Observational Study Medicine ®

OPEN

Mobile phone application for self-assessment

of acute stroke patients
A tool for extended care and follow-up
∗
Hong Chang, MN , Jie Zhao, MN, Yuchen Qiao, MN, Hui Yao, BN, Xiaojuan Wang, BN,
Juanmin Li, BN, Jia Liu, MD, PhD

Abstract
We aimed to assess the accuracy of self-assessment for acute stroke patients via mobile phone application-based scales and
determine the value and prospect of clinical use.
A cross-sectional study was designed and acute stroke patients were enrolled. We pushed the modified Rankin scale (mRS) and
activities of daily living (ADL) scale to patients via mobile phone application for self-assessment on the day before they were out of
hospital. We compared the results from nurse assessment and self-assessment.
Around 50 patients with the average age 51.72 ± 12.40 completed the self-assessment. A total of 27 patients self-assessed the
scales, while caregivers of other 23 patients completed the assessment. In comparison with patient assessment and nurse assessment,
significant difference was found in ADL score (P = .004), but was not found in mRS score (P > .05). When comparing caregiver
assessment with nurse assessment, no significant difference could be found either in ADL score (P > .05) or in mRS score (P > .05). The
kappa value for self-assessment and nurse agreement of ADL was 0.720 (P = .000), with sensitivity 96.8% and specificity 82.0%. The
kappa value for self-assessment and nurse agreement of mRS was 0.718 (P = .000), with sensitivity 97.6% and specificity 92.4%.
In summary, mobile phone application-based scales are generally accurate, economical and convenient for self-assessment of acute
stroke patients with acceptable reliability in our small scale study. Caregivers can serve as the proper assessor when patients are out of
hospital. Therefore, it is promising but still need to be further confirmed how practical to use this application in extended care and follow-up.
Abbreviations: x ± s = mean ± standard deviation, ADL = activities of daily living, M = median, mRS = modified Rankin scale, OS
= operation system, P25 = 25th percentile, P75 = 25th percentile.
Keywords: acute stroke, extended care, mobile phone application, self-assessment

1. Introduction spend the rehabilitation phase at home or in health service

institution of community. In order to improve the extended care
With population aging, the incidence of stroke is rapidly
and provide the information for follow-up, it is of great
increased with high morbidity, mortality and disability.[1,2]
importance to regularly obtain the accurate results of assessment.
Epidemiological investigation suggests there are 7 to 8 million
In general, there are mainly 2 ways of assessing, that is, face to
stroke patients in Chinese mainland with the morbidity of 220 to
face in hospital and follow-up via mobile equipment.
250 per 100 thousand.[3] In which, 70% to 80% of patients suffer
With the development of modern information technology,
the sequelae in different extent.[4] Thus, activities of daily living
mobile health application provides the possibility for improving
are severely affected together with lower life quality and heavier
the pattern of extended care and follow-up, and makes it more
burden of family and society.[5–7]
timesaving, convenient and economic. For instance, mobile
Due to Chinese limited health resource, most stroke patients
phone application can be used in self-assessment of patients with
stay in hospital for therapy just during the acute phase. Then they
better operability, practicability and compliance. On the other
hand, the information is easily acquired and analyzed by the
Editor: Bernhard Schaller. hospital with professional guidance.
HC and JZ both contributed equally to this study. In this study, we aimed to assess the accuracy of self-assessment
The authors have no conflicts of interest to disclose. for acute stroke patients via mobile phone application-based
Department of Neurology, Xuanwu Hospital, Capital Medical University, Beijing, scales and determine the value and prospect of clinical use.
China.
∗
Correspondence: Hong Chang, Department of Neurology, Xuanwu hospital,
Capital Medical University, Changchun Street 45, Beijing 100053, China 2. Patients and methods
(e-mail: Changhong19791111@126).
A cross-sectional study was designed. The study flow was
Copyright © 2018 the Author(s). Published by Wolters Kluwer Health, Inc.
illustrated as Figure 1.
This is an open access article distributed under the terms of the Creative
Commons Attribution-Non Commercial-No Derivatives License 4.0 (CCBY-NC-
ND), where it is permissible to download and share the work provided it is
2.1. Patients
properly cited. The work cannot be changed in any way or used commercially
without permission from the journal. We enrolled the stroke patients during September and October
Medicine (2018) 97:26(e11263) 2016 from stroke unit, Xuanwu Hospital, Capital Medical
Received: 14 February 2018 / Accepted: 5 June 2018 University, Beijing. Inclusion criteria were: aged 16 or more;
[Link] clearly diagnosed as acute stroke (onset within 12 hours before

1
Chang et al. Medicine (2018) 97:26 Medicine

application on smartphone of patient or caregiver with

agreement.

