, it embodies tion.
two components, a health state description followed by
an evaluation. The respondent classifies his or her prevail- In view of the sequence of completion, the EQ VAS score
ing state of health by selecting one of three different levels relates to that which the individual thinks about the
of problem severity within each of five health domains. health state in which s/he has declared her/himself to be.
The levels are none, moderate and severe/extreme (coded Moving from the first to the second part of the EQ-5D
1 through 3, respectively), whilst the domains are mobil- questionnaire requires the subject to translate her/his
ity, capacity for self-care, conduct of usual activities, pain/ description of personal well-being, represented by the
discomfort and anxiety/depression, ordered as such. The extent of health problems in five dimensions, into a uni-
conscious health states are therefore limited to 243 sever- dimensional value of health, however s/he cares to define
ity/domain vectors, ranging from 11111 (no problems in it. Whilst we would certainly expect individuals to inter-
any domain) to 33333 (severe problems in all five pret the presence of more health problems, each of higher
domains). Having located the current health state, the severity, as poorer rather than better health, the specific
respondent then evaluates his or her health using a visual translation for each person must remain essentially sub-
analogue scale (VAS). This is a vertical, calibrated, line, jective.
bounded at 0 ("worst imaginable health state") and at
100 ("best imaginable health state"). Respondents indi- Comparing EQ-5D classifications with VAS scores
cate where they perceive their present state of health to lie, amounts to an exploration of differential item function-
relative to these anchors. ing within the instrument [12]. Specifically, we hypothe-
sise that there exist group variables which contribute
Although the VAS was always integral to the EQ-5D, its systematically towards determining individuals' EQ VAS
role changed as the instrument evolved. The EQ-5D's scores, independently of those individuals' health states as
descriptive system was designed to allow the reported classified by the EQ-5D. We anticipate some degree of
health states to be evaluated, by assigning to each a qual- classification-independent variation for several reasons,
ity or value weight (index score). Initially, the VAS was the first being socio-demography. Age and education have
used to generate these weights; large population samples already been offered as explanations for the diversity in
were invited to value defined states by indicating appro- EQ VAS scores assigned by the general public to nominal
priate VAS positions [2]. Over time, however, the instru- health classifications described, for example, as "excel-
ment's developers came to favour alternative methods of lent" or "fair" [13]. Material deprivation and ethnic back-
evaluating health states [3]. In the operational (self- ground have been advanced as potential explanations for
report) version of the EQ-5D instrument, the VAS was divergences between self-reported and actual health states
retained to provide complementary information: "If the in the US population [14]. Second, it is probable that eval-
health status index is based on a set of weights derived uation is influenced by psychological disposition. Perceiv-
from values from general population samples, this ing oneself to be in control of one's own health has been
implies that the index can be regarded as a societal [sic] shown to influence positively both self-reported health
valuation of the respondent's health state, in contrast to status [15] and, more generally, subjective well-being
the respondent's or patient's own assessment of his/her [16]. Third, the VAS is continuous between 0 and 100,
health state (EQ VAS scores)" [[4] p.11]. whereas the classification scheme offers only three choices
of severity. EQ-5D subjects have reported feeling that the
There is an extensive body of research on the use of the three-level choice is too coarse to describe their circum-
VAS in population studies. It is now evident, for example, stances precisely [17]. Individuals with minor mobility
that VAS-derived utility weights differ from those elicited problems only, for example, are likely to recognise the
using the time trade-off or standard gamble techniques possibility of better health states, yet all might agree that
[5,6]. The weights can vary between populations [7]. Pop- the problems themselves are insufficient to merit assign-
ulation VAS ratings of conditions can differ from self-rat- ment to "moderate" or "severe" in the EQ-5D's mobility
ings, especially when the condition is more severe [8]. In domain. All these individuals would classify themselves
clinical studies involving subjects actually suffering from as having no health problems, yet all would record EQ
illnesses or disabilities, EQ VAS scores have been shown VAS scores of less than 100. That all would choose pre-
to be responsive to the symptoms and severities indicated cisely the same VAS score seems improbable, especially in
by condition-specific instruments [9-11]. In comparison view of the variety of representational heuristics which
with these lines of enquiry, however, relatively little atten- individuals are known to employ [18].

Page 2 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2008, 6:94 [Link]

Methods populations [26,27]. The HADS assesses depression and

Study sample anxiety independently on two sub-scales. Comparison of
Our analysis used data collected during the TOMBOLA the item scores for each sub-scale with established cut-off
randomised controlled trial, a multi-centre UK study of values enables the investigator to identify possible or
the management of women recording low-grade cytologi- probable cases of anxiety or depression. TOMBOLA
cal abnormalities (pre-cancers) as a result of routine cervi- recruits also completed the Multi-dimensional Health
cal screening. TOMBOLA was instigated because of Locus of Control Scale (MHLCS), an instrument which
uncertainty over the most effective means of managing locates subjects' perceived source of control over their
such women, the principal alternatives being immediate own health [28]. MHLCS comprises three ordinal sub-
referral to colposcopy, with treatment if indicated, and scales, each consisting of six statements. To each state-
cytological surveillance (Papanicoulou smear tests at six- ment, one of six levels of agreement is assigned (scored 1
monthly intervals) until the abnormality is seen to regress through 6), enabling summation to a sub-scale total. The
or progress [19]. The immediate referral of all cases had Internal sub-scale assesses the extent to which the subject
been thought unjustified until meta-analysis suggested perceives his/her health to be under his/her own direct
that women with low-grade abnormalities were at risk of control. The External (or "powerful others") sub-scale
eventually developing invasive cancer despite continued assesses the perceived importance of other people, for
surveillance [20]. Compared with referral, surveillance example, physicians and family, in determining health,
was believed to result in more defaults from follow-up whilst the Chance sub-scale assesses the perceived impor-
and in more pre-cancers being missed [21]. On the other tance of luck or fate. Each sub-scale has a range of 6-to-36,
hand, colposcopy-for-all is the more expensive option with higher values indicating stronger beliefs in that par-
[22] and might give rise to over-diagnosis and to unneces- ticular source of control.
sary treatment and cervical damage [23].
The HADS, EQ-5D and MHLCS data were all scored
As is required for all national trials in the UK, TOMBOLA's according to the conventional algorithms. Subjects with
recruitment and analysis protocol, including the research missing HADS or EQ-5D data were excluded from the
reported in this paper, had been granted full ethical analysis. The MHLCS algorithms permit imputation when
approval [24]. Cervical screening subjects are typically data is missing in part, although imputation was neces-
asymptomatic and are, on average, younger than the gen- sary only in a small number of cases (< 3 per cent). The
eral population. The TOMBOLA sample was in relatively EQ-5D index scores were derived from the current UK tar-
good general health, except in one respect. Having been iff [29,30], which has a maximum value of 1 (for health
informed of their abnormal cytology results, many of the state 11111) and a minimum value of -0.59 (for 33333).
women displayed elevated levels of anxiety and depres- Material deprivation was represented by the small-area
sion [25]. At the time of recruitment, TOMBOLA subjects Carstairs score, a composite measure comprising four
provided basic socio-demographic information and com- poverty-associated variables and based on data collected
pleted an array of quality-of-life and attitude question- during the decennial national census [31]. All such areas
naires, comprising both context-specific instruments and are ranked and divided into national quintiles, ranging
the EQ-5D. They were then randomised into two trial from the least- to the most-deprived. Each subject was
arms and managed accordingly. The control arm of the assigned to one of these quintiles, as determined by their
trial replicated current UK practice, namely, cytological home address.
surveillance. Those randomised to the active arm were
referred immediately to colposcopy, receiving treatment Analysis
as required (the current management practice for high- We modelled EQ VAS scores using ordinary least squares
grade abnormalities). The majority of subjects in both linear regression. Given the hypothesis under investiga-
arms completed a further array of questionnaires at 12 tion, the model contained EQ-5D health state classifica-
months after recruitment. tions as independent variables. In addition we included,
first, the HADS classification, to appraise the possibility
Measures that the EQ-5D classifications pertaining to the principal
With anxiety and distress expected to be the principal morbidity were insufficient in themselves to explain
morbidity, TOMBOLA employed the Hospital Anxiety health state values. Second, we included the MHLCS
and Depression Scale (HADS) as a specific measurement scores, anticipating that individuals who believed that
instrument. The HADS was developed to identify "case- they controlled their own health destinies would report
ness" with respect to anxiety, mood disorders and depres- higher subjective values of their health state. Finally, we
sion in non-psychiatric settings. It has been validated as a included a range of socio-demographic variables, with no
screening tool in a clinical context and has been used as a necessary expectation of sign on the coefficients, on the
primary instrument in investigations of both patients and basis of previous reports of associations between health

