A systematic review of randomized controlled trials of pharmacological therapy in osteoarthritis of the hip.

Objective. To systematically review all randomized controlled trials (RCT) of pharmacological therapy in osteoarthritis (OA) of the hip. To determine which nonsteroidal antiinflammatory drug (NSAID) is the most effective, and which NSAID is the most toxic. Methods. A MEDLINE search was used to identify RCT of pharmacological therapy in patients with OA of the hip published between 1966 and August 1994. Qualitative assessments were performed using a quality scoring system designed for NSAID trials in rheumatoid arthritis. Both the design and analysis aspects of the trials were evaluated, each aspect rated on a scale of 0 to 8. A quantitative method, which calculates the ratio of improvement produced by one NSAID to that produced by another, was used to rate the relative efficacy of different NSAID with respect to pain relief. Toxicity comparisons were made according to the authors' findings. Results. 43 RCT were identified, and of these, 39 evaluated NSAID while 4 evaluated only analgesics. The median design and analysis scores were 2 and 4, respectively. 6 NSAID were included in at least 5 trials; of these, indomethacin was rated more effective in 5 of its 7 comparisons, but more toxic in 7 of 12 comparisons. Only 5 of the 29 (17%) NSAID comparisons found statistically significant differences in efficacy. Conclusion. NSAID trials in patients with OA of the hip appear to be weakened by the lack of standardization of case definition of OA, and also by the lack of standardization of outcome assessments. No recommendations for the choice of specific NSAID therapy in hip OA can be offered based on this analysis.