2.3. Assessment
One day before moving out of the hospital, the scales were
pushed via software platform. Professional nurse would guide
and training the patient and caregiver how to use the application.
The face-to-face assessment of professional nurse was regarded as
the golden standard. These professional nurses were senior nurse
serving in neurology, who were independent and blinding in
developing study protocol and patient’s demographics. The self-
assessment of patient or caregiver would be compared with
golden standard. Meanwhile, the nurse and patient were paired in
the future extended care and follow-up.

2.4. Scales
The assessment tools were the modified Rankin scale (mRS) and
activities of daily living (ADL) scale (Fig. 2). The mRS is widely used
scale for measuring the degree of disability or dependence in the daily
activities of people who suffered stroke or other causes of
neurological disability,[8,9] in which the score ranges from 0 to 6,
suggesting from perfect health without symptoms to death: 0—no
symptoms; 1—no significant disability despite symptoms; able to
carry out all usual duties and activities; 2—slight disability. Able to
look after own affairs without assistance, but unable to carry out all
previous activities; 3— moderate disability. Requires some help, but
able to walk unassisted; 4—moderately severe disability. Unable to
walk without assistance and unable to attend to own bodily needs
without assistance; 5—severe disability. Requires constant nursing
care and attention, bedridden, incontinent; 6—dead.[10] ADL scale is
a world-recognized scale with most frequently used in assessing
activities of daily living of stroke patient. It is simple with good
reliability,[11] validity and sensitivity in predicting therapeutic effect,
length of stay and prognosis. The classification of result as Barthel
index was below: functional independence (61–100); moderate
dependence (41–60); and severe dependence (0–40).[12]

Figure 1. The study flow.

admission); owned a smartphone; alive. Informed consent from

patient and caregiver must be acquired before included.

2.2. Application
The application was designed by HRCD Science and Technology
Ltd., focusing on self-assessment of acute stroke patients with
Figure 2. The assessment tools were mRS and ADL scale in Chinese and
friendly interface. It could be installed on smartphone Mac
pushed via software platform. ADL = activities of daily living, mRS = modified
operation system (OS) or Android OS. During the hospitaliza- Rankin scale.
tion, professional nurse helped to introduce and install the

2
Chang et al. Medicine (2018) 97:26 [Link]

2.5. Statistical analysis 13.73, mean ± standard deviation [SD]). While 6 patients were
We used Epidata 3.1 to set up the database with double- drop out with clear private reasons. In the 50 patients, 49 were
personnel data entry, and SPSS18.0 (SPSS Inc, Chicago, IL) to ischemic stroke and one was hemorrhagic stroke. The score of
analyze the data. Self-assessors consisted of patients and National Institute of Health stroke scale was ranged from 0 to 22,
caregivers, which were analyzed and compared to nurses, with 4.60 ± 5.99 (mean ± SD). 27 patients self-assessed the scales,
respectively. Chi-square tests were performed and the difference while caregivers of other 23 patients completed the assessment.
was considered statistically significant when P was <.05. The Age distribution and education level of patient and assessor were
agreement of self-assessment and nurse assessment were illustrated in Figure 3.
measured by kappa coefficient. In comparison with patient assessment and nurse assessment,
significant difference was found in ADL score (P = .004, Table 1),
but was not found in mRS score (P > .05, Table 1). When
2.6. Ethical considerations comparing caregiver assessment with nurse assessment, no
This study was approved by the ethics committee of Capital significant difference could be found either in ADL score
Medical University. All the subjects were informed with the study (P > .05, Table 1) or in mRS score (P > .05, Table 1).
procedures before entering the study. The consistency of ADL score between self-assessment and
nurse assessment was substantial with kappa value 0.720
(P = .000, Table 2). The sensitivity and specificity were 96.8%
3. Results and 82.0%, respectively. While the kappa value of mRS score
Around 56 acute stroke patients were enrolled from stroke unit, was 0.718 (P = .000, Table 2), which meant a substantial
Xuanwu Hospital, Capital Medical University, during September consistency between self-assessment and nurse assessment. The
and October 2016, in which 50 patients (33 males and 17 sensitivity and specificity were 97.6% and 92.4%, respectively
females) completed the study with the age 24 to 88 (59.84 ± (Table 2).

