Translation, Inter-rater Reliability, Agreement, and Internal Consistency of the Japanese Version of the Cumulated Ambulation Score in Patients after Hip Fracture Surgery.

OBJECTIVES: The aim of this study was to translate the Cumulated Ambulation Score (CAS) from English into Japanese in cooperation with different types of healthcare providers and to investigate its inter-rater reliability and internal consistency.
METHODS: Two physical therapists at each of three general hospitals in Japan measured the mobility of 50 consecutive post-operative hip fracture patients on two occasions between 2 and 6 days after surgery using the Japanese version of the CAS (CAS-JP). We analyzed the inter-rater reliability and agreement using both the linear weighted kappa and the interclass correlation coefficient; we also analyzed the internal consistency using Cronbach's alpha coefficient.
RESULTS: The mean age of patients was 81 (SD: 11.6) years and 82% were women. Approximately half of the patients had severe cognitive impairment. Kappa was ≥ 0.93 for the three mobility activities and for the total CAS-JP score, the percentage agreement was ≥ 0.98, the ICC was ≥ 0.95, and Cronbach's alpha coefficient was 0.85.
CONCLUSIONS: We found that the CAS-JP possessed good inter-rater reliability, agreement, and internal consistency. The CAS-JP is a reliable and easy-to-use evaluation tool suitable for daily clinical practice across different healthcare providers to monitor mobility in older hip fracture patients in Japan. We suggest that CAS-JP be evaluated in future studies for use in younger patients and in other patient groups with mobility problems. ©2020 The Japanese Association of Rehabilitation Medicine.

INTRODUCTION

Post-operative hip fracture patients experience a loss of mobility.[1],[2]) To reduce morbidity, mortality, and improve physical function, early mobilization and rehabilitation after surgery is recommended.[3]) Because a large majority of patients who undergo hip fracture surgery are elderly and less physically active in general, healthcare providers need valid and reliable measurement methods to monitor the level and improvement of mobility of individual patients. Not only patients with hip fracture,[4],[5],[6],[7]) but also geriatric patients,[8],[9]) patients with community-acquired pneumonia,[10]) and patients who undergo acute high-risk abdominal surgery or knee-arthroplasty experience different levels of in-hospital mobility problems.[11],[12]) Therefore, by monitoring the early mobility function after admission or surgery, healthcare providers will be able to better plan and optimize patient care.

The Cumulated Ambulation Score (CAS) is a valid and reliable evaluation tool for assessing a patient’s mobility by observing three basic activities[3]): (1) getting in and out of bed, (2) sitting and rising from a chair (with armrests), and (3) indoor walking (with or without a walking aid).[13]) Each of these three CAS activities is scored from 0 (unable to perform) to 1 (assistance required) to 2 (independent) and the scores are summed to provide a one-day total CAS score of 0–6 points, with 6 points indicating an independent ambulatory status.[13]) The CAS is routinely used in other countries for patients with hip fracture,[3],[4],[5]) for those undergoing joint replacement,[14],[15]) and for other elderly health disorders.[11]) CAS scores cumulated for postoperative days 1–3 (3-day CAS) has proven valid as a prediction of the short-term clinical outcome of patients after hip fracture surgery.[3]) Moreover, the CAS shows a lower floor effect than the de Morton Mobility Index and the Chair Stand Test for physically frail populations, such as patients with hip fracture.[16]) Therefore, the CAS can be used to monitor the trajectory of a patient’s mobility function, especially in the crucial early stage after hip fracture surgery.

The CAS was developed in Denmark; it was subsequently translated into several languages including English, Italian, Spanish, and Turkish and is now used in many countries.[17],[18],[19]) Approved versions are also available in Swedish, Norwegian, Indonesian-Bahasa, and French, and the CAS is used at the national level in the Danish and Irish multidisciplinary hip fracture database. In Japan, the mean age of patients with hip fracture has been increasing, as has also been seen in other countries.[20],[21]) Furthermore, the life expectancy of people in Japan is one of the highest worldwide. Consequently, healthcare providers in Japan need valid and reliable measurements to monitor and guide the rehabilitation needs of the aged population.

