G Model

JINJ 7649 No. of Pages 7

Injury, Int. J. Care Injured xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Injury
journal homepage: [Link]/locate/injury

Pre-operative factors associated with increased mortality in elderly

patients with a hip fracture: A cohort study in a developing country
Kristian A. Espinosaa,* , Amparo Gómez Gélveza , Liliana P. Torresa ,
María Fernanda Garcíaa,b , Omar R. Peñaa
a
Department of Orthopaedics & Traumatology, Hospital Universitario de La Samaritana, Bogotá, Colombia
b
Pontificia Universidad Javeriana, Bogotá, Colombia

A R T I C L E I N F O A B S T R A C T

Background: Hip fractures are a public health problem worldwide, and several factors are involved with
Keywords: post-operative mortality. The aim of this study was to identify the pre-operative factors associated with
Hip fracture
increased mortality in elderly patients with hip fractures in a developing country during the first post-
Mortality
operative year.
Elderly
Surgical delay Methods: An ambidirectional cohort study was conducted with patients  65 years of age who underwent
Multiple comorbidities hip surgery due to a hip fracture caused by a fall from a standing position. Socio-demographic data, time
Developing country to surgery, and comorbidities measured by the Charlson Comorbidity Index (CCI) were recorded. One-
year mortality from all causes was the primary outcome, and 30-day and 6-month mortality were the
secondary outcomes. Log-rank test was used to evaluate survival, and Cox’s proportional hazard
regression was used to detect the factors associated with increased mortality.
Results: 478 patients who underwent hip surgery were included in this study. The mean age was
80.2  9.9, and 297 (62%) were females. There were 150 (31.4%) deaths at the end of the first follow-up
year, and the mean of surgical delay was 8.8 days  6.4. Patients who underwent surgery during the first
4 days (Log-rank test < 0.001) after hip fracture occurred and patients with a CCI  2 (Log-rank
test < 0.001) showed better survival (90%), comparing to mortality (52%) of patients with a CCI  3 and
surgical delay > 4 days. The age  80 years (Hazard ratio 2.55 (HR), 95% confidence interval (CI) 1.70 to
3.84, p < 0.001), CCI  3 (HR 1.61, 95% CI 1.14–2.26, p 0.006), surgical delay > 4 days (HR 2.41, 95% CI 1.38–
4.21, p 0.006), and haemoglobin < 10 g/dl (HR 1.51, 95% CI 1.06–2.15, p 0.02) were associated with
increased 1-year mortality. In addition, 30-day mortality was associated with age  80 years (HR 4.15,
95% CI 1.98–8.70, p < 0.001), CCI  3 (HR 1.80, 95% CI 1.08–2.99, p 0.023), pre-surgical time >48 h (HR 3.0,
95% CI 1.58–5.92, p 0.001), and surgical delay > 4 days (HR 3.0, 95% CI 1.33–6.81, p 0.008); and 6-month
mortality was associated with surgical delay > 4 days (HR 2.72, 95% CI 1.42–5.23, p 0.003), and
haemoglobin < 10 g/dl (HR 1.56, 95% CI 1.04–2.33, p < 0.028).
Conclusions: Surgical delay greater than 4 days and Charlson Comorbidity Index  3 were found as factors
associated with increased mortality, along with anaemia < 10 g/dl and age  80 years. A similar mortality
rate was found in this study compared to the rates reported by the literature, despite a surgical delay of
8.8 days.
© 2018 Elsevier Ltd. All rights reserved.

Introduction it is expected that this problem will increase exponentially to 2.6

million by 2025 and to 4.5 million by 2050, secondary to the aging
Hip fractures (HFs) are a public health problem worldwide [1], of the population [3].
and it is the most common fracture among osteoporotic elderly After an elderly person has had a hip fracture, mortality rates
people [2]. Every year HFs affect more than 1.5 million people, and vary depending on the time of measurement. The proportion of
mortality at 1 month is 13.3% (1.2%–16.3%), at 6 months is 15.8%
(7.9%–26.7%) and at 1 year is 24.5% (7.8%–35%) [1]. There are
* Corresponding author at: Carrera 8 No. 0-55 SUR, Third Floor, Department of
systematic reviews (SRs) that were conducted to identify the pre-
Orthopaedics & Traumatology, Bogotá, Colombia. Tel.: (+57) 14077075, Ext: 10372. operative factors associated with increased mortality. Surgical
E-mail address: kristianesga@[Link] (K.A. Espinosa). delay has been associated with an increased risk of death [4–6].

[Link]
0020-1383/© 2018 Elsevier Ltd. All rights reserved.