Figure 3. Age distribution and education level of patient and assessor.

3
Chang et al. Medicine (2018) 97:26 Medicine

Table 1
Result of self-assessment (patient or caregiver) and nurse assessment scale.
Patients assessment versus nurse assessment
ADL mRS
Assessment frequency x ±s M (P25, P75) x ±s M (P25, P75)
Patient assessment 27 81.11 ± 26.61 90.00 (75.00, 100.00) 1.93 ± 1.49 1 (1, 3)
Nurse assessment 27 74.81 ± 23.88 80.00 (65.00, 90.00) 2.19 ± 1.50 2 (1, 4)
z 2.850 1.941
p 0.004 0.052

Caregiver assessment versus nurse assessment

ADL mRS
Assessment frequency x ±s M (P25, P75) x ±s M (P25, P75)
Caregiver assessment 23 49.13 ± 34.53 40.00 (20.00, 95.00) 3.52 ± 1.38 4 (3, 4)
Nurse assessment 23 47.83 ± 31.69 45.00 (15.00, 85.00) 3.57 ± 1.41 4 (3, 5)
z 0.715 1.000
p 0.475 0.317
x ± s = mean ± standard deviation, ADL = activities of daily living, M = median, mRS = modified Rankin scale, P25 = 25th percentile, P75 = 25th percentile.

4. Discussion old, well-educated, ever getting benefit from exercise, willing to

get health guidance via media. Our scholars have focused on the
In the recent years, modern information technology has rapidly use of application in the management of stroke. However, the
developed and played an important role in solving the complexity reliability is still unknown.
of health care problems, especially the use of smartphones and This research determines the accuracy of mobile phone
applications provides new methods for the management of application-based scales, which is pushed through software
chronic diseases. Advantages of mobile phone application mainly platform focusing on self-assessment of acute stroke patients. The
include wide use, accurate positioning, instant information application is time and money-saving for data acquisition and
delivery and simple operation.[13,14] Due to the limitation of feedback to physician, and convenient for assessor to operate. It
Chinese economic level and cultural background, there are low can be potentially applied in the populations with different
acceptances of mobile phone application-based investigation by education levels and other backgrounds. Actually, the age of
patients with chronic diseases. In the past survey toward 218 application users (patients or their caregivers) ranged from 24 to
chronic patients,[15] only about half of them (53.7%) would like 76 years old (with the average 51.72 ± 12.40). While education
to make use of smartphone application to improve physical levels ranged from illiteracy to university degree, in which
activities. In which, the individuals below are prone to accepting university degree were 36% and middle-school education were
smartphone application, including those who are below 44 years 62%. Caregiver assessment is much closer to nurse assessment in

Table 2
Consistency test of self-assessment (patient or caregiver) and nurse assessment.
Cross-tabulation of paired ratings (activities of daily living)
Nurse assessment
0-40 41-60 61-100 Total
Self-assessment 0–40 11 3 1 15
41–60 1 4 0 5
61–100 0 3 27 30
Total 12 10 28 50
Kappa coefficient 0.720
P value 0.000

Cross-tabulation of paired ratings (modified Rankin scale)

Nurse assessment
0 1 2 3 4 5 Total
Self-assessment 0 1 3 0 0 0 0 4
1 2 9 3 0 0 0 14
2 0 0 4 0 1 0 5
3 0 0 0 4 1 0 5
4 0 0 0 0 15 1 16
5 0 0 0 0 0 6 6
Total 3 12 7 4 17 7 50
Kappa coefficient 0.718
P value .000

