Development of A Simplified Chinese Version of the Hip Disability and Osteoarthritis Outcome Score (hoos): Cross-Cultural Adaptation and Psychometric Evaluation
X. Wei,Z. Wang,C. Yang,B. Wu,X. Liu,H. Yi,Z. Chen,F. Wang,Y. Bai,J. Li,X. Zhu,M. Li
DOI: https://doi.org/10.1016/j.joca.2012.08.018
IF: 7.507
2012-01-01
Osteoarthritis and Cartilage
Abstract:Osteoarthritis (OA) has a profound impact on health-related quality of life1Veenhof C. Huisman P.A. Barten J.A. Takken T. Pisters M.F. Factors associated with physical activity in patients with osteoarthritis of the hip or knee: a systematic review.Osteoarthritis Cartilage. 2012; 20: 6-12Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar. Increasing importance has been attached to utilization of disease-specific, self-reported outcome measures2Thorborg K. Roos E.M. Bartels E.M. Petersen J. Hölmich P. Validity, reliability and responsiveness of patient-reported outcome questionnaires when assessing hip and groin disability: a systematic review.Br J Sports Med. 2010; 44: 1186-1196Crossref PubMed Scopus (126) Google Scholar, such as the Hip Disability and Osteoarthritis Outcome Score (HOOS) instrument3Klassbo M. Larsson E. Mannevik E. Hip disability and osteoarthritis outcome score. An extension of the Western Ontario and McMaster Universities Osteoarthritis Index.Scand J Rheumatol. 2003; 32: 46-51Crossref PubMed Scopus (313) Google Scholar. China is the most populous country in the world with 1.3 billion people. Hence, we translated and adapted the HOOS into a Simplified Chinese version (SC-HOOS) and validated it in a cohort of native Chinese-speaking patients with hip OA, relative to the Short Form Health Survey (SF-36), a visual analog scale (VAS), and the Harris hip score (HHS) test. Psychometric testing for internal consistency, test–retest reliability, construct validity, and responsiveness was conducted. The SC-HOOS showed satisfactory internal consistency, test–retest reliability, construct validity, and responsiveness when evaluated in Chinese-speaking patients with hip OA. A total of 131 consecutive patients (58 men, 73 women) with a diagnosis of primary hip OA were recruited from the Department of Orthopedics of our medical university between December 2010 and August 2011, and enrolled in accordance with the quality criteria described by Terwee et al.4Terwee C.B. Bot S.D. de Boer M.R. van der Windt D.A. Knol D.L. Dekker J. et al.Quality criteria were proposed for measurement properties of health status questionnaires.J Clin Epidemiol. 2007; 60: 34-42Abstract Full Text Full Text PDF PubMed Scopus (6038) Google Scholar They had a mean age of 51.3 (9.1) years and a mean body mass index (BMI) of 25.2 (3.0) kg/m2. The project was approved by the Human Research Ethics Committee of our institution, and each patient signed a written informed consent. The inclusion criteria were: age >18 years, ability to read and speak Chinese, and primary hip OA diagnosis according to the criteria of the American College of Rheumatology5Altman R. Alarcon G. Appelrouth D. Bloch D. Borenstein D. Brandt K. et al.The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hip.Arthritis Rheum. 1991; 34: 505-514Crossref PubMed Scopus (1537) Google Scholar. The exclusion criteria were: history of leg or spine surgery, tumors, infection, rheumatologic disease, ankylosing spondylitis, and/or neuropathologies; inability or unwillingness to complete questionnaires independently. Patients in whom surgical treatment [total hip replacement (THR) subgroup] was deemed necessary (N = 52) were allowed 3 months postoperatively to finish their questionnaires. Statistical analysis was performed using the Statistical Package for the Social Sciences, version 18.0 (SPSS, Chicago, IL). Mean values are reported with standard deviations (SDs). Intraclass correlation coefficient (ICC) values are reported with 95% confidence intervals (CIs). P values of <0.05 were considered significant. The translation and cross-cultural adaptation were performed according to previously published guidelines regarding dual forward translation, synthesis of the dual translations using resolution by consensus, backward translation into English to reveal any discrepancies, reconciliation by expert committee consensus, and a test of the pre-final SC-HOOS6Guillemin F. Bombardier C. Beaton D. Cross-cultural adaptation of health-related quality of life measures: literature review and proposed guidelines.J Clin Epidemiol. 