Page 3 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2008, 6:94 [Link]

values and socio-demographic factors. Carstairs scores VAS score was 87.0 (SD 10.7); for the remainder, it was
were not included as potential explanatory variables 74.5 (SD 17.5). The index and EQ VAS scores were signif-
because they proved to be collinear with the majority of icantly correlated (r = 0.51, p < 0.01).
individual characteristics.
Table 1 displays the characteristics of the recruitment sam-
To assess the stability of any relationship, we modelled ple, both by Carstairs quintile and overall. Differences in
changes in the EQ VAS score over the 12 months between sample composition as defined by Carstairs quintile were,
the two questionnaire arrays using, as independent varia- for the categorical variables, subjected to the chi-squared
bles, changes in the EQ-5D and in the HADS classifica- test. Differences for continuous variables were subjected
tions, plus the socio-demographic variables. We to one-way analysis of variance with Bonferroni adjust-
investigated the impact of one further factor in this second ment. Women drawn from quintiles characterised as
model, hypothesising that EQ VAS scores reported at the being less-deprived were more likely to be older, white,
second round of questionnaires would have been influ- cohabiting, non-smoking and with formal academic qual-
enced by the allocation to trial arms, for two reasons. First, ifications. The prevalence of HADS-assessed anxiety and
recruits to clinical trials necessarily accept that they cannot depression, and the likelihood of not working, increased
pre-determine the management method to which they with deprivation quintile. The MHLCS scores indicated
will be assigned, yet agreement to be randomised need that women from the most-deprived quintile placed more
not imply indifference to the randomisation outcome. It emphasis on both external factors and chance as control-
is established that a preference for the new practice under lers of health. Increased deprivation was associated with
investigation (i.e. an intervention not routinely available) lower mean EQ-5D index scores and lower mean EQ VAS
is a principal explanation for volunteering to participate scores.
in trials [32], whilst an unwillingness to risk randomisa-
tion away from current practice was found to be a princi- All of the Table 1 variables, with the exception of the EQ-
pal explanation for refusal to participate in TOMBOLA 5D index score, were candidates for the first regression
[33]. It is therefore likely than a prior preference for the analysis. Age and MHLCS were entered as continuous var-
new intervention (immediate colposcopy) was wide- iables, whilst the remaining variables (plus the EQ-5D
spread amongst TOMBOLA recruits. Second, by 12 classifications by severity and domain) were entered as
months, the cervical abnormalities detected in women dummies. Owing to the very small numbers of women
randomised to the colposcopy arm would have been reporting level 3 problems in the mobility, self-care and
resolved according to protocol. A proportion of women usual activities domains (n = 4, 0 and 11, respectively),
randomised to current practice, however, remained under those with problems at levels 2 or 3 were combined for
surveillance, and the uncertainties over their abnormali- these dimensions. The regression was estimated and re-
ties remained unresolved. We therefore judge that women estimated after excluding variables with insignificant coef-
randomised to the current practice arm of the trial (sur- ficients, to produce the model displayed in Table 2. The
veillance) might rate their health as worse, by virtue of signs associated with the EQ-5D domain coefficients, and
being denied the intervention which they had sought and the relative magnitudes associated with the severity of
of failing, in some cases, to have their uncertainties problem reported, are as would be expected. More severe
resolved. health problems in any dimension evidently gave rise to a
lower EQ VAS value for self-reported health. For any given
Results EQ-5D health state classification, the EQ VAS score was
The initial analysis was based on data from the recruit- lower if the respondent had a university degree, was a cur-
ment questionnaire array for 3,132 subjects. All were aged rent cigarette smoker, was non-white, was likely to be anx-
between 20 and 59 years. 53 different EQ-5D vectors were ious and/or depressed as assessed by the HADS, or located
represented in this recruitment sample, although 11111 control over her health in others. The EQ VAS score was
(no health problems in any of the five domains) was the higher if the respondent was older, or located control over
most frequently cited, by 53.9 per cent of subjects. Only her health in herself.
3.9 per cent of subjects recorded an index score at or
below 0.6, the lowest being -0.23. A further 41.8 per cent Matched EQ-5D and HADS data over two time points
recorded scores higher than 0.6 but up to and including (recruitment and 12 months thereafter) were available for
0.85. The proportions of EQ VAS scores up to 60, and 2,176 of the subjects. Of these, 50.6 per cent had been
higher than 60 but up to and including 85, were 7.2 and randomised after recruitment to immediate colposcopy,
45.8 per cent, respectively. 24.9 per cent of subjects leaving the remainder to undergo cytological surveillance
recorded scores of 90 and above, including 5.4 per cent (current practice). The data enabled us to calculate, for
who recorded the maximum score of 100. For those indi- each individual, (i) the change in the EQ VAS score over
viduals recording the 11111 health state, the mean EQ the period, (ii) changes in the severity of health problems

Page 4 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2008, 6:94 [Link]

Table 1: Sample characteristics, by Carstairs quintiles

Least-deprived Most-deprived
1 2 3 4 5 Full sample χ2 or F p=

Composition of sample, % 14.4 18.9 16.2 26.5 24.1 100.0

Ethnicity, % non-white 1.6 2.4 2.4 3.5 7.3 3.7 38.4 < 0.01
Cohabiting or married, % 61.6 60.0 64.7 51.2 45.4 55.1 65.2 < 0.01
Ever had children, % 54.7 54.5 61.0 51.9 53.2 54.6 11.3 0.02
Employment, %
Full-time 57.1 55.9 51.1 49.6 45.8 51.2 54.9 < 0.01
Part-time 23.1 22.5 27.5 21.4 21.0 22.7
Student 8.2 7.1 6.3 11.6 11.8 9.5
Not working 11.6 14.4 15.0 17.4 21.4 16.6
Training and qualifications, %
None 18.0 22.3 26.7 26.2 30.7 25.5 58.9 < 0.01
Via employment 17.3 20.5 21.2 18.0 21.1 19.6
Up to University level 29.8 29.8 30.7 28.0 29.0 29.3
University degree 34.9 27.4 21.4 27.8 19.1 25.6
Current cigarette smoker, % 26.8 26.0 27.9 38.2 44.8 34.2 81.4 < 0.01
HADS anxiety, %
Not a case 61.7 64.1 57.2 55.9 52.3 57.7 30.8 < 0.01
Possible 18.8 17.5 22.0 19.8 19.4 19.5
Probable 19.5 18.4 20.8 24.4 28.3 22.9
HADS depression, %
Not a case 93.9 92.6 91.8 91.1 89.3 91.5 18.2 < 0.01
Possible 5.4 4.8 5.0 7.1 7.3 6.1
Probable 0.7 2.6 3.2 1.8 3.4 2.4
Mean age, years 35.0 35.5 35.1 32.3 30.7 33.4 26.6 < 0.01
Mean MHLCS score
Internal 26.2 26.5 26.3 26.1 25.8 26.2 2.4 0.05
External 16.1 16.4 16.8 16.7 17.5 16.8 5.0 < 0.01
Chance 17.9 18.7 19.2 19.0 19.1 18.9 5.1 < 0.01
Mean EQ-5D score
Index 0.911 0.891 0.890 0.880 0.863 0.884 6.3 < 0.01
VAS 83.6 82.0 82.6 80.6 78.6 81.1 9.5 < 0.01

in each of the five EQ-5D domains, (iii) the change in the were also included, as was a dummy variable representing
likelihood of HADS caseness. With respect to (ii), we con- trial randomisation. The regression was estimated and re-
structed two dummy variables for each domain, one tak- estimated after excluding variables with insignificant coef-
ing the value of unity if the severity of health problem had ficients, to produce the model displayed in Table 3. The
increased (for example, a move from level 1 to level 2), coefficients for the EQ-5D variables were as anticipated;
the other being unity if it had decreased (for example, a an increased (decreased) severity of problem in any single
move from level 3 to level 1). Likewise, with respect to domain contributed to a fall (rise) in the EQ VAS score
(iii), dummies represented the likelihood of caseness from the recruitment baseline. For given changes in health
increasing over time (for example, a move from no-case to state, the EQ VAS score fell (rose) if the likelihood of
probable anxiety) or decreasing (for example, a move HADS-caseness of anxiety and/or depression increased
from probable to possible depression). In this two-period (decreased). In the initial estimation, none of the coeffi-
sample, the likelihood of HADS-anxiety and HADS- cients for the socio-demographic variables had achieved
depression caseness changed for 35.9 and 12.3 per cent of significance, implying that the VAS response to changing
subjects, respectively. The EQ VAS scores changed for 85.0 health states was independent of such factors. For a given
per cent of subjects, with a mean fall over the period of 1.5 health state and HADS-caseness, those randomised to the
(SD 15.1, IQR -5 to 10, range ± 75). immediate colposcopy arm of the trial (i.e. away from cur-
rent practice) reported an increase in EQ VAS score.
Movements in the EQ-5D domains and changes in the
HADS likelihood of caseness were entered into a regres- Moving between the Table 2 and the Table 3 models
sion model as independent variables, with the fall in EQ entailed the exclusion of 956 women from the sample.
VAS score as the dependant variable. The socio-demo- Although all of these women had supplied sufficient data
graphic and MHLCS variables used in the previous model at recruitment, they failed to supply EQ-5D scores or other

Page 5 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2008, 6:94 [Link]

Table 2: Regression, predicting VAS score ciations between individual characteristics and depriva-
tion levels evident in Table 1 were reproduced: excluded
β T-ratio p=
women, for example, were significantly more likely to be
Constant 83.23 49.32 < 0.01 young, smokers, unemployed and uneducated. It is not
Age, years 0.05 2.36 0.02 evident, however, that excluding cases between the Table
University degree = 1 -1.25 -2.29 0.02 2 and Table 3 models necessarily compromised the find-
Current smoker = 1 -2.84 -5.63 < 0.01 ings. First, the addition of Carstairs dummy variables to
Ethnicity, non-white = 1 -4.17 -3.35 < 0.01 the Table 3 model produced insignificant coefficients for
HADS any (at p = 0.27 or greater), suggesting that VAS changes
Possible anxiety = 1 -1.65 -2.61 0.01
were independent of deprivation. Second, re-estimating
Probable anxiety = 1 -3.22 -4.52 < 0.01
Possible depression = 1 -6.32 -5.90 < 0.01 the Table 2 model using the smaller, Table 3, sample, did
Probable depression = 1 -5.71 -3.45 < 0.01 not affect the formulation. No new variables appeared
MHLCS and no signs on existing variables changed, although the
Internal 0.29 -3.33 < 0.01 coefficients for age and EQ-5D mobility did become sta-
External -0.25 -2.62 0.01 tistically insignificant.
EQ-5D domain and level
Mobility 2/3 = 1 -4.61 -6.07 < 0.01
In the light of the relatively low coefficients of determina-
Self-care 2/3 = 1 -6.35 -8.60 < 0.01
Usual activities 2/3 = 1 -6.77 -5.88 < 0.01 tion, an analysis of residuals was conducted for each of
Pain/discomfort 2 = 1 -5.11 -9.19 < 0.01 the regressions. In each case, the scatter-plot of residuals
Pain/discomfort 3 = 1 -12.94 -10.79 < 0.01 against predicted values revealed a random pattern in the
Anxiety/depression 2 = 1 -5.47 5.17 < 0.01 distribution of outliers, and the normal probability plot
Anxiety/depression 3 = 1 -19.56 -5.91 < 0.01 was essentially linear.
Adjusted r2 0.32