Please cite this article in press as: K.A. Espinosa, et al., Pre-operative factors associated with increased mortality in elderly patients with a hip
fracture: A cohort study in a developing country, Injury (2018), [Link]
G Model
JINJ 7649 No. of Pages 7

2 K.A. Espinosa et al. / Injury, Int. J. Care Injured xxx (2018) xxx–xxx

Other associated factors which have also shown strong evidence of assessed by two physicians. Socio-demographic data were
mortality are multiple comorbidities, advanced aged, male gender recorded as well as living arrangements, which were divided into
and high American Anaesthetists Society (ASA) Score grading [1]. categories, such as: alone, in a nursing home or with someone.
Clinical guidelines recommend that elderly patients should Length of stay (LOS) was also measured, and it was calculated from
undergo surgery within the first 48 h after hospital admission [7,8]. the day of hospital admission to the day of discharge.
However, the effect of early surgery on mortality is still Four days were selected as the cut-off point for surgical delay
controversial, due to the low quality of evidence [7]. Particularly based on the characteristics of the hospital. As a referral hospital
in developing countries the uncertainty about the effect of early for isolated places, it was estimated a priori that patients can take
surgery may be even greater, as some studies have shown similar 2 days on average to be admitted to the institution, due to their
post-operative mortality rates, despite having a long delay before transportation by car ambulances. Another 2 days were added,
surgery. The time between the fall and subsequent fracture, and taking into account the recommendation that patients should
the hospital admission can be greater than 3 days. Patients can undergo hip surgery after admission by clinical guidelines [7,8].
undergo surgery 11 days after hospital admission, showing a Surgical delay time was also divided into pre-admission time, and
mortality rate of 13% at one year post-operative [9]. In contrast, in pre-surgical time. Pre-admission time was defined as the time
developed countries patients may be admitted to hospital during between the occurrence of the fracture and hospital admission,
the first 6 h after the trauma [10,11], and receive surgical treatment and pre-surgical time as the duration between hospital admission
within the first 24 h after the fracture [11], finding similar and surgery.
mortality rates at one year post-operative. The measured comorbidities were those included in the Charlson
Additionally, SRs on the effect of timing of hip surgery on Comorbidity Index (CCI), and a CCI  3 was selected as the group of
mortality have not included studies from developing countries, patients with a higher probability of mortality according to previous
despite having carried out exhaustive searches for this literature studies [12]. The CCI took into account 19 medical conditions, giving a
[4–6]. Conversely, SRs related to identifying factors associated with value depending on their severity. One point was considered for
increased mortality other than pre-operative time included studies pathologies such as: myocardial infarction, congestive heart failure,
from developing countries. However, they only accounted for less peripheral vascular disease, cerebrovascular disease, dementia,
than 3% of the studies [1]. For this reason evidence regarding chronic pulmonary disease, connective tissue disease, peptic ulcer
factors associated with increased mortality in developing coun- disease, mild liver disease and diabetes. Two points were assigned to
tries is limited. Therefore, more evidence should be generated in hemiplegia, moderate or severe renal disease, diabetes with end
order to address treatment strategies. organ damage, any tumour, leukaemia, and lymphoma. Three and six
The objective of this study was to identify the pre-operative points were given to moderate/severe liver disease, and metastatic
factors associated with increased mortality of elderly patients with solid tumour or acquired immune deficiency syndrome (AIDS)
hip fractures in a developing country during the first post- respectively [13]. Appendix 1 shows the definition of each medical
operative year. Two hypotheses were addressed: 1. Surgical delay condition in the CCI.
would not have any effect on increasing mortality during the first Patients were also categorised based on the classification of the
post-operative year in a developing country. 2. Multiple comor- American Society of Anaesthesiologists (ASA), which stated five
bidities measured through Charlson Comorbidities Index would be categories. ASA I indicated healthy patients, ASA II and ASA III
associated with a higher risk of mortality. patients with mild and severe systemic diseases respectively. ASA
IV was assigned to patients with severe systemic diseases that
Materials and methods were a constant threat of life, and ASA V to moribund patients that
need an operation to survive [14]. In this study, patients were split
An ambidirectional cohort study was conducted with into two categories for the analysis, patients with ASA I or ASA II,
patients  65 years of age who underwent hip surgery due to a and patients with ASA  III.
hip fracture caused by a fall from a standing position. They were Hip fractures were classified into intracapsular and extracap-
treated during 2010 and 2015 at a referral hospital in Bogotá, sular fractures. The distinction between the two types of fractures
Colombia. Patients with hip fractures due to oncologic diseases, was made through x-rays by orthopaedic surgeons. Moreover,
periprosthetic fractures, multiple fractures and previous history of haemoglobin values of the patients were obtained at the moment
surgery due to an ipsilateral or contralateral hip fracture were of hospital admission, and a haemoglobin of 10 g/dl was
excluded. Patients who were not properly registered in the hospital considered as cut-off for the analysis [15].
records were also excluded from the study.
Patients included in this study were land transferred to the Statistical analyses
institution from cities or small towns of the state, and also from
isolated areas of the country. As soon as patients arrived to the Initially, a descriptive analysis of the variables was carried out in
hospital, they were treated by the departments of orthopaedics which categorical variables were presented in proportions and
and traumatology, internal medicine, and anaesthesiology. When percentages. For continuous variables measurements such as
it was necessary, some cases were evaluated by other medical or mean, standard deviation (SD) and range were used. Kaplan-Meier
surgical specialties. After patients were discharged, they had was used to construct cumulative survival curves to the exposed
follow-up care in our institution or in hospitals of their home- and unexposed groups for each of the two variables of interest,
towns. All of them had telephone follow-ups by the department of surgical delay >4 days and Charlson Comorbidity Index  3. The
orthopaedics and traumatology, through which mortality was Log-rank test determined if any difference found between the
obtained. One-year mortality from all causes was the primary exposed and non-exposed groups were statistically significant.
outcome, and 30-day and 6-month mortality were the secondary Cox’s proportional hazard regression was used to evaluate the
outcomes. effect not only of the two variables of interest, but also with other
The retrospective data were collected from clinical records by covariates on the survival; this statistical analysis was carried out
two physicians, and a third physician verified the retrieved for each of the outcomes. Variables were screened one-at-a-time
information. Discrepancies of the data were resolved between (univariate analysis), and then variables with p-values of 0.15 or
them, otherwise a fourth physician was consulted to reach less resulting from this analysis were included in the multivariate
consensus. The prospective data (mortality) was independently analysis. Backward elimination method was used to obtain the

