Problems persist in reporting of methods and results for the WOMAC measure in hip and knee osteoarthritis trials

PURPOSE: The Western Ontario and McMaster Universities Arthritis Index (WOMAC) is a commonly used outcome measure for osteoarthritis. There are different versions of the WOMAC (Likert, visual analogue or numeric scales). A previous review of trials published before 2010 found poor reporting and inconsistency in how the WOMAC was used. This review explores whether these problems persist. METHODS: This systematic review included randomised trials of hip and/or knee osteoarthritis published in 2016 that used the WOMAC. Data were extracted on the version used, score range, analysis and results of the WOMAC, and whether these details were clearly reported. RESULTS: This review included 62 trials and 41 reported the WOMAC total score. The version used and item range for the WOMAC total score were unclear in 44% (n = 18/41) and 24% (n = 10/41) of trials, respectively. The smallest total score range was 0–10 (calculated by averaging 24 items scored 0–10); the largest was 0–2400 (calculated by summing 24 items scored 0–100). All trials reported the statistical analysis methods but only 29% reported the between-group mean difference and 95% confidence interval. CONCLUSION: Details on the use and scoring of the WOMAC were often not reported. We recommend that trials report the version of the WOMAC and the score range used. The between-group treatment effect and corresponding confidence interval should be reported. If all the items of the WOMAC are collected, the total score and individual subscale scores should be presented. Better reporting would facilitate the interpretation, comparison and synthesis of the WOMAC score in trials. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s11136-018-1978-1) contains supplementary material, which is available to authorized users.