The CAS already has been used in Japan and has proven to be valid in reflecting the association of surgery within 24 h of hospital admission and the ambulatory status after hip fracture surgery.[4]) However, although the CAS used in that study was in Japanese and was based on the first English version of the CAS,[3]) that Japanese version was not translated according to international guidelines. Consequently, there is a need for an official Japanese version of the CAS. By translation and cross-cultural validation of the CAS into an official Japanese version, the CAS can be used in Japan as an easily applicable measurement of patients with basic mobility problems, thereby further promoting use of the CAS as an international multidisciplinary outcome measurement. The aim of this study was to translate and cross-culturally validate the English version of the CAS into Japanese and to investigate the inter-rater reliability, agreement, and internal consistency of the Japanese version of the CAS (CAS-JP).

METHODS

Translation Process

We followed the recommendations of Ramada-Rodilla et al.[22]) to translate the English version of the CAS manual and score-sheet into Japanese.[23]) Two Japanese physical therapists, one who is fluent in English (HH) and one who uses English in daily clinical practice in the U.S. (FI), independently translated the English version of the CAS manual and score-sheet into Japanese, taking the Japanese culture and language into consideration while maintaining the original intention of the test. A CAS-JP committee synthesized these two Japanese translations of CAS. A back translation of the synthesized version was conducted by a third person (JM) who is a native speaker of American English, holds a U.S. nursing license, and has experience in U.S. hospital wards, without knowledge of the original version of the CAS. This back translated English version was forwarded to Dr. Kristensen, who is one of the original developers of CAS, to ensure that the core concepts of the score remained intact.[3],[13]) We subsequently modified the Japanese CAS with some word corrections to retain the original CAS scoring procedure. Two physical therapists from each institution conducted pilot measurements with the provisional Japanese CAS; one of these physical therapists communicated with Dr. Kristensen, and the other was a novice CAS user. First, these two raters preliminarily measured CAS-JP to confirm all measurements. If additional comments were presented during this process, they were shared and described on the CAS-JP score manual with Dr. Kristensen’s agreement. Once the modification process was complete, the final version of the Japanese CAS (CAS-JP) was approved by CAS-JP committee members. CAS-JP is available as a supplement (Appendix 1).

The Cumulated Ambulation Scores

The CAS is a composite measurement that consists of the following three basic mobility activities: (i) getting in and out of bed with the sequence of events as follows: the patient starts in the supine position on the bed, moves to sitting, stands or transfers to a chair next to the bed, then returns to sitting on the bed, and then to the supine position on the bed (with or without assistance or aids); (ii) sit to stand to sit from a chair with armrests (with or without assistance or aids), and (iii) walking indoors (with or without walking aids).[13],[23]) These three basic activities are each allocated 0–2 points and the three scores are combined for a total one-day CAS score ranging from 0 (minimum) to the maximum of 6 points. A CAS score of 0 indicates total dependence, whereas 6 points indicates complete independence in mobility activities. A score of 2 was given to patients who could perform the activity without verbal or physical assistance, a score of 1 was given to those who required verbal or physical assistance by one or more persons, and a score of 0 was given for patients who were unable to do the activity despite human assistance.[13],[23])

Study Population

A total of 50 consecutive hip fracture patients who underwent reparative surgery and were admitted between January and June 2020 to three general hospitals (one urban, one suburban, and one rural hospital: 18, 12, and 20 patients, respectively) in Japan were analyzed. We included patients with hip fracture (trochanteric fracture or femoral neck fracture) aged 65 years and older. Patients with multiple trauma, loss of consciousness, or immobility before injury were excluded. The following descriptive information was extracted from each patient’s medical chart: weight, height, type of fracture, and type of surgery. Additionally, the following baseline demographics were evaluated at hospital admission: age, sex, body mass index (BMI), walking ability before injury, walking aid use before injury, the affected side, the type of fracture, the surgical procedure, and the results of the Pfeiffer short portable mental state questionnaire (SPMSQ, 0–10 points[24])). The patient’s walking ability and walking aid use was obtained from the patient or from the patient’s family if the patient had cognitive impairment. Walking ability was then classified into one of four categories: able to walk outside for 15 min or more, able to walk outside for less than 15 min, able to walk only inside, or unable to walk. Walking aid was categorized as: no aid use, use of a cane or walker outside only, or use of a cane or walker both inside and outside. The fracture type was assessed as a femoral neck fracture or a trochanteric fracture according to the fracture site. The surgical procedure was categorized as open reduction with internal fixation or hemiarthroplasty. With respect to cognitive function, we assessed the SPMSQ score between 7 and 14 days post-surgery to avoid possible acute delirium that would have influenced the evaluation. SPMSQ scores were categorized as follows: 0–2 points as no cognitive impairment, 3–4 points as mild cognitive impairment, 5–7 points as moderate cognitive impairment, and >8 points as severe cognitive impairment.[13],[23],[24])