Please cite this article in press as: K.A. Espinosa, et al., Pre-operative factors associated with increased mortality in elderly patients with a hip
fracture: A cohort study in a developing country, Injury (2018), [Link]
G Model
JINJ 7649 No. of Pages 7

K.A. Espinosa et al. / Injury, Int. J. Care Injured xxx (2018) xxx–xxx 3

final model. P-values < 0.05 were considered statically significant, which corresponded to 36 (7.5%) in-hospital deaths after surgery
and data were reported with Hazard ratio (HR) and 95% confidence and 33 (6.9%) deaths between hospital discharge and 30-day post-
interval (CI). Data were analysed using IBM SPSS Statistics 21. operative. 6-month mortality represented 24% (n = 115) of deaths.
The sample size calculation was performed based on the results Patients who underwent surgery during the first 4 days after
of a pilot study for surgical delay >4 days, and Charlson the hip fracture occurred, and patients with a CCI  2, showed
Comorbidity Index  3. For both variables we used the same better survival. In fact, patients who had a surgical delay > 4 days
statistical parameters such as a statistical power of 80%, 1:3 ratio, presented 37% (n = 120) of mortality vs 20% (n = 30) in the group
accrual period of 365 days and a significance level of 0.05, however, who had earlier surgery at 1-year follow-up (Fig. 2A), being the
we used in case of surgical delay >4 days and CCI  3 a mean difference between those groups statistically significant (Log-rank
survival of 250 days and 280 days respectively, and an increased test < 0.001). Similarly, 46% (n = 51) of patients with a CCI  3 died,
mortality rate of 30% for both variables. The required sample size to compared to 27% (n = 99) of the CCI  2 group; that difference
reject the null hypothesis was at least of 100 patients on the between those groups was also statistically significant (Log-rank
exposed group and 300 patients on the unexposed group for test < 0.001) (Fig. 2B) (Table 2).
surgical delay >4 days, and 105 patients on the exposed group and After selecting variables with p values  0.15 values from the
315 patients on the unexposed group for Charlson Comorbidity univariate analysis of each outcome (Table 3), a Cox’s proportional
Index  3. Taking into account the two sample size calculations, a hazard regression was carried out with them. As a result, a final
higher sample of 500 patients was randomly selected to ensure the model was obtained for 1-year mortality with the following
number of required patients due to the reduction of patients after statistically significant variables: age  80 yr, Charlson comorbidi-
applying the exclusion criteria. The sample size calculation was ty index  3, surgical delay > 4 days, and haemoglobin < 10 g/dl. In
performed using the PS Power and Sample Size Calculation addition, the final model for secondary outcomes such as 30-day
software version 3.1.2, 2014 [16]. mortality was made up by age  80 years, CCI  3, pre-surgical
time > 48 h, and surgical delay > 4 days; and 6-month mortality
Results outcome had statistically significant variables as surgical de-
lay > 4 days, and haemoglobin < 10 g/dl (Table 4).
500 patients  65 years of age who underwent hip surgery were
randomly selected from hospital records, and 478 patients were Comorbidities
included in this study after applying the exclusion criteria (Fig. 1).
In general, patients had a surgical delay of 8.8 days  6.4 on Within the pathologies evaluated using the Charlson comor-
average: the pre-admission time had a mean of 3.81 days  2.9, and bidity index, chronic pulmonary disease (chronic obstructive
the pre-surgical time presented a mean of 5.07 days  3.4. The pulmonary disease (COPD) and asthma) was the most common
mean of length of stay was 12.65 days  9.1. Table 1 shows the condition in 51% (n = 244) of the cases, followed by diabetes
characteristics of the general population, as well as patients with a mellitus with 16.9% (n = 81), dementia 14% (n = 67) and congestive
surgical delay >4 days and Charlson comorbidity index  3. heart failure 13.6% (n = 65). Other diseases such as myocardial
infarction and hemiplegia due to stroke were also presented in
Survival 8.4% (n = 40) and 4.2% (n = 20) of the patients respectively.
Leukaemia, lymphoma, moderate or severe liver disease, meta-
There were 150 (31%) deaths in the entire population at the end static solid tumour and AIDS were not found in this study.
of the first follow-up year. 30-day mortality was 14.4% (n = 69), Comorbidities such as myocardial infarction, chronic pulmonary

