Are sample sizes of randomized clinical trials in rheumatoid arthritis too large?

Abstract

We had the impression that randomized clinical trials (RCTs) frequently over enrolled patients. Thus, we surveyed power calculations in publications of RCTs of biologics in rheumatoid arthritis (RA) to assess over enrollment. A PubMed search identified 40 reports of original RCTs testing the efficacy of infliximab, etanercept, adalimumab, abatacept, rituximab and tocilizumab in patients with RA. As a first analysis, based on a two equal arms study with an alpha error of 0·05 and a power of 80% and of 90%, recalculation of the sample size was performed using the primary outcome results. In the second analysis, only those studies with equal number of patients in both arms and also in which all elements of a power calculation were given, were considered. New sample sizes were calculated based on the presented power elements in the related publications. In the first analysis, when we assigned a power of 80% and of 90%, 32 of 40 (80%) studies enrolled more than required number of patients, with a mean 131 ± 147 (SD), and 31 of 40 (78%) studies having had enrolled extra patients, with a mean 121 ± 147 (SD) in their treatment arms, respectively. Eleven studies qualified for the second analysis. There were still more patients with a mean of 48 ± 30 (SD) extra patients enrolled in the treatment arms. Most RCTs in RA enrol more patients than needed. This is costly and has the immediate consequence of exposing needless number of patients to potential harm.