Inter-rater Reliability

Inter-rater reliability for the CAS-JP was evaluated according to the Guidelines for Reporting Reliability and Agreement.[25]) The inter-rater reliability of the CAS-JP was assessed between two physical therapy raters at each of the three general hospitals (six physical therapists in total); each therapist did the first measurement of half of the consecutive hip fracture patients who had undergone surgery at their respective hospital. The first assessment of each patient by a rater from physical therapist group A was made as part of clinical practice before rehabilitation within postoperative days 2–6 and was followed by assessment by the other rater (physical therapist group B) on the same day. To reduce possible bias resulting from fluctuations in patients’ daily mobility function, the second rater measured the same patient 2 h after the first rater; the raters performed the measurements independently. No discussion of the ratings was allowed until the end of the study.[26]) Two days after the first assessment, all patients were evaluated again using the same procedure. Consequently, each patient was evaluated twice by each rater, which sums to a total of 200 CAS assessments conducted by the six raters. Therefore, reliability estimates were based on comparison of 100 CAS evaluations by rater group A and 100 by rater group B. If the planned day of assessment was a national holiday or an out-of-service day, the measurement day was adjusted accordingly.

Statistical Analysis

Continuous variables are presented as means with standard deviations (SDs) or as medians with interquartile ranges according to their distributions; categorical variables are described as numbers with percentages. We estimated linear weighted kappa values to assess the inter-rater reliability for the three mobility activities and the total CAS-JP score.[27]) A kappa value less than 0.2 was considered as slight agreement, between 0.21 and 0.40 as fair agreement, between 0.41 and 0.60 as moderate agreement, between 0.61 and 0.80 as substantial agreement, and more than 0.80 as almost perfect agreement.[28]) The percentage agreement between the two raters and the prevalence of CAS-JP scores for the three mobility activities and the total CAS-JP scores were described. To evaluate the measurement error for the three mobility activities and the total CAS-JP score, we also calculated the interclass correlation coefficient (ICC) with a two-way mixed model based on consistency measurement.[29]) The ICC ranges from 0 to 1, and cut off values were assessed as follows: poor agreement for ICC lower than 0.40, fair agreement for ICC ranging from 0.40 to 0.59, good agreement for ICC ranging from 0.60 and 0.74, and excellent agreement for more than 0.75.[30]) ICC values were used to calculate the standard error of measurement [SEM=SD × √(1-ICC)] and the smallest real difference (SRD=SEM × √2 × 1.96) to allow comparison of the measurement error with previous original CAS research performed in English.[13],[27],[31]) The score differentiation between the two raters was also visualized using a Bland-Altman plot. To test for internal consistency, Cronbach’s alpha coefficient was estimated.[32]) Cronbach’s alpha coefficient ranges from 0 to 1, and values lower than 0.70 were regarded as unacceptable, values between 0.70 and 0.80 as acceptable, and values higher than 0.80 as good.[33],[34]) The sample size was calculated according to the criteria of Nunnally and Hoskins et al. A total of 45 to 50 study patients were planned.[35],[36]) Written informed consent was obtained from patients or their relatives. Study approval was obtained from the institutional review board (E19HS-010, 19-R097). The type I error probability was set to 0.05 for all analyses. All statistical analyses were performed using Stata version 16.1 (Stata Corp, College Station, TX, USA).

