Loading...
Evidence suggests that smaller randomized controlled trials (RCTs) yield greater treatment-effect estimates than do larger RCTs. In this study, investigators determined the effects of sample size on treatment-effect estimates.
Researchers identified 93 meta-analyses (that included 735 RCTs) of therapeutic interventions with binary outcomes. RTCs within each meta-analysis were sorted by sample size (i.e., quarter 1: the 25% of trials with the smallest samples, through quarter 4: the 25% of trials with the largest samples). Compared with effect estimates in quarter 4, effect estimates in quarter 3 averaged 12% larger, effect estimates in quarter 2 were 17% larger, and effect estimates in quarter 1 were 32% larger. Individual trials also were s…