The Uses of Propensity Scores in Randomized Controlled Trials

Abstract

Propensity scores are a dimension-reduction technique used to quantify the differences between treatment groups. Though propensity scores were developed to address the issue of confounding in observational studies, they have also proven useful in randomized controlled trials where confounding is structurally absent. When applied to randomized controlled trials, propensity scores can ensure balance between groups at the time of randomization, account for chance imbalances in observed randomization, and generalize target results to target populations. In this article, we review propensity score methodology developed for randomized trials with these goals.