RESULTS

During the translation process, there were two major considerations: one was the difference of language use and the other was the discrepancy in bedside settings across hospitals. First, when we translated the original English version of CAS into Japanese (CAS-JP), the phrase “get in/out of bed” was freely translated into Japanese, resulting in the implication of different mobility actions. Therefore, we prioritized the original meaning rather than the technical term used in the physical therapeutic field, rishou, which mainly means getting out of bed but does not include getting into bed. The second challenging point was the concept of independence when it comes to putting on footwear. Different hospitals used different types of footwear, such as slippers or shoes. The original version of the CAS did not include any variance in footwear such as room shoes or slippers for the measurement of independence. Therefore, to maintain consistency across hospitals, we did not include putting on footwear in evaluating independence and added this point to the manual.

A total of 50 consecutive hip fracture patients were evaluated twice (with an interval of 2 h) on two separate occasions (200 measurements in total) using the CAS-JP. Evaluations were carried out by six physical therapists independently in three general hospitals in Japan (Table 1). All patients were evaluated between 2 and 10 days post-surgery.

Table 1.

 Patient demographics (n=50) and observed scores of the Japanese version of the Cumulated Ambulation Score

	Total (N=50)
Age (years)	80.68 (11.55)
Sex
Female	41 (82)
Male	9 (18)
Body mass index (kg/m2)	20.44 (3.19)
Pre-fracture walking ability
Able to walk outside for 15 min or more	20 (40)
Able to walk outside less than 15 min	11 (22)
Able to walk inside only	19 (38)
Unable to walk	0 (0)
Pre-fracture use of walking aid
No use	18 (36)
Use only outside (cane or walker)	19 (38)
Use both inside and outside (cane or walker)	13 (26)
Fracture side
Right	23 (46)
Left	27 (54)
Fracture type
Femoral neck	34 (68)
Trochanteric	16 (32)
Surgical procedure
Open reduction with internal fixation	25 (50)
Hemiarthroplasty	25 (50)
Short portable mental state questionnaire
No cognitive impairment	6 (12)
Mild cognitive impairment	6 (12)
Moderate cognitive impairment	11 (22)
Severe cognitive impairment	27 (54)
Cumulated Ambulation Score (CAS)
Getting in and out of bed (0–2)	1.10 (0.48)
Sit-to-stand-to-sit from chair with armrests (0–2)	1.30 (0.56)
Walking indoors with or without an aid (0–2)	0.63 (0.72)
Total one-day CAS (0–6)	3.02 (1.50)

Values are given as the number of patients with the percentage of the cohort in parentheses, with the exception of age, body mass index (BMI), short portable mental state questionnaire (SPMSQ), and cumulated ambulation score (CAS) which are given as the mean and standard deviation. The total number of CAS observations was 200 because measurement by raters from groups A and B were taken on two occasions.

The mean age of patients was 81 (SD, 11.6) years and the majority were women (n=41, 82%). In terms of the pre-fracture mobility, 62% of patients were able to walk outside before injury, and more than a two-thirds of patients either required no use of a walking aid or use of a walking aid outside only. With respect to hip fracture, 68% of patients had femoral neck fracture and half of the patients underwent hemiarthroplasty. Approximately half of the patients had severe cognitive impairment (Table 1).

Tables 2 and 3 show the results for inter-rater reliability. The weighted kappa for the three mobility activities (getting in and out of bed, sit-to-stand-to-sit from an armchair, walking indoors with or without an aid) and the total CAS-JP were excellent (0.96, 0.93, 0.97, and 0.97, respectively). The corresponding percentage agreement also was excellent (≥98%) (Table 2).

Table 2.

 Inter-rater reliability and agreement of the Japanese version of the Cumulated Ambulation Score between scores taken by two different physical therapists (1)

Activity	Linear weighted Kappa value (95%CI)	Observed agreement, n (%)	CAS score 0 to 2, n (%)
			0	1	2
Getting in and out of bed (0–2)	0.96 (0.81–1.0)	99 (99)	15 (7.5)	144 (72.0)	41 (20.5)
Sit-to-stand-to-sit from chair with armrests (0–2)	0.93 (0.77 − 1.0)	98 (98)	12 (6.0)	114 (57.0)	74 (37.0)
Walking indoors with or without an aid (0–2)	0.97 (0.82 − 1.0)	99 (99)	102 (51.0)	68 (34.0)	30 (15.0)
Total CAS (0–6)	0.97 (0.84 − 1.0)	99 (99)	n/a	n/a	n/a

CI, confidence interval.

Table 3.

