British Journal of Rheumatology 1991;30:326-329 METHODS OF ASSESSMENT USED IN ANKYLOSING SPONDYLITIS CLINICAL TRIALS: A REVIEW

Twenty non-steroidal anti-inflammatory drug (NSAID) trials in ankylosing spondylitis (AS) were reviewed to assess the frequency with which statistically significant differences had been detected between active drug and either a placebo or an NSAID-free washout period. Differences in pain severity were almost invariability detected, irrespective of the scale employed. In contrast, significant differences in axial movement were rarely detected in placebo controlled studies, and only about half of the variables detected significant improvement with respect to a washout period. From our data it is diffi-cult to differentiate whether the lack of difference with active therapy was due to inadequate sample size, non-responsive patients, or insensitive outcome measures. However, it is not surprising that between-drug differences are rarely detected in AS clinical trials of NSAIDs given our current inability to differentiate consistently an active treatment from a placebo and an active treatment phase from a washout period. KEY WORDS: Ankylosing spondylitis, Clinical trials, Non-steroidal anti-inflammatory drugs. THE importance of selecting appropriate outcome measures in musculoskeletal trials has been under-scored by several recent conferences and reviews [1], and is engendered in the guidelines of both inter-national agencies and regulatory bodies [2,3]