Fig. 1. Flow diagram of the patient selection process.

Please cite this article in press as: K.A. Espinosa, et al., Pre-operative factors associated with increased mortality in elderly patients with a hip
fracture: A cohort study in a developing country, Injury (2018), [Link]
G Model
JINJ 7649 No. of Pages 7

4 K.A. Espinosa et al. / Injury, Int. J. Care Injured xxx (2018) xxx–xxx

Table 1
Characteristics of the elderly patients with hip fractures.

Variables General population n = 478 (%) Surgical delay > 4 days n = 379 (%) CCI  3 n = 110 (%)
Age, mean [SD] 80.2 [9.9] 80.8 [8.8] 82.2 [6.1]
Sex
Male 181 (38) 138 (36) 41 (37)
Female 297 (62) 241 (64) 69 (63)
BMI, Kg/m2
 18.5 42 (9) 34 (9) 12 (11)
18.6–25 262 (55) 209 (55) 60 (55)
25.1 174 (36) 136 (36) 38 (34)
Living arrangements
Alone 67 (14) 54 (14) 4 (4)
Nursing home 57 (12) 41 (11) 17 (15)
With someone 354 (74) 284 (75) 89 (81)
Type of fracture
Extracapsular 378 (79) 293 (77) 93 (85)
Intracapsular 100 (21) 86 (23) 17 (15)
ASA
I–II 172 (36) 122 (32) 11 (10)
3 306 (64) 257 (68) 99 (90)
Haemoglobin, g/dl
<10 101 (21) 88 (23) 33 (30)
 10.1 377 (79) 291 (77) 77 (70)

CCI = Charlson Comorbidity Index; BMI = body mass index; ASA = classification of the American Society of Anaesthesiologists; SD = standard deviation.

Fig. 2. Kaplan-Meier analysis of the probability of survival among (A) patients with a surgical delay > 4 days, and patients with a surgical delay  4; as well as (B) patients in
the CCI  3 group, and patients in the CCI  2 group. CCI = Charlson comorbidity index.

Table 2
Time to surgery and Charlson Comorbidity Index in the general population and in patients with mortality.

Variables Total patients n = 478 30-day mortality n = 69 (14%)a 6-month mortality n = 115 (24%)a 1-year mortality n = 150 (31%)a
Time of Surgery
Surgical delay >4 days (%) 379 60 (16) 105 (28) 136 (36)
Pre-surgical time >48 h (%) 398 53 (13) 99 (25) 129 (32)
Pre-admission time >24 h (%) 272 47 (17) 77 (28) 98 (36)
Charlson Comorbidity Index
Mean CCI, [SD] 1.47 [1.30] 1.88 [1.32] 1.83 [1.29] 1.90 [1.36]
CCI  3 (%) 110 24 (22) 40 (36) 51 (46)
a
Cumulative mortality. CCI = Charlson Comorbidity Index; SD = standard deviation.

disease and hemiplegia were associated with increased mortality was statistically significant (Odds ratio (OR): 2.32, CI 95% 1.36–3.97,
during the first post-operative year (Table 5), after adjusting for p = 0.002). However, it was not observed at 30-day post-operative
age  80 years, male sex and surgical delay >4 days variables. mortality (OR: 1.7, CI 95% 0.95–2, p = 0.069) (Table 6). Conversely,
The group of patients with a CCI  3 and surgical delay >4 days patients with fewer comorbidities (CCI  2) and surgical delay
presented an increased mortality at 1-year post-operative, which  4 showed a reduced mortality that was statistically significant at

Please cite this article in press as: K.A. Espinosa, et al., Pre-operative factors associated with increased mortality in elderly patients with a hip
fracture: A cohort study in a developing country, Injury (2018), [Link]
G Model
JINJ 7649 No. of Pages 7

K.A. Espinosa et al. / Injury, Int. J. Care Injured xxx (2018) xxx–xxx 5

Table 3
Univariate analysis of factors associated with increased mortality.

