Clinical trialists generally accept the premise that surprising results in subgroups of a randomized, controlled trial most likely arise from chance. As we wrote nearly three decades ago, “We believe that the overall ‘average’ result of a randomized clinical trial is usually a more reliable estimate of treatment effect in the various subgroups examined than are the observed effects in individual subgroups.” When we wrote those words, trials involving patients with genetic markers or specific biologic targets were uncommon. Today, subgroups that are defined by such targets are, and we believe should be, analyzed with the view to making specific inferences about effects that may be unique to that group. Similarly, few trials in the 1980s were multinational; today many are. This article addresses the question of interpretation of observed variation in treatment effect among subgroups that are defined by country or geographic region. We do not address variability in observed effects according to the individual trial site, because in most multicenter trials, the sample size at each site is too small to expect reliable estimates of treatment effect.

Citado: Yusuf S, Wittes J. Interpreting Geographic Variations in Results of Randomized, Controlled Trials. N Engl J Med [Internet]. 2016 [citado 7 Nov 2017];375(23).