 Inter-rater reliability and agreement of the Japanese version of the Cumulated Ambulation Score between scores taken by two different physical therapists (2)

Activity	ICC (95%CI)	SEM	SRD
Getting in and out of bed (0–2)	0.96 (0.95–0.97)	0.10	0.28
Sit-to-stand-to-sit from chair with armrests (0–2)	0.94 (0.92–0.96)	0.14	0.39
Walking indoors with or without an aid (0–2)	0.98 (0.97–0.99)	0.10	0.28
Total CAS (0–6)	0.99 (0.98–0.99)	0.17	0.47

ICC, intraclass correlation coefficients; SEM, standard error of measurement; SRD, smallest real difference.

The ICC showed an excellent result (≥0.94) for the three mobility activities and the one-day CAS-JP (Table 3). The SEM and the SRD for the one-day CAS-JP were 0.17 and 0.47. The largest difference in scores between the two raters was 1 point, and the Bland-Altman plot indicated that there was no systematic between-rater bias over the three mobility activities or for the total CAS-JP scores (Figure 1). Cronbach’s alpha coefficient for the CAS-JP was 0.85.

Fig. 1.

 A Bland-Altman plot of two physical therapist raters for the Japanese version of the Cumulated Ambulation Score.

DISCUSSION

In this study, the English version of the CAS was translated and cross-culturally validated into Japanese based on the guideline used in previous research in which the CAS was translated.[22]) CAS-JP showed excellent inter-rater reliability, high agreement, and excellent internal consistency in our research patients. In our analysis, a total of 50 patients were enrolled and 200 measurements in total were carried out by the two physical therapist rater groups (groups A and B). The patients were evaluated in three hospitals located in urban (Tokyo), suburban (Saitama), and rural (Nagano) areas. The baseline characteristics of our data represent the general hip fracture population in Japan.[21])

As the age of patients with hip fracture increases in Japan along with the aging of the population, CAS-JP as applied to the hip fracture population can contribute to geriatric clinical research. Also, because life expectancy and the incidence of hip fracture are increasing around the world, the results of clinical research on patients with hip fracture in Japan can contribute to health care in other countries where there are predictions of a future increase in the number of hip fractures. Although our patients were slightly older than those in previous studies in Denmark, Spain, and Turkey, the reliability and consistency of the translated CAS-JP were consistent with previous studies.[17],[18],[19])

Because the CAS can be used with high reliability by different healthcare providers, including physical therapists, occupational therapists, and physicians,[13],[17],[18],[19]) we also included a nurse in the translation process to expand knowledge of the CAS across healthcare providers. The use of the CAS to monitor the progress of early recovery after hip fracture surgery and of other patient groups by different healthcare providers has been demonstrated. There is scope for further application of the CAS to a wider range of age groups with or without comorbidities across a broad range of countries.

Strength and Limitations

One strength of this study is that the excellent reliability estimates were based on evaluations by six physical therapists: two at each of the three hospitals forming rater groups A and B. Nonetheless, our research also had some limitations. First, we did not externally validate the CAS-JP, although the original version of the CAS was validated in previous studies.[16],[18]) Therefore, further study is needed to confirm the validity of the CAS-JP and to compare results with previous studies. Second, our study was restricted to general hospitals in Japan, although we achieved some diversification by selecting study sites in urban, suburban, and rural hospitals. Even so, these study sites may not represent all hospital settings across Japan. Finally, the number of patients with a CAS score of 0 was relatively small, whereas patients with a CAS score of 1 or 2 were sufficiently covered. Our data collection started from post-operative day 2, which explains the low number of patients with a CAS score of 0. However, a CAS score of 0, which means the inability to perform any mobility activity (e.g., a completely bedridden patient), should be relatively easy to evaluate and we assumed it to be reasonable to draw conclusions without a larger number of patients with a CAS score of 0.

We translated the original English version of the CAS into Japanese and demonstrated its excellent inter-rater reliability and consistency based on robust methods. The CAS is reliable and easy to use in daily clinical practice to monitor mobility by different healthcare providers. We therefore suggest that the CAS-JP be measured broadly for older patients with hip fracture in Japan and that it be evaluated throughout hospitals in Japan for use in younger patients and in other patient groups with mobility problems.