30-day mortality 6-month mortality 1-year mortality

Variables Hazard Ratio (95% CI) p value Hazard Ratio (95% CI) p value Hazard Ratio (95% CI) p value
Age  80 yr 1.46 (0.96 –2.23) 0.07 1.04 (0.69 –1.58) 0.82 2.77 (1.85 –4.16) <0.001
Male 1.25 (0.91 –1.73) 0.16 1.04 (0.75 –1.44) 0.77 1.31 (0.95 –1.81) 0.09
BMI  18.5 kg/m2 0.92 (0.54 –1.55) 0.76 0.87 (0.51 –1.46) 0.59 0.26 (0.79 –2.25) 0.26
Alone or nursing home 0.68 (0.46 –1.1) 0.06 1.2 (0.81 –1.76) 0.35 0.84 (0.57 –1.24) 0.40
Intracapsular fracture 1.12 (0.77 –1.64) 0.54 1.25 (0.85 –1.82) 0.24 1.17 (0.80 –1.71) 0.39
ASA  3 1.28 (0.88 –1.86) 0.19 1.13 (0.78 –1.64) 0.50 1.86 (1.28 –2.70) <0.001
CCI  3 1.30 (0.92 –1.83) 0.12 1.09 (0.77 –1.53) 0.60 1.90 (1.35 –2.67) <0.001
Pre-admission time >24 h 1.15 (0.82 –1.61) 0.40 1.03 (0.73 –1.44) 0.85 1.51 (1.08 – 2.11) 0.01
Pre-surgical time >48 h 1.51 (0.95 –2.39) 0.08 1.04 (0.66 –1.66) 0.84 1.25 (0.78 –1.98) 0.34
Surgical delay >4 days 1.74 (0.86 –3.50) 0.12 2.89 (1.51 –5.54) <0.001 2.8 (1.62 –4.87) <0.001
Haemoglobin <10 g/dl 1.86 (1.13 –3.07) 0.015 1.70 (1.14 –2.53) <0.01 1.8 (1.30 –2.60) <0.001

BMI = body mass index; ASA = classification of the American Society of Anaesthesiologists; CCI = Charlson Comorbidity Index; CI = confidence interval.

1-year post-operative (OR: 0.21, CI 95% 0.07–0.59, p = 0.003), and at

Table 4 30-day post-operative (OR: 0.26, CI 95% 0.09–0.73, p = 0.01)
Cox proportional-hazards regression to identify factors associated with increased
post-operative mortality.
(Table 6).

Hazard Ratio (95% CI) p value Discussion

30-day mortality
Age  80 yr 4.15 (1.98–8.70) <0.001 We hypothesised that surgical delay would not be associated
CCI  3 1.80 (1.08–2.99) 0.023
with post-operative mortality in a developing country. However,
Pre-surgical time > 48 h 3.0 (1.58–5.92) 0.001
Surgical delay >4 days 3.0 (1.33–6.81) 0.008 the hypothesis was rejected based on the results of this study. In
6-month mortality fact, surgical delay >4 days was associated with increased
Surgical delay >4 days 2.72 (1.42–5.23) 0.003 mortality during all post-operative periods of follow-up (at 30-
Haemoglobin < 10 g/dl 1.56 (1.04–2.33) 0.028
day, 6-month and 1-year). Similarly, the pre-surgical time >48 h
1-year mortality
Age  80 yr 2.55 (1.70–3.84) <0.001
was also associated with increased mortality, but only 30 days after
CCI  3 1.61 (1.14–2.26) 0.006 surgery. These findings are in agreement with the results of the
Surgical delay > 4 days 2.41 (1.38–4.21) 0.002 systematic reviews available in the literature, which show that
Haemoglobin < 10 g/dl 1.51 (1.06–2.15) 0.02 surgical delay is associated with a greater risk of death [4–6].
CCI = Charlson Comorbidity Index; CI = confidence interval

Table 5
Association of comorbidities included in the Charlson comorbidity index with post-operative mortality.

