Clinical trial design for evaluating combination therapies.

Because the differences in efficacy between a disease-modifying antirheumatic drug (DMARD) combination and its constituent drugs are likely to be smaller than those between placebo and the single DMARDs, it has been difficult to prove that any combination of DMARDs has either additive or synergistic benefit. The discriminatory power of the study design can be enhanced by careful attention to the details of patient selection, study design and duration, and choices of primary outcome measures and analytical methods. Use of the American College of Rheumatology (ACR) 'core set' of outcome measures and the proposed ACR 'definition of improvement' for rheumatoid arthritis (RA) clinical trials, with intent-to-treat analysis, in a balanced, prospective, double-blind randomized clinical trial of 1-2 yr duration will optimize the chances of discriminating between the clinical benefit of the combination and its components if a real difference is present.