4
Chang et al. Medicine (2018) 97:26 [Link]

both of ADL and mRS scores without significant difference. [2] Feigin VL, Forouzanfar MH, Krishnamurthi R, et al. Global and regional
burden of stroke during 1990-2010: findings from the Global Burden of
Therefore, caregivers can serve as the proper assessor when
Disease Study 2010. Lancet 2014;383:245–54.
patients are out of hospital. The limitation should be considered. [3] Zhao D, Liu J, Wang W, et al. Epidemiological transition of stroke in
This study is a pilot study with small sample size, which must be China: twenty-one-year observational study from the Sino-MONICA-
expanded in the future design. Beijing Project. Stroke 2008;39:1668–74.
In summary, mobile phone application-based scales are widely [4] Roming OM, Staven K. Determinants of change in quality of life from 1
to 6 months following acute stroke. Cerebrovasc Dis 2008;25:67–73.
accepted by patients and their caregivers with different ages and [5] Lynch EA, Cadilhac DA, Luker JA, et al. Education-only versus a
education levels, which are generally accurate, economical and multifaceted intervention for improving assessment of rehabilitation
convenient for self-assessment of acute stroke patients with needs after stroke; a cluster randomised trial. Implement Sci
acceptable reliability in our small scale study. It is promising but 2016;11:120.
[6] Edwards DF, Hahn MG, Baum CM, et al. Screening patients with stroke
still need to be further confirmed how practical to use this
for rehabilitation needs: validation of the post-stroke rehabilitation
application in extended care and follow-up. guidelines. Neurorehabil Neural Repair 2006;20:42–8.
[7] Hakkennes S, Hill KD, Brock K, et al. Selection for inpatient
Author contributions rehabilitation after severe stroke: what factors influence rehabilitation
assessor decision-making? J Rehabil Med 2013;45:24–31.
Conceptualization: Yuchen Qiao. [8] Newcommon NJ, Green TL, Haley E, et al. Improving the assessment of
Data curation: Hong Chang. outcomes in stroke: use of a structured interview to assign grades on the
modified Rankin Scale. Stroke 2003;34:377–8.
Formal analysis: Jie Zhao.
[9] Saver JL, Filip B, Hamilton S, et al. Improving the reliability of stroke
Funding acquisition: Hong Chang. disability grading in clinical trials and clinical practice: the Rankin
Investigation: Xiaojuan Wang. Focused Assessment (RFA). Stroke 2010;41:992–5.
Methodology: Hui Yao. [10] Banks JL, Marotta CA. Outcomes validity and reliability of the modified
Project administration: Hong Chang. Rankin scale: implications for stroke clinical trials: a literature review
and synthesis. Stroke 2007;38:1091–6.
Resources: Juanmin Li. [11] Collin C, Wade DT, Davies S, et al. The Barthel ADL index: a reliability
Software: Yuchen Qiao, Juanmin Li. study. Int Disabil Stud 1988;10:61–3.
Validation: Xiaojuan Wang, Juanmin Li. [12] Wade DT, Collin C. The Barthel ADL index: a standard measure of
Writing – original draft: Hong Chang, Jia Liu. physical disability? Int Disabil Stud 1988;10:64–7.
[13] Jeon E, Park HA. Development of a smartphone application for
Writing – review & editing: Jia Liu.
clinical-guideline-based obesity management. Healthc Inform Res
2015;21:10–20.
References [14] Montag C, Błaszkiewicz K, Sariyska R, et al. Smartphone usage in the
21st century: who is active on WhatsApp? BMC Res Notes 2015;8:331.
[1] Yusuf S, Reddy S, Ounpuu S, et al. Global burden of cardiovascular [15] Sun L, Jiao C, Wang Y, et al. A survey on the willingness to use physical
diseases: part I: general considerations, the epidemiologic transition, risk activity smartphone applications (apps) in patients with chronic diseases.
factors, and impact of urbanization. Circulation 2001;104:2746–53. Stud Health Technol Inform 2016;225:1032–3.