Comorbidities n (%) Unadjusted HR (95% CI) p value Adjusted HRa (95% CI) p value
Myocardial infarction 40 (8.4) 2.57 (1.66 –3.98) <0.001 2.50 (1.43–4.35) <0.001
Congestive heart failure 65 (13.6) 1.66 (1.11 –2.50) <0.01 1.41 (0.93–2.11) 0.09
Peripheral vascular disease 26 (5.4) 1.20 (0.63 –2.29) 0.56 0.95 (0.50– 1.82) 0.88
Cerebrovascular disease 11 (2.3) 0.77 (0.24 –2.44) 0.66 1.01 (0.31–3.24) 0.98
Dementia 67 (14) 1.72 (1.16 –2.56) <0.01 1.48 (0.99– 2.20) 0.05
Chronic pulmonary disease 244 (51) 1.58 (1.14 –2.20) <0.01 1.53 (1.10 –2.12) 0.01
Connective tissue disease 4 (0.8) 0.04 (0.00 –66.9) 0.41 0.00 (0.0– 1.2) 0.96
Peptic ulcer disease 48 (10) 1.01 (0.62 –1.76) 0.84 0.95 (0.57– 1.60) 0.87
Mild liver disease 4 (0.8) 0.04 (0.00 –66.9) 0.41 0.00 (0.0–1.2) 0.96
Diabetes mellitus 81 (16.9) 0.82 (0.52 –1.29) 0.39 0.87 (0.55– 1.38) 0.57
Hemiplegia 20 (4.2) 2.58 (1.42 –4.66) <0.01 2.14 (1.15 –3.96) <0.01
MS renal disease 28 (5.9) 1.18 (0.62 –2.25) 0.62 0.99 (0.52– 1.89) 0.98
DM with end organ damage 6 (1.3) 0.96 (0.23 –2.89) 0.95 1.13 (0.27– 4.61) 0.86
Any tumour 29 (6.1) 2.22 (1.34 –2.86) <0.01 1.50 (0.89– 2.51) 0.12
a
Adjusted for age  80 years, male sex and surgical delay > 4 days. CCI = Charlson comorbidity index; MS = moderate or severe; DM = diabetes mellitus; HR = hazard ratio;
CI = confidence interval.

Table 6
Subgroup analysis according to Charlson Comorbidity Index and surgical delay on mortality.

Subgroup Total patients 30-day mortality Mortality between 1 month and 1 year

n (%) Odds Ratio (95% CI) p value n (%) Odds Ratio (95% CI) p value
CCI  3 & surgical delay > 4 days 93 19 (20) 1.17 (0.95–2.0) 0.069 26 (28) 2.32 (1.36 –3.97) 0.002
CCI  3 & surgical delay  4 days 17 5 (29) 2.58 (0.88 –7.5) 0.084 1 (6) 0.29 (0.03 –2.27) 0.243
CCI  2 & surgical delay > 4 days 286 41 (14) 0.98 (0.58 –1.64) 0.94 50 (17) 1.1 (0.67 –1.79) 0.7
CCI  2 & surgical delay  4 days 82 4 (5) 0.26 (0.09 –0.73) 0.01 4 (5) 0.21 (0.07 –0.59) 0.003

CCI = Charlson Comorbidity Index; CI = confidence interval.

Please cite this article in press as: K.A. Espinosa, et al., Pre-operative factors associated with increased mortality in elderly patients with a hip
fracture: A cohort study in a developing country, Injury (2018), [Link]
G Model
JINJ 7649 No. of Pages 7