Discussion
It appears that very few studies directly comparable to
necessary data at 12 months. To investigate selection bias, ours have been conducted. One employing the same
the characteristics of the excluded women were compared method was based on EQ-5D data obtained from around
with the 2,176 supplying adequate data both at recruit- 1,200 inhabitants of a South African suburb [34]. This
ment and at 12 months. Sample composition by Carstairs study's regression model suggested that, over and above
quintile differed significantly (χ2 = 43.10, p < 0.01). Of health state classification, significantly lower VAS scores
the sample used in the Table 3 model, 36.2 per cent was were associated with the presence of disability, being
drawn from the two least-deprived quintiles, compared older, unemployment and being in the lowest possible
with only 26.8 per cent for the excluded women. The asso- income band. The South African model shares three simi-
larities with our own. First, coefficients for EQ-5D health
Table 3: Regression, predicting decrease in VAS score states were significant and appropriately signed and, sec-
ond, the reporting or detection of a co-morbidity (disabil-
β T-ratio p= ity in the South African case, distress in ours) resulted in a
lower VAS for a given EQ-5D health state. Third, eco-
Constant 1.19 2.31 0.02
nomic deprivation emerged as an independent influence,
EQ-5D, level increases
Mobility 6.60 3.19 < 0.01 explicitly in the South African model although implicitly
Self-care 13.31 3.28 < 0.01 in ours. The characteristics which predicted higher VAS
Usual activities 5.55 3.70 < 0.01 scores in our case (Table 2) – being older, having a univer-
Pain/discomfort 3.70 3.62 < 0.01 sity education, not smoking and being white – were least
Anxiety/depression 5.31 5.56 < 0.01 common amongst the most deprived (Table 1). Unlike
HADS, case more likely our own sample, however, the South African sample con-
Anxiety 3.25 3.56 < 0.01
tained both males and females across the full population
Depression 8.38 6.82 < 0.01
EQ-5D, level decreases age range; its mean age was around 17 years higher than
Usual activities -3.33 -2.15 0.03 was ours. Our explanation of the variance in the cross-sec-
Pain/discomfort -2.84 -3.02 < 0.01 tion model (Table 2) was slightly higher than that of the
Anxiety/depression -3.60 -3.81 < 0.01 South African model (r2 = 0.23).
HADS, case less likely
Anxiety -2.42 -3.02 < 0.01 Our basic approach is also comparable with that of an
Depression -3.56 -2.36 0.02
Israeli study of public perception of health-related quality
Randomised to immediate colposcopy = 1 -1.50 -2.53 0.01
Adjusted r2 0.15 of life [35]. Again, the sample contained both males and
females from across the full population age range, the

Page 6 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2008, 6:94 [Link]

mean age being around 25 years higher than ours. another, Blacks perceived extreme health problems to be
Approximately 2,000 subjects were asked to classify their associated with less disutility than did Hispanics [39]. A
health using the SF-36 quality of life instrument and to Swedish study concluded that differences in self-reported
value it on a numerical scale, 100 to -30, with zero indi- health between native and immigrant populations were
cating "dead". Values regressed on SF-36 scores and other only partially explained by economic and psycho-social
variables indicated that higher economic status, being factors [40]. Cultural differences might well extend
younger and being female were associated with a higher beyond non-monetary health state valuations, given that
value for self-reported health for given SF-36 scores. The significant differences in valuations of risk reduction by
regression explained 52 percent of the variance of the VAS ethnic background have been demonstrated in a contin-
scores, virtually all deriving from the SF-36 domain gent valuation study [41].
scores. The relatively high coefficient of determination is
probably accounted for by the SF-36 being a more com- The presence of anxiety and depression effects in both the
prehensive descriptive system in comparison with the EQ- Table 2 and Table 3 models was perhaps the most surpris-
5D. Its 36 questions combine into eight independent ing result, given that both types of health problem figure
multi-items scales and two summary dimensions. explicitly in the EQ-5D classification instrument. "Anxi-
ety/depression" is one of the five named domains.
Turning towards explanations of particular variables in Although they had been given the opportunity to record
our models, predicting the sign on an age coefficient their anxiety/depression problems directly, individuals
defies intuition. We note that the positive sign on our who were more likely to be suffering from HADS-anxiety
coefficient contrasts with the negative sign identified in and/or HADS-depression recorded EQ VAS scores dispro-
the South African and Israeli models, although this might portionately low in relation to the severity of their prob-
well result from the absence of elderly individuals in our lems as they themselves had classified them. Changes in
sample. Our age coefficient pertains to a narrower age EQ VAS scores were determined by changes in the likeli-
range. It is possible that the cigarette smokers valued their hood of HADS-identified anxiety and/or depression, in
health states lower relative to non-smokers simply by vir- addition to any change in the assigned health state. By
tue of being smokers. The messages of public health pro- inference, the EQ-5D health state description system must
motion initiatives over the past few decades have have been inadequate to represent values of that which
emphasised constantly the damage to health entailed by constituted anxiety and depression to individuals in such
cigarette smoking. "Nearly all smokers regret having circumstances. In respect of our data, it might be felt that
started smoking. Regretful smokers are those who believe the problem arises by virtue of the absence of substantial
themselves to be addicted. These regretful smokers report numbers of subjects exhibiting distress and mood disor-
that smoking has lowered their quality of life and will ders, evidenced by a majority classifying themselves as
continue to do so in the future. Although they are more 11111. However, the coarseness of the anxiety/depression
likely to perceive that there are benefits of quitting, they classification has also been demonstrated for samples of
have tried to quit multiple times, they have failed, and patients wherein the majority were experiencing major
now they fear the future consequences to their health" anxiety disorders and depressive episodes [42,43].
[[36] p.349]. It follows that, if the smoker wants to give up
smoking, then the best imaginable health state entails The coefficients of determination for our regression mod-
being a non-smoker which, by definition, s/he is not. els indicate that the model specifications leave a large part
Non-smokers, of course, face no such impediment when of the variance unexplained. The analysis of residuals sup-
defining their best imaginable health state. ports the belief that the unexplained portion is attributa-
ble to randomness in individual choices. Indeed, an
The lower value placed on health by those with a univer- experiment involving the valuation of hypothetical states
sity education replicates the greater distance between using VAS and time trade off methods concluded that
index and EQ VAS score found for those with longer peri- "individual response patterns (unrelated to age or other
ods of schooling in a US study [37]. Why the possession identifiable respondent characteristics) were the main
of a university degree should influence individuals' evalu- source of 'noise' in the scores" [[44] p.9]. This having been
ation of their own health status levels remains unclear, said, individual response patterns are, in principle, ame-
however. A similar comment can made with respect to nable to psychological analysis, and the inability to detect
ethnicity, although an ethnic influence on both classifica- an explanation might simply point to insufficient data.
tion and valuation has already been identified within the Our models identified two psychological factors explain-
US population. In one US study, Asians were found to be ing individual responses. First, women randomised to a
significantly more likely than Whites to classify them- new, experimental, method of management recorded a
selves as EQ-5D state 11111, even allowing for objective smaller fall in mean EQ VAS score for a given change in
health conditions, education and income [38]. In health state classification. This result is consistent with

Page 7 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2008, 6:94 [Link]