1993; 46: 1417-1432Abstract Full Text PDF PubMed Scopus (4008) Google Scholar, 7Beaton D.E. Bombardier C. Guillemin F. Ferraz M.B. Guidelines for the process of cross-cultural adaptation of self-report measures.Spine (Phila Pa 1976). 2000; 25: 3186-3191Crossref PubMed Scopus (6664) Google Scholar. Our examination of the final SC-HOOS (see Appendix) was consistent with recommendations for cross-cultural validation studies of patient-reported outcomes8Roos E.M. Davis A.M. Recommendations for publication of cross-cultural validation studies of patient-reported outcomes (PROs) in Osteoarthritis and Cartilage.Osteoarthritis Cartilage. 2012; 20: 4-5Abstract Full Text Full Text PDF PubMed Scopus (2) Google Scholar. All 131 patients completed the SC-HOOS, the SF-36, VAS, and HHS in an outpatient hospital room. Patients were asked to finish the SC-HOOS first, before the other tests, and the time frame for completing the SC-HOOS was 15 min. The HOOS includes five subscales: pain (10 items), other symptoms (10 items), function in daily living (ADL) (17 items), function in sports and recreation (Sport/Rec) (four items), and hip-related quality of life (QoL) (four items). A five-point Likert scale was used and each item was given a score of 0 (none), 1 (mild), 2 (moderate), 3 (severe), or 4 (extreme). All of the item scores within each subscale were summed, divided by the maximum score, and then deducted from 100, such that more extreme symptoms resulted in a larger deduction from 100, and thus a lesser subscale score. The normalized scores, from 0 (indicating extreme symptoms) to 100 (indicating no symptoms), for each subscale were plotted in an outcome profile for each participant. The HHS is a multidimensional observational assessment that contains questions about pain, function, deformity, and range of motion, with a total score ranging from 100 (no disability) to 0 (maximum disability)9Harris W.H. Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty. An end-result study using a new method of result evaluation.J Bone Joint Surg Am. 1969; 51: 737-755PubMed Google Scholar. The SF-36 contains eight domains: physical functioning (PF), role-physical (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role-emotional (RE), and mental health (MH). Each raw subscale scores was transformed to a 100-point scale. The SF-36 has been translated into Chinese and thoroughly tested. Finally, the VAS allows patients to rate pain intensity along a 100-mm line ranging from "no pain" to "worst pain imaginable". Floor and ceiling effects exceeding 15% were considered significant4Terwee C.B. Bot S.D. de Boer M.R. van der Windt D.A. Knol D.L. Dekker J. et al.Quality criteria were proposed for measurement properties of health status questionnaires.J Clin Epidemiol. 2007; 60: 34-42Abstract Full Text Full Text PDF PubMed Scopus (6038) Google Scholar. The SC-HOOS subscale scores were well-distributed, with no floor or ceiling effects. To evaluate acceptability, patients were asked about any difficulties that had been encountered. The data were checked for missing or multiple responses and the completeness of the SC-HOOS was calculated. Missing data were treated as recommended by Nilsdotter et al.10Nilsdotter A.K. Lohmander L.S. Klassbo M. Roos E.M. Hip disability and osteoarthritis outcome score (HOOS)-validity and responsiveness in total hip replacement.BMC Musculoskelet Disord. 2003; 30: 10Crossref Scopus (644) Google Scholar Most (127/131; 96.9%) of the patients responded that they did not have any difficulties with completing the SC-HOOS. The remaining four patients left the question unanswered. The average time for patients to finish the SC-HOOS was 10.4 (3.2) min, similar to that reported previously. The total numbers of improperly answered items in the total test group, the retest subgroup, and the THR subgroup were relatively few at 352/5240 (3.6%), 143/2400 (6.0%), and 97/2080 (4.7%), respectively. The correct completion rates for the entire SC-HOOS were 96.4%, 94.0%, and 95.3% for total test group, the retest subgroup, and the THR subgroup, respectively. Internal consistency of the SC-HOOS subscales was evaluated by calculating Cronbach's α coefficient, where α > 0.80 and α > 0.90 were regarded as good and excellent, respectively4Terwee C.B. Bot S.D. de Boer M.R. van der Windt D.A. Knol D.L. Dekker J. et al.Quality criteria were proposed for measurement properties of health status questionnaires.J Clin Epidemiol. 