6 K.A. Espinosa et al. / Injury, Int. J. Care Injured xxx (2018) xxx–xxx

Delayed surgery has been a controversial topic during the last compared to the mortality (10%) in the group of the healthier
five decades [17]. There are still clinicians that prefer to delay patients (CCI  2) who received quicker surgical treatment
surgical treatment in order to conduct an investigation and (surgical delay  4 days). It is noteworthy that 30-day mortality
stabilisation of pre-operative medical conditions [18], despite the in patients with a CCI  3 and surgical delay >4 days was 20%,
evidence from systemic reviews and clinical practice guidelines which is 9% lower than the mortality (29%) in patients with the
(CPG) that show an association of surgical delay with increased same CCI  3 but with a surgical delay  4 days. These findings
risk of mortality and complications [4–8]. Uncertainty in suggest that surgical delay for patients who were “more sick”
clinicians opinions persists due to the evidence that was based (CCI  3) was not a determining factor on mortality. Rather, the
on mainly retrospective observational studies which could have determining factor was the number of comorbidities they had.
inadequate statistical analyses or low sample sizes. In addition, Therefore, if the medical approach is focused on the optimiza-
the evidence regarding this topic was ranked as low quality by tion of pathologies rather than timing, this could greatly reduce
CPG. The last systematic reviews on this topic were published mortality in this specific subgroup of patients.
between 2010 and 2012 [4–6]. Since then many studies have been In this study delayed surgery (>4 days), multiple comorbid-
conducted with results in favour of and against the benefit of ities (CCI  3), haemoglobin <10 gr/dl and age >80 years are
early surgery on mortality. Therefore, based on these new associated with increased one-year mortality post-operative.
findings it is essential to carry out a new SR in order to help There are several other factors that may play a role in mortality;
the decision makers. perhaps other variables that were not measured, such as pre-
It is noteworthy to mention that the mortality rates found in operative functionality or cognitive impairment. Both variables
this study (14% at 30 days, 24% at 6 months and 31% at 1 year have already been found to be factors associated with increased
postoperative) are similar to the reported mortality rates by other mortality [1]. Given the multiple factors involved and the
studies [1], despite our mean surgical delay of 8.8 days. This is complexity of the patients, an orthogeriatric model for hip
similar to the values found in another study that was conducted in fracture patients should be implemented. This integrated
a developing country, which reported a lower one-year mortality approach has shown a less prolonged pre-surgical period,
rate (13%), but with a longer mean of surgical delay (16 days) [9]. In length of stay and lower mortality and complication rates. It has
our study delayed surgery was associated with an increased risk of also shown an improvement of post-operative functionality
death after performing a multivariate analysis. However the [22,23].
uncertainty remains, as the lack of randomisation in this study This study has several strengths that should be noted.
could have caused a selection bias in which healthier patients were According to our research, this is the first study in Colombia
more easily referred for surgery [4]. Currently there are only two with these characteristics, and one of the few studies to be
randomised clinical trials (RCT). One trial was inaccurate in many conducted in developing countries. In addition, this study did not
aspects [4,19] and the other was a pilot study [10]. For this reason a present missing data, which avoids attrition bias. However, this
high quality RCT should be performed to clarify this issue. study also has several limitations. Firstly, this is an observational
The reasons for delayed surgery have been thoroughly study that by nature is intrinsically prone to several biases such
described in the literature. There are many causes of pre- as: selection, performance and information bias. Secondly, the
hospital delay, such as a patient being transferred from another used epidemiologic design was an ambidirectional cohort study,
hospital, or a patient chooses to delay their visit to hospital [20]. which had a retrospective and a prospective component. In order
In this study 80% of the patients were transferred from hospitals to overcome the retrospective pitfall, a second researcher verified
in isolated locations by car ambulances. This is a determining all the collected data to increase the quality of the information.
factor in the delayed arrival of our patients to the hospital. Therefore we tried to limit this effects of this problem as much as
Although in Colombia there is air transportation to some possible.
hospitals, we currently do not have this facility. Therefore if In conclusion a surgical delay greater than 4 days and a Charlson
implemented, this could provide a possible solution to reduce Comorbidity Index  3 were found as factors associated with
the pre-hospital time of our patients. Moreover, there are increased post-operative mortality, as well as other factors such as
several causes of delay in surgical treatment after hospital anaemia or patients older than 80 years. The proportion of
admission, such as admission on weekend or holidays, drugs’ mortality during the post-operative period in elderly patients was
hold (aspirin, antiplatelet, warfarin), medical consultations, similar to the rates reported by the medical literature, despite a
delayed diagnosis and delayed admission from the emergency mean surgical delay of 8.8 days. Clinicians and healthcare
room due to patient’s bed shortage [20]. We believe these causes institutions should strive to modify risk factors such as surgical
also apply to our clinical setting. It is worth noting that the delay, assessing and stabilizing comorbidities of patients promptly
exacerbations and decompensations of pathologies can be due in order to perform surgery on the day of, or the day after hospital
to the fracture or to the prolonged pre-hospital delay. admission.
We proposed a hypothesis that patients with multiple
comorbidities (Charlson Comorbidity Index  3) would be Funding source
associated with a higher mortality. This hypothesis was
confirmed by the results of this study. CCI  3 was associated There was no external funding source to support this study.
with increased mortality at 30-day, 6-month and 1-year post-
operative. We found that the results of SRs matched our patients Ethical standards
results [1]. In the post-hoc analysis medical conditions such as
myocardial infarction, chronic pulmonary disease (COPD and The study was approved by both an internal medical board and
asthma) and hemiplegia due to stroke were associated with Ethics Committee of the Hospital Research Center. The study was
increased 1-year mortality. Indeed these diseases are the performed in accordance with the Ethical standards of the 1964
leading causes of death worldwide [21]. In addition, when Declaration of Helsinki and its later amendments.
patients were split according to their Charlson comorbidity
Index and the time of surgical delay (Table 6), we found a higher Conflict of interest
mortality (48%) during the first post-operative year in the group
of “more sick” patients (CCI  3) with a surgical delay >4 days, The authors declare that they have no conflict of interest.

Please cite this article in press as: K.A. Espinosa, et al., Pre-operative factors associated with increased mortality in elderly patients with a hip
fracture: A cohort study in a developing country, Injury (2018), [Link]
G Model
JINJ 7649 No. of Pages 7