our prior expectation that the self-perceived health of 10. Parkin D, Rice N, Jacoby A, Doughty J: Use of a visual analogue
scale in a daily patient diary: modelling cross-sectional time-
women undergoing a less-preferred method of manage- series data on health-related quality of life. Social Science and
ment, which is, in itself, slower in resolving uncertainties, Medicine 2004, 59:351-360.
would be poorer than those undergoing the more- 11. Bushnell DM, Martin ML, Ricci J-F, Bracco A: Performance of the
EQ-5D in patients with irritable bowel syndrome. Value in
favoured alternative. Second, and again in keeping with Health 2006, 9(2):90-97.
our prior expectation, the quality of self-reported health 12. Teresi JA, Fleishman JA: Differential item functioning and health
for any health state was higher amongst individuals with assessment. Quality of Life Research 2007, 16:33-42.
13. Llach XB, Herdman M, Schiaffino A: Determining correspond-
stronger Internal, and weaker External, loci of control. ence between scores on the EQ-5D "thermometer" and a 5-
Whilst it is likely that part of the variation in VAS scores is point categorical rating scale. Medical Care 1999, 37(3):671-677.
14. Franks P, Muennig P, Lubetkin E, Jia H: The burden of disease asso-
genuinely random, we would nominate personality fac- ciated with being African-American in the United States and
tors, such as extroversion and conscientiousness, as strong the contribution of socio-economic status. Social Science and
candidates to fill at least some of the explanatory void in Medicine 2006, 62:2469-2478.
15. Gebhardt WA, Doef MP van der, Paul LB: The Revised Health
future research. Indeed, personality factors have been Hardiness Inventory (RHHI-24); psychometric properties
shown to be significant predictors of self-perceived health, and relationship with self-reported health and health behav-
ior in two Dutch samples. Health Education Research 2001,
independently of actual health problems [45]. 16(5):579-592.
16. DeNeve KM, Copper H: The happy personality: a meta-analysis
Conclusion of 137 personality traits and subjective well-being. Psychologi-
cal Bulletin 1998, 124(2):197-229.
The results confirm our hypothesis that there exist group 17. Devlin NJ, Hansen P, Selai C: Understanding health state valua-
variables which contribute systematically towards deter- tions: a qualitative analysis of respondents' comments. Qual-
mining EQ VAS scores independently of EQ-5D health ity of Life Research 2004, 13:1265-1277.
18. van Osch SMC, Stiggelbout AM: Understanding VAS valuations:
state classification. In our study, these variables comprised qualitative data on the cognitive process. Quality of Life Research
psychological disposition, socio-demographic factors, 2005, 14:2171-2175.
19. Lindeque BG: Management of cervical premalignant lesions.
management method and clinically-important distress. Best Practice and Research Clinical Obstetrics and Gynaecology 2005,
19(4):545-561.
Competing interests 20. Soutter WP, Fletcher A: Invasive cancer of the cervix in women
with mild dyskaryosis followed up cytologically. British Medical
The authors declare that they have no competing interests. Journal 1994, 308:1421-1423.
21. Flannelly G, Anderson D, Kitchener HC, Mann EMF, Campbell M,
Fisher P, Walker F, Templeton AA: Management of women with
Acknowledgements mild and moderate cervical dyskaryosis. British Medical Journal
The TOMBOLA trial was funded by the United Kingdom Medical Research 1994, 308:1399-1403.
Council, the English National Health Service and the Scottish National 22. Flannelly G, Campbell MK, Meldrum P, Torgerson DJ, Templeton A,
Health Service. Kitchener HC: Immediate colposcopy or cytological surveil-
lance for women with mild dyskariosis: a cost effectiveness
analysis. Journal of Public Health Medicine 1997, 19(4):419-423.
References 23. Jones MH: The management dilemma of the mildly abnormal
1. Rabin R, De Charro F: EQ-5D: a measure of health status from smear: fact or fiction? Annals of the Academy of Medicine of Singa-
the EuroQoL Group. Annals of Medicine 2001, 33(5):337-343. pore 1998, 27:666-670.
2. Greiner W, Weijnen T, Nieuwenhuizen M, Oppe S, Badia X, Bussch- 24. Cotton SC, Sharp L, Little J, Duncan I, Alexander L, Cruickshank ME,
bach J, Buxton M, Dolan P, Kind P, Krabbe P, et al.: A single Euro- Gray NM, Jenkins D, Philips Z, Robertson A, et al.: Trial of manage-
pean currency for EQ-5D health states. Results from a six- ment of borderline and other low-grade abnormal smears
country study. European Journal of Health Economics 2003, (TOMBOLA): trial design. Contemporary Clinical Trials 2006,
4:222-231. 27:449-471.
3. Brooks R, with the EuroQoL Group: EuroQoL: the current state 25. Gray NM, Sharp L, Cotton SC, Masson LF, Little J, Walker LG, Avis
of play. Health Policy 1996, 37:53-72. M, Philips Z, Russell I, Whynes D, et al.: Psychological effects of a
4. Krabbe P, Weijnen T: Guidelines for analysing and reporting low-grade abnormal cervical smear test result: anxiety and
EQ-5D outcomes. In The measurement and valuation of health status associated factors. British Journal of Cancer 2006, 94:1253-1262.
using EQ-5D: a European perspective Edited by: Brooks R, Rabin R, De 26. Herrmann C: International experiences with the Hospital
Charro F. Dordrecht: Kluwer Academic Publishers; 2003:7-19. Anxiety and Depression Scale: a review of validation data
5. de Wit A, Busschbach JJV, de Charro FT: Sensitivity and perspec- and clinical results. Journal of Psychosomatic Research 1997,
tive in the valuation of health status: whose values count? 42(1):17-41.
Health Economics 2000, 9:109-126. 27. Bjelland I, Dahl AA, Haug TT, Neckelmann D: The validity of the
6. Happich M, von Lengerke T: Valuing the health state 'tinnitus': Hospital Anxiety and Depression Scale: an updated litera-
differences between patients and the general public. Hearing ture review. Journal of Psychosomatic Research 2002, 52:69-77.
Research 2005, 207:50-58. 28. Wallston KA: The validity of the Multidimensional Health
7. Johnson JA, Ohinmaa A, Murt B, Sintonen H, Coons SJ: Comparison Locus of Control Scales. Journal of Health Psychology 2005,
of Finnish and U.S.-based visual analog scale valuations of the 10(4):623-631.
EQ-5D measure. Medical Decision Making 2000, 20:281-289. 29. Dolan P, Gudex C, Kind P, Williams A: The time trade-off
8. McPherson K, Myers J, Taylor WJ, McNaughton HK, Weatherall M: method: results from a general population study. Health Eco-
Self-valuation and societal valuations of health state differ nomics 1996, 5:141-154.
with disease severity in chronic and disabling conditions. 30. Dolan P: Modeling valuations for EuroQol health states. Med-
Medical Care 2004, 42(11):1143-1151. ical Care 1997, 35(11):1095-1108.
9. Sandblom G, Carlsson P, Sigsjö P, Varenhorst E: Pain and health- 31. Morgan O, Baker A: Measuring deprivation in England and
related quality of life in a geographically defined population Wales using 2001 Carstairs scores. Health Statistics Quarterly
of men with prostate cancer. British Journal of Cancer 2001, 2006, 31:28-33.
85(4):497-503. 32. King M, Nazareth I, Lampe F, Bower P, Chandler M, Morou M, Sibbald
B, Lai R: Impact of participant and physician intervention pref-

Page 8 of 9
(page number not for citation purposes)
Health and Quality of Life Outcomes 2008, 6:94 [Link]

erences on randomized trials: a systematic review. Journal of

the American Medical Association 2005, 293:1089-1099.
33. Sharp L, Cotton SC, Alexander L, Williams E, Gray NM, Reid JM, on
behalf of the TOMBOLA Group: Reasons for participation and
non-participation in a randomised controlled trial: postal
questionnaire surveys of women eligible for TOMBOLA
(Trial Of Management of Borderline and Other Low-grade
Abnormal smears). Clinical Trials 2006, 3:431-442.
34. Jelsma J, Ferguson G: The determinants of self-reported health-
related quality of life in a culturally and socially diverse South
African community. Bulletin of the World Health Organization 2004,
82:206-212.
35. Mozes B, Maor Y, Shmueli A: Do we know what global ratings of
health-related quality of life measure? Quality of Life Research
1999, 8:269-273.
36. Fong GT, Hammond D, Laux FL, Zanna MP, Cummings KM, Borland
R, Ross H: The near-universal experience of regret among
smokers in four countries: findings from the International
Tobacco Control Policy Evaluation Survey. Nicotine and
Tobacco Research 2004, 6(Supplement 3):S341-S351.
37. Franks P, Lubetkin EI, Melnikov J: Do personal and societal pref-
erences differ by socio-demographic group? Health Economics
2007, 16:319-325.
38. Fu AZ, Kattan MW: Racial and ethnic differences in preference-
based health status measure. Current Medical Research and Opin-
ion 2006, 22(12):2439-2448.
39. Shaw JW, Johnson JA, Chen S, Levin JR, Coons SJ: Racial/ethnic dif-
ferences in preferences for the EQ-5D health states: results
from the US valuation study. Journal of Clinical Epidemiology 2007,
60(5):479-490.
40. Lindström M, Sundquist J, Östergren P-O: Ethnic differences in
self reported health in Malmö in southern Sweden. Journal of
Epidemiology and Community Health 2001, 55:97-103.
41. Whynes DK, Frew EJ, Philips ZN, Covey J, Smith RD: On the
numerical forms of contingent valuation responses. Journal of
Economic Psychology 2007, 28(4):462-476.
42. Supina AL, Johnson JA, Patten SB, Williams JVA, Maxwell CJ: The
usefulness of the EQ-5D in differentiating among persons
with major depressive episode and anxiety. Quality of Life
Research 2007, 16(5):749-754.
43. Günther OH, Roick C, Angermeyer MC, König H-H: The respon-
siveness of EQ-5D utility scores in patients with depression:
a comparison with instruments measuring quality of life, psy-
chopathology and social functioning. Journal of Affective Disorders
2008, 105(1–3):81-91.
44. Essink-Bot MLE, Stuifbergen MC, Meerding WJ, Looman CWN, Bon-
sel GJ, the VOTE group: Individual differences in the use of the
response scale determine valuations of hypothetical health
states: an empirical study. BMC Health Services Research 2007,
7(62):.
45. Goodwin R, Engstrom G: Personality and the perception of
health in the general population. Psychological Medicine 2002,
32:325-332.

Publish with Bio Med Central and every

scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical researc h in our lifetime."
Sir Paul Nurse, Cancer Research UK

Your research papers will be:

available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright

Submit your manuscript here: BioMedcentral

[Link]

Page 9 of 9
(page number not for citation purposes)
 Journal of Clinical Epidemiology 60 (2007) 34e42

Quality criteria were proposed for measurement properties

of health status questionnaires
Caroline B. Terweea,*, Sandra D.M. Bota, Michael R. de Boera,b,
Daniëlle A.W.M. van der Windta,c, Dirk L. Knola,d, Joost Dekkera,e,
Lex M. Boutera, Henrica C.W. de Veta
a
EMGO Institute, VU University Medical Center, Van der Boechorststraat 7, 1081 BT Amsterdam, The Netherlands
b
Department of Ophthalmology, VU University Medical Center, Amsterdam, The Netherlands
c
Department of General Practice, VU University Medical Center, Amsterdam, The Netherlands
d
Department of Clinical Epidemiology and Biostatistics, VU University Medical Center, Amsterdam, The Netherlands
e
Department of Rehabilitation Medicine, VU University Medical Center, Amsterdam, The Netherlands
Accepted 29 March 2006

Abstract
Objectives: Recently, an increasing number of systematic reviews have been published in which the measurement properties of health
status questionnaires are compared. For a meaningful comparison, quality criteria for measurement properties are needed. Our aim was to
develop quality criteria for design, methods, and outcomes of studies on the development and evaluation of health status questionnaires.
Study Design and Setting: Quality criteria for content validity, internal consistency, criterion validity, construct validity, reproducibil-
ity, longitudinal validity, responsiveness, floor and ceiling effects, and interpretability were derived from existing guidelines and consensus
within our research group.
Results: For each measurement property a criterion was defined for a positive, negative, or indeterminate rating, depending on the de-
sign, methods, and outcomes of the validation study.
Conclusion: Our criteria make a substantial contribution toward defining explicit quality criteria for measurement properties of health
status questionnaires. Our criteria can be used in systematic reviews of health status questionnaires, to detect shortcomings and gaps in
knowledge of measurement properties, and to design validation studies. The future challenge will be to refine and complete the criteria
and to reach broad consensus, especially on quality criteria for good measurement properties. Ó 2006 Elsevier Inc. All rights reserved.
Keywords: Reproducibility; Reliability; Validity; Responsiveness; Guidelines; Criteria

