Cross-cultural adaptation and validation of the French version of the Hip disability and Osteoarthritis Outcome Score (HOOS) in hip osteoarthritis patients.

OBJECTIVE: To translate and adapt the Hip disability and Osteoarthritis Outcome Score (HOOS) into French and to evaluate the psychometric properties of this new version, by testing feasibility, internal consistency, construct validity, reliability and responsiveness, in patients with hip osteoarthritis (OA).

METHODS: The French version of the HOOS was developed according to published international guidelines to ensure content validity. The new version was then evaluated in two symptomatic hip OA populations, one with no indication for joint replacement (medical group), and the other waiting for total hip replacement (THR) (surgery group). The psychometric properties assessed were feasibility: percentage of responses, floor and ceiling effects; internal consistency using Cronbach's alpha; construct validity by evaluating correlations with the Lequesne's index and the visual analogic scale (VAS) for pain (Pearson's correlation coefficient); reliability: intra-class correlation coefficient (ICC), Bland et Altman representation; responsiveness by comparing the results of before and 1 month after injection of hyaluronic acid (medical group) and by comparing the pre and 3 months post THR results (surgery group) by calculating standardized response mean (SRM) and effect size (ES).

RESULTS: A total of 88 patients were recruited; 58 in the medicine group (mean age=61.8+/-9 years, range 42-81, 70% women) and 30 in the surgery group (mean age=67.5+/-9 years, range 50-81, 68% women). The percentage of item responses was excellent (99%). Neither a floor nor a ceiling effect was observed, except for a ceiling effect (17.8% of patients with worst possible score) observed prior to surgery in the sport and recreation subscale. The internal consistency was good for four of the five HOOS subscales. As expected, the strongest correlations were observed between all HOOS subscales and Lequesne's index or VAS pain, indicating good construct validity. The reliability was good, with an ICC>0.8 for all subscales. The responsiveness was good for all domains 1 month after hyaluronic acid injection (ES ranging from 0.73 to 1.86 and SRM from 0.51 to 1.04) and high for all domains 3 months after THR (ES ranging from 1.47 to 2.08 and SRM ranging from 1.97 to 3.24).

CONCLUSION: The French version of HOOS demonstrated good psychometric properties and appears to be useful for the evaluation of patient-relevant outcome whatever the severity of hip OA. This study provides a basis for the use of this French version of the HOOS in future clinical trials.