Methods for assessing and controlling placebo effects

Abstract

The placebo serves as an indispensable control in many randomized trials. When analyzing the benefit of a new treatment, researchers are often confronted with large placebo effects that diminish the treatment effect. Various alternative methods have been proposed for analyzing placebo and treatment effects in studies where large placebo effects are expected or have already occurred. This paper presents an overview of methodological work that has been proposed for assessing and/or controlling for placebo effects in randomized trials. Throughout this paper, two main approaches are discussed. The first approach considers designs that represent alternatives to the classical placebo-controlled randomized trial design. Separately, the second approach considers adopting new methods for the statistical analysis of placebo and treatment effects to be implemented after the data have been collected using a classical randomized trial design.