1. Introduction criteria for good measurement properties are needed to

legitimize what the best questionnaire is.
The number of available health status questionnaires has
Several articles offer criteria for the evaluation of ques-
increased dramatically over the past decades. Consequently,
tionnaires. Probably the best-known and most comprehen-
the choice of which questionnaire to use is becoming a
sive criteria are those from the Scientific Advisory
major difficulty. Recently a large number of systematic
Committee (SAC) of the Medical Outcomes Trust [12]. The
reviews have been published of available questionnaires SAC defined eight attributes of instrument properties that
measuring a specific concept in a specific population, for
warrant consideration in evaluation. These include (1) con-
example [1e11]. In these systematic reviews, typically,
ceptual and measurement model, (2) validity, (3) reliability,
the content and measurement properties of the available
(4) responsiveness, (5) interpretability, (6) respondent and
questionnaires are compared. In analogy to systematic
administrative burden, (7) alternative forms, and (8) cultural
reviews of clinical trials, criteria are needed to determine
and language adaptations (translations). Within each of these
the methodological quality of studies on the development
attributes, specific criteria were defined by which instru-
and evaluation of health status questionnaires. In addition,
ments should be reviewed. Similar criteria have been defined,
e.g., by Bombardier and Tugwell [13], Andresen [14], and
McDowell and Jenkinson [15]. What is often lacking in these
* Corresponding author. Tel.: þ31-20-4448187; fax: þ31-20-4446775. criteria, however, are explicit criteria for what constitutes
E-mail address: [Link]@[Link] (C.B. Terwee). good measurement properties. For example, for the
0895-4356/06/$ e see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi: 10.1016/[Link].2006.03.012
 C.B. Terwee et al. / Journal of Clinical Epidemiology 60 (2007) 34e42 35

assessment of validity it is often recommended that hypoth- judge the relevance and comprehensiveness of the
eses about expected results should be tested, but no criteria items. For example, a questionnaire developed to
have been defined for how many hypotheses should be con- measure functional status of patients with shoulder
firmed to justify that a questionnaire has good validity. No problems may be less valid to measure functional sta-
criteria have been defined for what constitutes good agree- tus of patients with wrist/hand problems, because
ment (acceptable measurement error), good responsiveness, some items may be less relevant for these patients
or good interpretability, and no criteria have been defined (e.g., lifting above shoulder level), whereas important
for the required sample size of studies assessing measure- items for patients with wrist/hand problems may be
ment properties. missing (e.g., buttoning a shirt). The relevance of
As suggested by the SAC [12], we took on the challenge items may also depend on disease severity. An ade-
to further discuss and refine the available quality criteria for quate description of the target population is therefore
studies on the development and evaluation of health status important for judging the comprehensiveness and the
questionnaires, including explicit criteria for the following applicability of the questionnaire in (other) populations.
measurement properties: (1) content validity, (2) internal e Concepts that the questionnaire is intended to measure.
consistency, (3) criterion validity, (4) construct validity, To judge the suitability of a questionnaire for a specific
(5) reproducibility, (6) responsiveness, (7) floor and ceiling purpose, it is important that authors provide a clear
effects, and (8) interpretability. We used our criteria in two framework of what the overall concept to be measured
systematic reviews comparing the measurement properties is. Relevant concepts can be defined in terms of symp-
of questionnaires for shoulder disability [1] and for visual toms; functioning (physical, psychological, and so-
functioning [4], and revised them based on our experiences cial); general health perceptions; or overall quality of
in these reviews. Our criteria can also be used to detect life [18]. These different outcome levels should clearly
shortcomings and gaps in knowledge of measurement be distinguished and measured by separate subscales.
properties, and to design validation studies. For physical functioning it is important to distinguish
In this article we define our quality criteria for measure- between capacity (what a patient thinks he can do)
ment properties, discuss the difficult and sometimes arbi- and performance (what a patient actually does).
trary choices we made, and indicate future challenges. e Item selection and item reduction. The methods for
We emphasize that, just like the criteria offered by the item selection, item reduction, and the execution of
SAC and others, our criteria are open to further discussion a pilot study to examine the readability and compre-
and refinement. Our aim is to contribute to the development hension should be justified and reported. Items in the
of explicit quality criteria for the design, methods, and out- questionnaire must reflect areas that are important to
comes of studies on the development and evaluation of the target population that is being studied. Therefore,
health status questionnaires. the target population should be involved during item
selection. In some guidelines it is recommended that
developers start with a large number of items and apply
2. Content validity item reduction techniques to select a small number of
final items. This strategy, however, does not guarantee
Content validity examines the extent to which the con-
a better content validity, because a comprehensive set
cepts of interest are comprehensively represented by the
of items can also be achieved without item reduction.
items in the questionnaire [16]. To be able to rate the
Therefore, we do not consider this to be mandatory.
quality of a questionnaire, authors should provide a clear
e Interpretability of the items. Completing the question-
description of the following aspects regarding the develop-
naire should not require reading skills beyond that of
ment of a questionnaire:
a 12-year-old to avoid missing values and unreliable
e Measurement aim of the questionnaire, i.e., discrimi- answers [19]. That means that items should be short
native, evaluative, or predictive [17]. The measure- and simple and should not contain difficult words or
ment aim is important, because different items may jargon terms. Moreover, items should not consist of
be valid for different aims. For example, a question two questions at the same time [19]. Furthermore,
on stiffness could be a valid item of a discriminative the time period to which the questions refer should
questionnaire used to measure the impact of osteoar- be clearly stated and justified.
thritis on quality of life (to distinguish between pa-
tients with different levels of quality of life), but We give a positive rating for content validity if a clear
would be considered invalid for an evaluative ques- description is provided of the measurement aim, the target
tionnaire used as an outcome measure in a pain med- population, the concepts that are being measured, and the
ication trial, because it is unlikely to be changed by item selection. Furthermore, the target population should
pain medication. have been involved during item selection, as well as either
e Target population, i.e., the population for which the investigators or experts. If a clear description is lacking,
questionnaire was developed. This is important to content validity is rated as indeterminate.
36 C.B. Terwee et al. / Journal of Clinical Epidemiology 60 (2007) 34e42

3. Internal consistency a positive rating for criterion validity if convincing argu-

ments are presented that the used standard really is ‘‘gold’’
Internal consistency is a measure of the extent to which
and if the correlation with the gold standard is at least 0.70.
items in a questionnaire (sub)scale are correlated (homoge-
neous), thus measuring the same concept. Internal consis-
tency is an important measurement property for 5. Construct validity
questionnaires that intend to measure a single underlying Construct validity refers to the extent to which scores on
concept (construct) by using multiple items. In contrast, a particular instrument relate to other measures in a manner
for questionnaires in which the items are merely different that is consistent with theoretically derived hypotheses con-
aspects of a complex clinical phenomenon that do not have cerning the concepts that are being measured [17,19]. Con-
to be correlated, such as in the Apgar Scale [20], internal struct validity should be assessed by testing predefined
consistency is not relevant [21,22]. hypotheses (e.g., about expected correlations between mea-
An internally consistent (homogeneous or unidimen- sures or expected differences in scores between ‘‘known’’
sional) scale is achieved through good construct definitions, groups). These hypotheses need to be as specific as possi-
good items, then principal component analysis or exploratory ble. Without specific hypotheses, the risk of bias is high be-
factor analysis, followed by confirmatory factor analysis. cause retrospectively it is tempting to think up alternative
When internal consistency is relevant, principal component explanations for low correlations instead of concluding that
analysis or factor analysis should be applied to determine the questionnaire may not be valid. We therefore give a pos-
whether the items form only one overall scale (dimension) itive rating for construct validity if hypotheses are specified
or more than one [23,24]. In case that there is no prior hypoth- in advance and at least 75% of the results are in correspon-
esis regarding the dimensionality of a questionnaire, explor- dence with these hypotheses, in (sub)groups of at least 50
atory principal component analysis or factor analyses can be patients.
applied. But if there is a clear hypothesis regarding the factor
structure, e.g., because of an existing theoretical model or be-
6. Reproducibility
cause the factor structure has been determined previously,
confirmatory factor analysis should be used [25,26]. The Reproducibility concerns the degree to which repeated
number of subjects included in a factor analysis is a matter measurements in stable persons (testeretest) provide simi-
of debate. Rules-of-thumb vary from four to 10 subjects lar answers. We believe that it is important to make a dis-
per variable, with a minimum number of 100 subjects to en- tinction between reliability and agreement [29,30].
sure stability of the varianceecovariance matrix [27]. Agreement concerns the absolute measurement error, i.e.,
After determining the number of (homogeneous) (sub)- how close the scores on repeated measures are, expressed
scales, Cronbach’s alpha should be calculated for each in the unit of the measurement scale at issue. Small mea-
(sub)scale separately. Cronbach’s alpha is considered an ad- surement error is required for evaluative purposes in which
equate measure of internal consistency. A low Cronbach’s one wants to distinguish clinically important changes from
alpha indicates a lack of correlation between the items in measurement error. Reliability concerns the degree to
a scale, which makes summarizing the items unjustified. which patients can be distinguished from each other, de-
A very high Cronbach’s alpha indicates high correlations spite measurement error [19]. High reliability is important
among the items in the scale, i.e., redundancy of one or for discriminative purposes if one wants to distinguish
more items. Furthermore, a very high Cronbach’s alpha is among patients, e.g., with more or less severe disease (as
usually found for scales with a large number of items, be- in diagnostic applications). Reliability coefficients (intra-
cause Cronbach’s alpha is dependent upon the number of class correlation coefficients (ICC)) concern the variation
items in a scale. Note that Cronbach’s alpha gives no infor- in the population (interindividual variation) divided by
mation on the number of subscales in a questionnaire, be- the total variation, which is the interindividual variation
cause alpha can be high when two or more subscales plus the intraindividual variation (measurement error),
with high alphas are combined. Nunnally and Bernstein expressed as a ratio between 0 and 1.
[28] proposed a criterion of 0.70e0.90 as a measure of The time period between the repeated administrations
good internal consistency. In our experience, however, should be long enough to prevent recall, though short
manydin our view goodd(subscales of) questionnaires enough to ensure that clinical change has not occurred. Of-
have higher Cronbach’s alphas. We give a positive rating ten, 1 or 2 weeks will be appropriate, but there could be
for internal consistency when factor analysis was applied reasons to choose otherwise. Therefore, we do not rate
and Cronbach’s alpha is between 0.70 and 0.95. the appropriateness of the time period, but only require that
this time period is described and justified.