2007; 60: 34-42Abstract Full Text Full Text PDF PubMed Scopus (6038) Google Scholar. The Cronbach's α coefficients for the subscales (see Table I) were high (0.865–0.968), especially for the pain and ADL subscales, indicating good internal consistency.Table IInternal consistency, construct validity, reliability, and responsiveness of the SC-HOOSParameterSC-HOOS subscale (No. items)Symptoms (5)Pain (10)ADL (17)Sport/Rec (4)QoL (4)Internal consistencyCronbach's α0.8830.9390.9680.8650.87Construct validity indicated by Pearson correlation coefficient, r (P value), vs indicated instrumentsSF-36 domains PF0.771 (<0.0001)0.728 (<0.0001)0.769 (<0.0001)0.733 (<0.0001)0.734 (<0.0001) RP0.420 (<0.0001)0.441 (<0.0001)0.451 (<0.0001)0.418 (<0.0001)0.427 (<0.0001) BP0.636 (<0.0001)0.701 (<0.0001)0.628 (<0.0001)0.666 (<0.0001)0.712 (<0.0001) GH0.508 (<0.0001)0.485 (<0.0001)0.510 (<0.0001)0.412 (<0.0001)0.446 (<0.0001) VT0.230 (0.0082)0.291 (0.0008)0.276 (0.0014)0.211 (0.0157)0.261 (0.0026) SF0.463 (<0.0001)0.458 (0.0001)0.473 (<0.0001)0.429 (0.0001)0.479 (<0.0001) RE0.248 (0.0043)0.252 (0.0037)0.266 (0.0021)0.239 (0.0061)0.308 (0.0003) MH0.221 (0.0113)0.182 (0.0374)0.190 (0.0299)0.264 (0.0023)0.217 (0.0129)VAS−0.765 (<0.0001)−0.786 (<0.0001)−0.766 (<0.0001)−0.714 (0.0001)−0.777 (<0.0001)HHS0.898 (<0.0001)0.848 (<0.0001)0.887 (<0.0001)0.827 (0.0001)0.893 (<0.0001)Test–retest reliability, mean (SD) or ICC value (CI range)Test score46.3 (18.0)46.6 (17.2)46.2 (18.1)42.5 (16.9)44.6 (19.7)Retest score48.2 (16.9)47.3 (16.8)45.8 (17.6)44.1 (17.0)42.4 (19.8)Score change1.8 (6.0)0.8 (4.0)−0.4 (5.4)1.6 (8.9)−2.2 (8.2)ICC (95% CI)0.940 (0.902–0.964)0.973 (0.955–0.984)0.956 (0.921–0.973)0.862 (0.780–0.915)0.913 (0.859–0.947)Responsiveness pre-THR vs 3 months after THR, mean (SD)*N = 52; higher scores represent less pain. For comparison, the HHS yielded values of 28.7 (6.9), 51.3 (15.5), 22.6 (9.6), 3.28, and 2.35, respectively, for the responsiveness parameters.Pre-THR score27.8 (11.5)31.8 (8.7)30.0 (10.6)27.7 (10.6)27.7 (9.8)Post-THR score57.4 (9.1)60.8 (11.1)58.3 (9.6)54.5 (9.6)53.7 (8.4)Change29.6 (9.5)29.0 (12.8)28.3 (13.1)26.8 (12.3)29.4 (13.4)ES2.573.332.672.533.01SRM3.122.272.162.182.19∗ N = 52; higher scores represent less pain. For comparison, the HHS yielded values of 28.7 (6.9), 51.3 (15.5), 22.6 (9.6), 3.28, and 2.35, respectively, for the responsiveness parameters. Open table in a new tab Sixty patients were randomly selected to be in the retest subgroup according to a computer generated randomized number table. These patients were asked to complete the SC-HOOS again at home 7 days after they had completed it the first time, and then to return it by mail once they finished it. A 1-week interval was chosen because it is too brief for obvious post-treatment clinical changes to be apparent, and also because 1 week is the time that THR patients needed to wait for surgery, enabling them to participate in the reliability evaluation. An ICC (two-way random effects model) was calculated to quantify test–retest reliability. An ICC value of 0.60–0.80 indicated good reliability, and an ICC >0.80 indicated excellent reliability11Landis J.R. Koch G.G. The measurement of observer agreement for categorical data.Biometrics. 1977; 33: 159-174Crossref PubMed Scopus (51055) Google Scholar. Mean subscale scores, score changes from test to retest, ICCs, and CIs are listed in Table I. The inter-test ICCs (0.862–0.973) indicated excellent test–retest reliability, indicating that this instrument yields similar results across trials and thus has excellent stability. Bland–Altman plots12Bland J.M. Altman D.G. Measuring agreement in method comparison studies.Stat Methods Med Res. 1999; 8: 135-160Crossref PubMed Scopus (6646) Google Scholar (Fig. 1) of test–retest differences between subscale measures for individual patients relative to overall means of the two sessions showed no systematic bias between sessions, indicating good test–retest agreement of the subscales13Ornetti P. Parratte S. Gossec L. Tavernier C. Argenson J.N. Roos E.M. et al.Cross-cultural adaptation and validation of the French version of the hip disability and osteoarthritis outcome score (HOOS) in hip osteoarthritis patients.Osteoarthritis Cartilage. 