A statistic to avoid being misled by the “winners curse”

The “winners curse”, the term applied to the finding that RCTs, that report large effect sizes [e.g. relative risks of>5] worryingly often, when replicated are shown to be substantively exaggerated. This can result in undue adoption of a drug in clinical practice and from a methods perspective when replication is attempted, using these results to estimate sample size, subsequent RCTs will be underpowered to detect the true treatment effect. [1] in the past it has only been possible to show this retrospectively; Lord et  al in this issue now propose two statistics to predict this prospectively the type S Error Risk and the Exaggeration Ratio.
Source: Journal of Clinical Epidemiology - Category: Epidemiology Authors: Tags: Editorial Source Type: research