4. Criterion validity 6.1. Agreement

Criterion validity refers to the extent to which scores on The measurement error can be adequately expressed
a particular instrument relate to a gold standard. We give as the standard error of measurement (SEM) [30]. The
 C.B. Terwee et al. / Journal of Clinical Epidemiology 60 (2007) 34e42 37

SEM equals the square root of the error variance of an AN- used, the weighted Kappa coefficient is identical to the
OVA analysis, either including systematic differences ICCagreement [19].
(SEMagreement) or excluding them (SEMconsistency). Many Often 0.70 is recommended as a minimum standard for
authors fail to describe how they calculated the SEM. We reliability [28]. We give a positive rating for reliability
believe that systematic differences should be considered when the ICC or weighted Kappa is at least 0.70 in a sample
part of the measurement error, because we want to distin- size of at least 50 patients.
guish them from ‘‘real’’ changes, e.g., due to treatment.
Therefore, we prefer SEMagreement. The SEM can be con- 7. Responsiveness
verted into the smallest detectable change
(SDC 5 1.96  O2  SEM), which reflects the smallest Responsiveness has been defined as the ability of a ques-
within-person change in score that, with P ! 0.05, can be tionnaire to detect clinically important changes over time,
interpreted as a ‘‘real’’ change, above measurement error, even if these changes are small [37]. A large number of def-
in one individual (SDCind) [31,32]. The SDC measurable initions and methods were proposed for assessing respon-
in a group of people (SDCgroup) can be calculated by siveness [38]. We consider responsiveness to be a measure
dividing the SDCind by On [32,33]. of longitudinal validity. In analogy to construct validity, lon-
Another adequate parameter of agreement is described gitudinal validity should be assessed by testing predefined
by Bland and Altman [34]. Their limits of agreement equal hypotheses, e.g., about expected correlations between
the mean change in scores of repeated measurements changes in measures, or expected differences in changes be-
(meanchange) 6 1.96  standard deviation of these changes tween ‘‘known’’ groups [38]. This shows the ability of
(SDchange). The limits of agreement are often reported a questionnaire to measure changes if they really have hap-
because they are easily interpretable. Note that SDchange pened. Futhermore, the instrument should be able to distin-
equals O2  SEMconsistency. guish clinically important change from measurement error.
For evaluative purposes, the absolute measurement error Responsiveness should therefore be tested by relating the
should be smaller than the minimal amount of change in the SDC to the MIC, as described under agreement (see Section
(sub)scale that is considered to be important (minimal im- 6.1). This approach equals Guyatt’s responsiveness ratio
portant change (MIC)). Therefore, the MIC of a (sub)scale (RR), in which the clinically important change (MIC) is re-
should be defined (see under interpretability). lated to the between-subject variability in within-subject
We give a positive rating for agreement if the SDC changes in stable subjects (SDchange; the same as in the limits
(SDCind for application in individuals and SDCgroup for use of agreement) [39]. The RR should thus be at least 1.96 (at
in groups) or the limits of agreement (upper or lower limit, the value of 1.96 the MIC equals the SDCind, which is
depending on whether the interest is in improvement or de- 1.96  SDchange). Another adequate measure of responsive-
terioration) are smaller than the MIC. Because this is a rela- ness is the area under the receiver operating characteristics
tively new approach and not yet commonly presented, we (ROC) curve (AUC) [40], which is a measure of the ability
also give a positive rating if authors provide convincing argu- of a questionnaire to distinguish patients who have and have
ments (e.g., based on their experience with the interpretation not changed, according to an external criterion. We consider
of the questionnaire scores) that the agreement is acceptable. an AUC of at least 0.70 to be adequate.
In both cases, we consider a sample size of at least 50 pa-
tients adequate for the assessment of the agreement parame- 8. Floor or ceiling effects
ter, based on a general guideline by Altman [35].
Floor or ceiling effects are considered to be present if
more than 15% of respondents achieved the lowest or high-
6.2. Reliability
est possible score, respectively [41]. If floor or ceiling ef-
The ICC is the most suitable and most commonly used fects are present, it is likely that extreme items are
reliability parameter for continuous measures. Many au- missing in the lower or upper end of the scale, indicating
thors fail to describe which ICC they have used, e.g., an limited content validity. As a consequence, patients with
ICC for consistency (ICCconsistency) or an ICC for agree- the lowest or highest possible score cannot be distinguished
ment (ICCagreement) [19,36]. Because systematic differences from each other, thus reliability is reduced. Furthermore, the
are considered to be part of the measurement error, responsiveness is limited because changes cannot be mea-
ICCagreement (two-way random effects model, or ICC sured in these patients. We give a positive rating for (the ab-
(A,1) according to McGraw and Wong [36]) is preferred. sence of) floor and ceiling effects if no floor or ceiling
The Pearson correlation coefficient is inadequate, because effects are present in a sample size of at least 50 patients.
systematic differences are not taken into account [19].
For ordinal measures, the weighted Cohen’s Kappa coeffi-
9. Interpretability
cient should be used. The absolute percentage of agreement
is inadequate, because it does not adjust for the agreement Interpretability is defined as the degree to which one can
attributable to chance. When quadratic weights are being assign qualitative meaning to quantitative scores [42].
38 C.B. Terwee et al. / Journal of Clinical Epidemiology 60 (2007) 34e42

Investigators should provide information about what extremely heterogeneous study population, the evaluated
(change in) score would be clinically meaningful. Various measurement properties are also rated as indeterminate.
types of information can aid in interpreting scores on
a questionnaire: (1) means and SD of scores of (subgroups
11. Overview table
of) a reference population (norm values); (2) means and SD
of scores of relevant subgroups of patients who are ex- In the final comparison of the measurement properties of
pected to differ in scores (e.g., groups with different clinical different questionnaires, one has to consider all ratings to-
diagnoses, age groups, gender groups, primary vs. second- gether when choosing between different questionnaires.
ary care setting); (3) means and SD of scores of patients be- We recommend to compose a table that provides an over-
fore and after treatment(s) of known efficacy [43]; and (4) view of all ratings, such as the example given in Table 2.
means and SD of scores of subgroups of patients based on In Table 2 the results are presented from our systematic re-
patients’ global ratings of change. For example, a positive view of all questionnaires measuring disability in patients
rating was given if mean scores and SD are presented of with shoulder complaints (because there is no gold standard
at least four subgroups. For example, if means and SD for disability, criterion validity was not assessed) [1]. In
are presented for a general population (norm values), Table 2 all ratings for each questionnaire are presented sep-
stratified by gender and age groups. In addition, an MIC arately for each specific population or setting. For example,
should be defined to enable interpretation of change scores the Shoulder Pain and Disability Index (SPADI) was eval-
over time and sample size calculations. The MIC has been uated in several populations. Construct validity and respon-
defined as ‘‘the smallest difference in score in the domain siveness were rated positively for outpatients (c), but rated
of interest which patients perceive as beneficial and would as indeterminate for primary care patients (b) and hospital
mandate, in the absence of troublesome side effects and ex- patients (d). With this table one can make an evidence-
cessive cost, a change in the patient’s management’’ [44]. based choice for the questionnaire with the best measure-
Various distribution-based and anchor-based methods have ment properties, taking into account those measurement
been proposed. Anchor-based approaches use an external properties that are most important for a specific application
criterion to operationalize an important change. Distribu- (e.g., reliability when using a questionnaire for discrimina-
tion-based approaches are based on statistical characteris- tion and responsiveness when using it for evaluation of
tics of the sample [45]. We recommend an anchor-based a treatment effect) and the population and setting in which
method, to determine the MIC because distribution-based the questionnaire is going to be used.
methods do not provide a good indication of the importance
of the observed change. We consider a sample size of at
12. Discussion
least 50 patients adequate to determine the MIC.
We developed quality criteria for the design, methods,
and outcomes of studies on the development and evaluation
10. Population-specific ratings of measurement
of health status questionnaires. Nine measurement proper-
properties
ties were distinguished: content validity, internal consis-
A summary of the criteria for measurement properties of tency, criterion validity, construct validity, reproducibility,
health status questionnaires is presented in Table 1. Each longitudinal validity, responsiveness, floor and ceiling
property is rated as positive, negative, or indeterminate, de- effects, and interpretability.
pending on the design, methods, and outcomes of the study. Our criteria are mostly opinion based because there is no
Measurement properties differ between populations and empirical evidence in this field to support explicit quality
settings. Therefore, the evaluation of all measurement prop- criteria. They are useful rules of thumb, but other investiga-
erties needs to be conducted in a population and setting that tors may want to make their own choices.
is representative for the population and setting in which the We did not summarize the quality criteria into one over-
questionnaire is going to be used. The setting refers to the all quality score, as is often done in systematic reviews of
testing conditions, e.g., self-completed or interview, and lan- randomized clinical trials [46]. An overall quality score as-
guage. A clear description of the design of each individual sumes that all measurement properties are equally impor-
study has to be provided, including population (diagnosis tant, which is probably not the case. We consider content
and clinical features, age, and gender); design (e.g., lan- validity as one of the most important measurement proper-
guage version, time between the measurements, completion ties. Only if the content validity of a questionnaire is ade-
before or after treatment); testing conditions (e.g., question- quate, one will consider using the questionnaire, and
naires completed at home or in a waiting room, self of in in- evaluation of the other measurement properties is useful.
terview); and analyses of the data. If a clear description of Furthermore, the aim of the questionnaire demands differ-
the design of the study is lacking, the evaluated measure- ent qualities of the questionnaire with respect to reproduc-
ment properties are rated as indeterminate. In addition, if ibility and responsiveness. Discriminative questionnaires
any important methodological weakness in the design or require a high level of reliability to be able to distinguish
execution of the study is found, e.g., selection bias or an between persons. Evaluative questionnaires require a high
 C.B. Terwee et al. / Journal of Clinical Epidemiology 60 (2007) 34e42 39

Table 1
Quality criteria for measurement properties of health status questionnaires
Property Definition Quality criteriaa,b
1. Content validity The extent to which the domain of þA clear description is provided of the measurement aim, the target population,
interest is comprehensively sampled by the concepts that are being measured, and the item selection AND target popu-
the items in the questionnaire lation and (investigators OR experts) were involved in item selection;
?A clear description of above-mentioned aspects is lacking OR only target
population involved OR doubtful design or method;
No target population involvement;
0No information found on target population involvement.