2010; 18: 522-529Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar. Good construct validity means that a questionnaire correlates well with measures of the same construct (convergent validity) while correlating poorly with measures of different constructs (divergent or discriminant validity). Pearson correlation coefficients (r) of the SC-HOOS subscales scores were determined relative to scores from the eight SF-36 domains, the VAS, and the HHS. The correlations were judged as poor (r = 0–0.20), fair (r = 0.21–0.40), moderate (r = 0.41–0.60), very good (r = 0.61–0.80), or excellent (r = 0.81–1.0). It was hypothesized that the SC-HOOS subscales should correlate strongly with the PF and BP domains of the SF-36 (especially for pain and ADL), moderately with the GH, RP, and SF domains, and poorly with the VT, MH, and RE (mental health related) domains of the SF-36. The SC-HOOS subscales were expected to correlate well with the HHS and VAS. The construct validity evaluation results obtained by comparing SC-HOOS subscales versus SF-36 domains, VAS, and HHS are shown in Table I. Pearson correlation analyses between the SC-HOOS and the SF-36 indicated that all of the SC-HOOS subscales correlated robustly with the PF (r = 0.728–0.771, P < 0.0001) and BP (r = 0.628–0.712, P < 0.0001) domains of the SF-36. SC-HOOS subscales correlated moderately with the RP (r = 0.418–0.451, P < 0.0001), GH (r = 0.412–0.510, P < 0.0001), and SF (r = 0.429–0.479, P < 0.0001) domains of the SF-36, and correlated weakly with the VT (r = 0.211–0.291, P = 0.0008–0.0152), RE (r = 0.239–0.308, P = 0.0004–0.0061), and MH (r = 0.182–0.264, P = 0.0023–0.0374) domains. All five SC-HOOS subscales correlated strongly with the VAS results (r = −0.714–0.786, P < 0.0001) and the HHS results (r = 0.827–0.898, P < 0.0001). The SC-HOOS exhibited satisfactory construct validity. Its strong correlations with the PF and BP domains of the SF-36, HHS, and VAS indicate that the SC-HOOS has good convergent validity. Meanwhile, its weak correlations with the VT, RE, and MH domains of the SF-36 demonstrate the divergent or discriminant validity of the SC-HOOS. There were excellent correlations between the SC-HOOS subscales and the HHS, which measures some similar constructs (pain and function). However, it is worth pointing out that the HHS is administered by a physician, whereas the HOOS is a self-report questionnaire that can be administered by mail. The responsiveness4Terwee C.B. Bot S.D. de Boer M.R. van der Windt D.A. Knol D.L. Dekker J. et al.Quality criteria were proposed for measurement properties of health status questionnaires.J Clin Epidemiol. 2007; 60: 34-42Abstract Full Text Full Text PDF PubMed Scopus (6038) Google Scholar, 14Davies G.M. Watson D.J. Bellamy N. Comparison of the responsiveness and relative effect size of the western Ontario and McMaster Universities Osteoarthritis Index and the short-form Medical Outcomes Study Survey in a randomized, clinical trial of osteoarthritis patients.Arthritis Care Res. 1999; 12: 172-179Crossref PubMed Google Scholar of the SC-HOOS was tested by comparison of the preoperative and 3-month postoperative scores of the THR groups (N = 52; Student's t test). Standardized response mean (SRM), defined as the mean change between these time points, divided by the SD of this change, was calculated. Effect size (ES), defined as the mean change between preoperative results and 3-month postoperative results divided by the SD of the preoperative HOOS score, was also calculated. These data are summarized in Table I. The scores for all subscales improved postoperatively compared to preoperative values (P < 0.001). All of the subscales showed excellent responsiveness. The SRMs and ESs of the SC-HOOS subscales were similar to the SRM and ES of the HHS. Similar to other studies reporting high responsiveness of HOOS instruments10Nilsdotter A.K. Lohmander L.S. Klassbo M. Roos E.M. Hip disability and osteoarthritis outcome score (HOOS)-validity and responsiveness in total hip replacement.BMC Musculoskelet Disord. 2003; 30: 10Crossref Scopus (644) Google Scholar, 13Ornetti P. Parratte S. Gossec L. Tavernier C. Argenson J.N. Roos E.M. et al.Cross-cultural adaptation and validation of the French version of the hip disability and osteoarthritis outcome score (HOOS) in hip osteoarthritis patients.Osteoarthritis Cartilage. 