K.A. Espinosa et al. / Injury, Int. J. Care Injured xxx (2018) xxx–xxx 7

Acknowledgments [9] Vidal E, Moreira-Filho D, Pinheiro R, Souza RC, Almeida L, Camargo Jr. K, et al.
Delay from fracture to hospital admission: a new risk factor for hip fracture
mortality? Osteoporos Int 2012;23(12):2847–53.
The authors would like to express their gratitude to Elaine Pelan [10] Buse G, Bhandari M, Sancheti P, Rocha S, Winemaker M, Adili A, et al.
for carrying out a spellcheck. Likewise, the authors acknowledge Accelerated care versus standard care among patients with hip fracture: the
Gloria Stella Bernal, María del Rosario Cardenas and the rest of the HIP ATTACK pilot trial. CMAJ 2014;186(1):E52–e60.
[11] Muhm M, Arend G, Ruffing T, Winkler H. Mortality and quality of life after
members of the Statistics department of the Hospital Universitario proximal femur fracture-effect of time until surgery and reasons for delay. Eur
de La Samaritana, for their assistance in getting the list of patients J Trauma Emerg Surg 2013;39(3):267–75.
treated for a hip fracture at the hospital. [12] Folbert EC, Hegeman JH, Vermeer M, Regtuijt EM, van der Velde D, Ten Duis
HJ, et al. Improved 1-year mortality in elderly patients with a hip fracture
following integrated orthogeriatric treatment. Osteoporos Int 2017;28
Appendix A. Supplementary data (1):269–77.
[13] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying
prognostic comorbidity in longitudinal studies: development and validation. J
Supplementary data associated with this article can be found, in
Chronic Dis 1987;40(5):373–83.
the online version, at [Link] [14] Owens WD, Felts JA, Spitznagel Jr. EL. ASA physical status classifications: a
study of consistency of ratings. Anesthesiology 1978;49(4):239–43.
References [15] Maxwell MJ, Moran CG, Moppett IK. Development and validation of a
preoperative scoring system to predict 30 day mortality in patients
undergoing hip fracture surgery. Br J Anaesth 2008;101(4):511–7.
[1] Hu F., Jiang C, Shen J, Tang P, Wang Y. Preoperative predictors for mortality [16] Dupont WD, Plummer Jr. WD. Power and sample size calculations for studies
following hip fracture surgery: a systematic review and meta-analysis. Injury involving linear regression. Controlled Clin Trials 1998;19(6):589–601.
2012;43(6):676–85. [17] Davie IT, MacRae WR, Malcolm-Smith NA. Anesthesia for the fractured hip: a
[2] Panula J, Pihlajamaki H, Mattila VM, Jaatinen P, Vahlberg T, Aarnio P, et al. survey of 200 cases. Anesth Analg 1970;49(1):165–70.
Mortality and cause of death in hip fracture patients aged 65 or older: a [18] Parker M, Johansen A. Hip fracture. BMJ Clin Res Ed 2006;333(7557):27–30.
population-based study. BMC Musculoskelet Disord 2011;12:105. [19] Swanson C, Day G, Yelland C, Broome J, Massey L, Richardson H, et al. The
[3] Abrahamsen B, van Staa T, Ariely R, Olson M, Cooper C. Excess mortality management of elderly patients with femoral fractures. A randomised
following hip fracture: a systematic epidemiological review. Osteoporos Int controlled trial of early intervention versus standard care. Med J Aust Internet
2009;20(10):1633–50. 1998;169(10):515–8.
[4] Moja L, Piatti A, Pecoraro V, Ricci C, Virgili G, Salanti G, et al. Timing matters in [20] Cha YH, Ha YC, Yoo JI, Min YS, Lee YK, Koo KH. Effect of causes of surgical delay
hip fracture surgery: patients operated within 48 hours have better outcomes. on early and late mortality in patients with proximal hip fracture. Arch Orthop
A meta-analysis and meta-regression of over 190, 000 patients. PLoS One Trauma Surg 2017;137(5):625–30.
2012;7(10):e46175. [21] Organization WH. The top 10 causes of death. World Health Organization;
[5] Shiga T, Wajima Z, Ohe Y. Is operative delay associated with increased 2017 Updated 2017. Available from: [Link]
mortality of hip fracture patients? Systematic review, meta-analysis, and factsheets/fs310/en.
meta-regression. Can J Anaesth 2008;55(3):146–54. [22] Doshi HK, Ramason R, Azellarasi J, Naidu G, Chan WL. Orthogeriatric model for
[6] Simunovic N, Devereaux PJ, Sprague S, Guyatt GH, Schemitsch E, Debeer J, et al. hip fracture patients in Singapore: our early experience and initial outcomes.
Effect of early surgery after hip fracture on mortality and complications: Arch Orthop Trauma Surg 2014;134(3):351–7.
systematic review and meta-analysis. CMAJ 2010;182(15):1609–16. [23] Kammerlander C, Roth T, Friedman SM, Suhm N, Luger TJ, Kammerlander-
[7] Centre NCG. The management of hip fracture in adults. NICE; 2011 Updated Knauer U, et al. Ortho-geriatric service–a literature review comparing different
2014. Available from: [Link] models. Osteopor Int 2010;21(Suppl. 4):S637–46.
guideline-183081997.
[8] Surgeons AAoO. Management of hip fractures in the elderly. American
Academy of Orthopaedic Surgeons; 2014 Available from: [Link]
org/research/guidelines/HipFxGuideline_rev.pdf.

Please cite this article in press as: K.A. Espinosa, et al., Pre-operative factors associated with increased mortality in elderly patients with a hip
fracture: A cohort study in a developing country, Injury (2018), [Link]