2. Internal The extent to which items in a (sub)scale þFactor analyses performed on adequate sample size (7 * # items and >100)
consistency are intercorrelated, thus measuring the AND Cronbach’s alpha(s) calculated per dimension AND Cronbach’s alpha(s)
same construct between 0.70 and 0.95;
?No factor analysis OR doubtful design or method;
Cronbach’s alpha(s) !0.70 or O0.95, despite adequate design and method;
0No information found on internal consistency.

3. Criterion validity The extent to which scores on a þConvincing arguments that gold standard is ‘‘gold’’ AND correlation
particular questionnaire relate to a gold with gold standard >0.70;
standard ?No convincing arguments that gold standard is ‘‘gold’’ OR doubtful design or
method;
Correlation with gold standard !0.70, despite adequate design and method;
0No information found on criterion validity.

4. Construct validity The extent to which scores on a þSpecific hypotheses were formulated AND at least 75% of the results are in
particular questionnaire relate to other accordance with these hypotheses;
measures in a manner that is consistent ?Doubtful design or method (e.g., no hypotheses);
with theoretically derived hypotheses Less than 75% of hypotheses were confirmed, despite adequate design and
concerning the concepts that are being methods;
0No information found on construct validity.
measured

5. Reproducibility
5.1. Agreement The extent to which the scores on þMIC ! SDC OR MIC outside the LOA OR convincing arguments that agreement
repeated measures are close to each is acceptable;
other (absolute measurement error) ?Doubtful design or method OR (MIC not defined AND no
convincing arguments that agreement is acceptable);
MIC > SDC OR MIC equals or inside LOA, despite adequate design and method;
0No information found on agreement.
5.2. Reliability The extent to which patients can be þICC or weighted Kappa > 0.70;
distinguished from each other, despite ?Doubtful design or method (e.g., time interval not mentioned);
measurement errors ICC or weighted Kappa ! 0.70, despite adequate design and method;
(relative measurement error) 0No information found on reliability.

6. Responsiveness The ability of a questionnaire to detect þSDC or SDC ! MIC OR MIC outside the LOA OR RR O 1.96 OR
clinically important changes over time AUC > 0.70;
?Doubtful design or method;
SDC or SDC > MIC OR MIC equals or inside LOA OR
RR < 1.96 OR AUC ! 0.70, despite adequate design and methods;
0No information found on responsiveness.

7. Floor and ceiling The number of respondents who þ<15% of the respondents achieved the highest or lowest possible scores;
effects achieved the lowest or highest possible ?Doubtful design or method;
score O15% of the respondents achieved the highest or lowest possible scores,
despite adequate design and methods;
0No information found on interpretation.

8. Interpretatability The degree to which one can assign þMean and SD scores presented of at least four relevant subgroups of patients
qualitative meaning to quantitative and MIC defined;
scores ?Doubtful design or method OR less than four subgroups OR no MIC defined;
0No information found on interpretation.
MIC 5 minimal important change; SDC 5 smallest detectable change; LOA 5 limits of agreement; ICC 5 Intraclass correlation; SD, standard deviation.
a
þ 5 positive rating; ? 5 indeterminate rating;  5 negative rating; 0 5 no information available.
b
Doubtful design or method 5 lacking of a clear description of the design or methods of the study, sample size smaller than 50 subjects (should be at
least 50 in every (subgroup) analysis), or any important methodological weakness in the design or execution of the study.
40 C.B. Terwee et al. / Journal of Clinical Epidemiology 60 (2007) 34e42

Table 2
Summary of the assessment of the measurement properties of all questionnaires measuring disability in patients with shoulder complaints [1]
Reproducibility
Content Internal Construct Floor or
Questionnaire validity consistency validity Agreement Reliability Responsiveness ceiling effect Interpretability
SDQ-UK þ 0 þ 0 0 0 þ (b);  (a) 0
SIQ þ ? þ þ 0 þ þ þ
OSQ þ ? þ þ 0 þ þ þ
SDQ-NL ? 0 þ 0 0 þ (b); ? (b)  (b) þ
RC-QOL þ 0 þ ? 0 0 þ ?
DASH þ 0 þ (c,d); ? (c) þ (c) þ (c) þ (c); ? (d) þ (c) þ
WOSI þ 0 þ 0 þ ? 0 0
SSRS  0 ? (c) ? (d) ? (d) ? (d) þ (d) þ
SRQ þ ? ? ? ? ? 0 ?
SST þ ? þ (c); ? (d) ? (c,d) ? (d) ? (c,d) þ (c) þ
WOOS þ 0 ? 0 ? ? 0 0
SSI 0 0 þ (c) ? (d) ? (d) ? (d) þ (c) 0
UEFS  þ ? 0 0 ? þ þ
ASES  ? þ (c); ? (d) ? (d) ? (c,d) ? (c,d) þ (d); ? (c) þ
SPADI  ? þ (c); ? (b,c) þ (c); ? (d) ? (c,d) þ (c); ? (b,d) þ (b,c) þ
UEFL  0 þ 0 0 0  (a) 0
Rating: þ 5 positive; 0 5 intermediate; - 5 poor; ? 5 no information available.
Kinds of study population(s) used in the studies: (a) community, (b) primary care, (c) outpatients’ clinic, and (d) hospital patients.
SDQ-UK 5 Shoulder Disability Questionnaire (English version); SIQ 5 Shoulder Instability Questionnaire; OSQ (Oxford) 5 Shoulder Questionnaire;
SDQ-NL 5 Shoulder Disability Questionnaire (Dutch version); RC-QOL 5 Rotator Cuff Quality-of-Life Measure; DASH 5 Disabilities of the Arm, Shoul-
der and Hand Scale; WOSI 5 Western Ontario Shoulder Instability Index; SSRS 5 Subjective Shoulder Rating Scale; SST 5 Simply Shoulder Test;
SSI 5 Shoulder Severity Index; UEFS 5 Upper Extremity Function Scale; ASES 5 American Shoulder and Elbow Surgeons Standardized Shoulder Assess-
ment Form.

level of agreement to be able to measure important of studies on the development and evaluation of health sta-
changes. Evaluative questionnaires should be responsive tus questionnaires therefore needs to be improved.
to change, whereas discriminative questionnaires do not In applying our criteria in two systematic reviews [1,4],
necessarily need to be responsive to change. we found that several measurement properties are often not
We recommend to always compose a table that provides properly assessed or analyzed, nor clearly reported. Impor-
an overview of all ratings of all questionnaires, such as Ta- tant information on content validity is often very limited.
ble 2, which facilitates an assessment of all ratings together The aim of a questionnaire, i.e., whether the questionnaire
when choosing the most suitable questionnaire for a specific was designed for discriminative or evaluative purposes, is
application. Two important issues should be kept in mind. often not described, and the concepts that the questionnaire
Firstly, with our approach, poor quality questionnaires is intended to measure are often ill defined. Furthermore,
can be given positive ratings for some measurement proper- item selection is often poorly described. A statement like
ties. For example, the Upper Extremity Functional Limita- ‘‘A preliminary questionnaire was developed and com-
tion Scale (UEFL) received a positive rating for construct pleted by 30 patients.. A subset of these patients was in-
validity despite a negative rating for content validity. By terviewed and each question was assessed for clinical
considering all ratings together, one may decide for exam- relevance, importance, and ease of completion’’ [47] does
ple to choose the Shoulder Rating Questionnaire (SRQ) or not justify that the items comprehensively represent all is-
the Western Ontario Osteoarthritis of the Shoulder Index sues that are important to the target population. This ham-
(WOOS), with positive ratings for content validity, over pers judgment about the applicability of the questionnaire
the UEFL, despite the fact that these questionnaires re- in a given population.
ceived indeterminate ratings for construct validity. Sec- One would assume that the number of scales corre-
ondly, questionnaires with the highest number of positive sponds with the number of dimensions identified in factor
ratings are not necessarily the best questionnaires. The rat- analysis, but this is often not the case. We found that sev-
ings depend on the availability of information and the qual- eral questionnaires claimed to cover more than one dimen-
ity of reporting on the assessment of the measurement sion, but consisted of one scale only, or vice versa [1].
properties. For example, a questionnaire can be given many Many authors fail to specify hypotheses for the assess-
indeterminate ratings if measurement properties are not yet ment of construct validity or the hypotheses are not very
evaluated because the questionnaire is newly developed. specific or informative. For example, to validate a shoulder
Furthermore, poorly reported validation studies will lead disability questionnaire, the authors tested the hypothesis
to low ratings for questionnaires that are not necessarily that ‘‘restriction of shoulder movement on examination
poor in design or performance. The quality of reporting correlates with disability score’’ [48]. An informative
 C.B. Terwee et al. / Journal of Clinical Epidemiology 60 (2007) 34e42 41