2010; 18: 522-529Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar, 15Lee Y.K. Chung C.Y. Koo K.H. Lee K.M. Lee D.J. Lee S.C. et al.Transcultural adaptation and testing of psychometric properties of the Korean version of the Hip Disability and Osteoarthritis Outcome Score (HOOS).Osteoarthritis Cartilage. 2011; 19: 853-857Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar, we found that all of the subscales had high responsiveness in Chinese hip OA patients receiving THR. The responsiveness of the SC-HOOS was similar to that of the HHS in terms of SRM and ES values, strengthening our inference that the SC-HOOS has excellent responsiveness. We did not test the responsiveness of the SC-HOOS in patients undergoing non-surgical treatments due to the inadequate sample size following division of patients with respect to differing drug regimes, a lack of suitable controls, and the potential for a placebo effect13Ornetti P. Parratte S. Gossec L. Tavernier C. Argenson J.N. Roos E.M. et al.Cross-cultural adaptation and validation of the French version of the hip disability and osteoarthritis outcome score (HOOS) in hip osteoarthritis patients.Osteoarthritis Cartilage. 2010; 18: 522-529Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar. Several hip joint-specific, self-reported disability instruments, such as the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and Oxford hip score, have been developed for people with hip disabilities and are used extensively in research and clinical settings. However, the WOMAC is limited in that it only assesses pain, stiffness, and functional limitations. The HOOS includes more questions and subscales, and enables a more comprehensive evaluation, than the WOMAC. With the globalization of clinical research and the vigorous development of health-related quality of life research in China, there is an increasing need for valid questionnaires in Chinese that have solid psychometric properties but are not administratively burdensome. The HOOS, being a multidimensional and internationally used questionnaire designed to subjectively assess patients with hip disabilities is suitable for meeting this aim3Klassbo M. Larsson E. Mannevik E. Hip disability and osteoarthritis outcome score. An extension of the Western Ontario and McMaster Universities Osteoarthritis Index.Scand J Rheumatol. 2003; 32: 46-51Crossref PubMed Scopus (313) Google Scholar. The present results provide strong evidence that the SC-HOOS is similarly reliable and valid for evaluating pain and functional status in Chinese patients, similar to versions of the HOOS in French, Dutch, and Korean13Ornetti P. Parratte S. Gossec L. Tavernier C. Argenson J.N. Roos E.M. et al.Cross-cultural adaptation and validation of the French version of the hip disability and osteoarthritis outcome score (HOOS) in hip osteoarthritis patients.Osteoarthritis Cartilage. 2010; 18: 522-529Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar, 15Lee Y.K. Chung C.Y. Koo K.H. Lee K.M. Lee D.J. Lee S.C. et al.Transcultural adaptation and testing of psychometric properties of the Korean version of the Hip Disability and Osteoarthritis Outcome Score (HOOS).Osteoarthritis Cartilage. 2011; 19: 853-857Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar, 16de Groot I.B. Reijman M. Terwee C.B. Bierma-Zeinstra S.M. Favejee M. Roos E.M. et al.Validation of the Dutch version of the hip disability and osteoarthritis outcome score.Osteoarthritis Cartilage. 2007; 15: 104-109Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar. Our study has some limitations. Firstly, the sample was limited in size and may not fully represent the Chinese population. The validity of the SC-HOOS in areas beyond mainland China remains to be tested. Additionally, the responsiveness of the SC-HOOS was not assessed in patients receiving conservative treatments. Thus, it will be important to validate the SC-HOOS in a broader population and to probe the responsiveness of this instrument in hip OA patients receiving non-surgical treatments. In conclusion, the HOOS was successfully cross-culturally adapted into Simplified Chinese. The SC-HOOS was demonstrated to have good acceptability, internal consistency, test–retest reliability, construct validity, and responsiveness. This work will enable the HOOS to become widely used by physicians and researchers in mainland China.