hypothesis should evaluate, as directly as possible, specific References

claims made for a theory [49], and should thus include a jus-
[1] Bot SD, Terwee CB, van der Windt DA, Bouter LM, Dekker J, de
tification of the expected direction and magnitude of the Vet HC. Clinimetric evaluation of shoulder disability questionnaires:
correlation. Without testing specific hypotheses, the risk a systematic review of the literature. Ann Rheum Dis 2004;63:335e41.
of bias is high. Practically all authors of instrument evalu- [2] Jorstad EC, Hauer K, Becker C, Lamb SE. Measuring the psycholog-
ation studies conclude that their instrument is valid, ical outcomes of falling: a systematic review. J Am Geriatr Soc
2005;53:501e10.
whereas objective counting of the number of hypotheses
[3] Daker-White G. Reliable and valid self-report outcome measures in
that were confirmed frequently indicates otherwise [1,4]. sexual (dys)function: a systematic review. Arch Sex Behav 2002;31:
Responsiveness is often ill defined and not well as- 197e209.
sessed. For example, many authors use multiple indices [4] de Boer MR, Moll AC, de Vet HC, Terwee CB, Volker-Dieben HJ,
of responsiveness and conclude that their instrument is re- van Rens GH. Psychometric properties of vision-related quality of
life questionnaires: a systematic review. Ophthalmic Physiol Opt
sponsive when all indices point in the same direction
2004;24:257e73.
[50e52]. Different authors, however, have pointed to the [5] Edwards B, Ung L. Quality of life instruments for caregivers of
conceptual and algebraic differences between different in- patients with cancer. Cancer Nurs 2002;25:342e9.
dices of responsiveness, showing that different indices [6] Garratt AM, Brealey S, Gillespie WJ. Patient-assessed health instru-
may lead to different conclusions [53,54]. Furthermore, ment for the knee: a structured review. Rheumatology 2004;43:
1414e23.
many indices of responsiveness should actually be looked
[7] Hallin P, Sullivan M, Kreuter M. Spinal cord injury and quality of life
upon as measures of the magnitude of a treatment effect, measures: a review of instrument psychometric quality. Spinal Cord
which, in itself, tell us nothing about the quality of the in- 2000;38:509e23.
strument to serve its purpose [38]. [8] Haywood KL, Garratt AM, Fitzpatrick R. Quality of life in older peo-
We recommend using criteria like ours as an aid in the ple: a structured review of generic self-assessed health instruments.
Qual Life Res 2005;14:1651e68.
design and reporting of studies on the development and
[9] Haywood KL, Garratt AM, Dawes PT. Patient-assessed health in an-
evaluation of health status questionnaires. This should lead kylosing spondylitis: a structured review. Rheumatology (Oxford)
to an improvement in the quality of (reporting of) such 2005;44:577e86.
studies. [10] Dziedzic KS, Thomas E, Hay EM. A systematic search and critical
review of measures of disability for use in a population survey of
hand osteoarthritis (OA). Osteoarthritis Cartilage 2005;13:1e12.
13. Future challenges [11] Ettema TP, Droes RM, de Lange J, Mellenbergh GJ, Ribbe MW.
A review of quality of life instruments used in dementia. Qual Life
One might argue that our criteria are not discriminative Res 2005;14:675e86.
enough to distinguish between good and very high-quality [12] Scientific Advisory Committee of the Medical Outcomes Trust.
questionnaires. This would be important when many Assessing health status and quality-of-life instruments: attributes
high-quality questionnaires are available, but in our experi- and review criteria. Qual Life Res 2002;11:193e205.
[13] Bombardier C, Tugwell P. Methodological considerations in func-
ence, within the field of health status and health-related
tional assessment. J Rheumatol 1987;14(Suppl 15):6e10.
quality of life measurement, this is not (yet) the case. [14] Andresen EM. Criteria for assessing the tools of disability outcomes
Therefore, we believe that our criteria work well to separate research. Arch Phys Med Rehabil 2000;81(Suppl 2):S15e20.
the wheat from the chaff. The next step would be to further [15] McDowell I, Jenkinson C. Development standards for health mea-
refine and complete the criteria, e.g., by deciding how spe- sures. J Health Serv Res Policy 1996;1:238e46.
[16] Guyatt GH, Feeny DH, Patrick DL. Measuring health related quality
cific the hypotheses for testing construct validity or respon-
of life. Ann Intern Med 1993;118:622e9.
siveness should be, and by including criteria for the [17] Kirschner B, Guyatt G. A methodological framework for assessing
methods and results of studies using Item Response Theory health indices. J Chronic Dis 1985;38:27e36.
(IRT) models. Furthermore, broad consensus is needed, es- [18] Wilson IB, Cleary PD. Linking clinical variables with health-related
pecially on the criteria for good measurement properties. quality of life. JAMA 1995;273:59e65.
[19] Streiner DL, Norman GR. Health measurement scales. A practical
Because the number of health status questionnaires is
guide to their development and use. New York: Oxford University
rapidly growing, choosing the right questionnaire for a spe- Press; 2003.
cific purpose becomes a time-consuming and difficult task. [20] Apgar V. A proposal for new method of evaluation of the newborn
An increasing number of systematic reviews are being pub- infant. Curr Res Anesth Analg 1953;32:260e7.
lished in which the measurement properties of health status [21] Fayers PM, Hand DJ. Causal variables, indicator variables and mea-
surement scales: an example from quality of life. J R Stat Soc A
questionnaires are being evaluated and compared. These
2002;165:233e61.
systematic reviews are important tools for evidence-based [22] Streiner DL. Being inconsistent about consistency: when coefficient
instrument selection. Explicit quality criteria for studies alpha does and doesn’t matter. J Pers Assess 2003;80:217e22.
on the development and evaluation of health status ques- [23] Floyd FJ, Widaman KF. Factor analysis in the development and re-
tionnaires are needed to legitimize what the best question- finement of clinical assessment instruments. Psychol Assess
1995;7:286e99.
naire is. Our criteria are a step in this direction. The future
[24] Streiner DL. Figuring out factors: the use and misuse of factor anal-
challenge will be to refine and complete the criteria and to ysis. Can J Psychiatry 1994;39:135e40.
reach broad consensus, especially on the quality criteria for [25] Bollen KA. Structural equations with latent variables. New York:
good measurement properties. Wiley; 1989.
42 C.B. Terwee et al. / Journal of Clinical Epidemiology 60 (2007) 34e42

[26] de Vet HCW, Ader HJ, Terwee CB, Pouwer F. Are factor analytical [41] McHorney CA, Tarlov AR. Individual-patient monitoring in clinical
techniques appropriately used in the validation of health status ques- practice: are available health status surveys adequate? Qual Life
tionnaires? A systematic review on the quality of factor analyses of Res 1995;4:293e307.
the SF-36. Qual Life Res 2005;14:1203e18. [42] Lohr KN, Aaronson NK, Alonso J, Burnam MA, Patrick DL,
[27] Kline P. The handbook of psychological testing. London: Routledge; Perrin EB, et al. Evaluating quality of life and health status instru-
1993. ments: development of scientific review criteria. Clin Ther 1996;18:
[28] Nunnally JC, Bernstein IH. Psychometric theory. 3rd ed. New York: 979e92.
McGraw-Hill; 1994. [43] Guyatt GH, Bombardier C, Tugwell PX. Measuring disease-specific
[29] de Vet HCW. Observer reliability and agreement. In: Armitage P, quality of life in clinical trials. CMAJ 1986;134:889e95.
Colton T, editors. Encyclopedia of biostatistics. Boston: John Wiley [44] Jaeschke R, Singer J, Guyatt GH. Measurement of health status.
& Sons Ltd.; 1998. p. 3123e8. Ascertaining the minimal clinically important difference. Control
[30] Stratford P. Reliability: consistency or differentiating among sub- Clin Trials 1989;10:407e15.
jects? Phys Ther 1989;69:299e300. [45] Crosby RD, Kolotkin RL, Williams GR. Defining clinically meaning-
[31] Beckerman H, Roebroeck ME, Lankhorst GJ, Becher JG, ful change in health-related quality of life. J Clin Epidemiol 2003;56:
Bezemer PD, Verbeek AL. Smallest real difference, a link between 395e407.
reproducibility and responsiveness. Qual Life Res 2001;10:571e8. [46] Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ,
[32] de Vet HCW, Bouter LM, Bezemer PD, Beurskens AJ. Reproducibil- Gavaghan DJ, et al. Assessing the quality of reports of randomized
ity and responsiveness of evaluative outcome measures. Theoretical clinical trials: is blinding necessary? Control Clin Trials 1996;17:
considerations illustrated by an empirical example. Int J Technol 1e12.
Assess Health Care 2001;17:479e87. [47] L’Insalata JC, Warren RF, Cohen SB, Altchek DW, Peterson MG.
[33] de Boer MR, de Vet HCW, Terwee CB, Moll AC, Völker-Dieben HJM, A self-administered questionnaire for assessment of symptoms and
van Rens GHMB. Change to the subscales of two vision-related quality function of the shoulder. J Bone Joint Surg Am 1997;79:738e48.
of life questionnaires are proposed. J Clin Epidemiol 2005;58:1260e8. [48] Croft P, Pope D, Zonca M, O’Neill T, Silman A. Measurement of
[34] Bland JM, Altman DG. Statistical methods for assessing agreement shoulder related disability: results of a validation study. Ann Rheum
between two methods of clinical measurement. Lancet 1986;307e10. Dis 1994;53:525e8.
[35] Altman DG. Practical statistics for medical research. London: Chap- [49] Strauss ME. Introduction to the special section on construct validity
man and Hall; 1991. of psychological tests: 50 years after Cronbach and Meehl (1955).
[36] McGraw KO, Wong SP. Forming inferences about some intraclass Psychol Assess 2005;17:395.
correlation coefficients. Psychol Methods 1996;1:30e46. [50] Beaton DE, Hogg-Johnson S, Bombardier C. Evaluating changes in
[37] Guyatt GH, Deyo RA, Charlson M, Levine MN, Mitchell A. Respon- health status: reliability and responsiveness of five generic health
siveness and validity in health status measurement: a clarification. status measures in workers with musculoskeletal disorders. J Clin
J Clin Epidemiol 1989;42:403e8. Epidemiol 1997;50:79e93.
[38] Terwee CB, Dekker FW, Wiersinga WM, Prummel MF, Bossuyt PM. [51] Stucki G, Liang MH, Fossel AH, Katz JN. Relative responsiveness of
On assessing responsiveness of health-related quality of life instru- condition-specific and generic health status measures in degenerative
ments: guidelines for instrument evaluation. Qual Life Res 2003;12: lumbar spinal stenosis. J Clin Epidemiol 1995;48:1369e78.
349e62. [52] Wright JG, Young NL. A comparison of different indices of respon-
[39] Guyatt G, Walter S, Norman G. Measuring change over time: assess- siveness. J Clin Epidemiol 1997;50:239e46.
ing the usefulness of evaluative instruments. J Chronic Dis 1987;40: [53] Stratford PW, Riddle DL. Assessing sensitivity to change: choosing
171e8. the appropriate change coefficient. Health Qual Life Outcomes
[40] Deyo RA, Centor RM. Assessing the responsiveness of functional 2005;3:23.
scales to clinical change: an analogy to diagnostic test performance. [54] Zou GY. Quantifying responsiveness of quality of life measures with-
J Chronic Dis 1986;39:897e906. out an external criterion. Qual Life Res 2005;14:1545e52.