Propensity Score Methods for Estimating Treatment Effects Using Observational Data
*Peter Austin, Institute for Clinical Evaluative Sciences, and University of Toronto, Canada 

Keywords:

Propensity score methods are increasingly being used in health services and comparative effectiveness research to estimate the effects of treatments, interventions, and exposures on outcomes using observational or non-randomized data. We will begin by briefly reviewing the design and analysis of randomized controlled trials (RCTs). Participants will then be introduced to the concept of the propensity score and how it can be estimated using observational data. We will then examine how the propensity score can be used for matching, weighting, stratification, or covariate adjustment to estimate treatment effects. We will discuss how the first three of these methods allow one to mimic some of the characteristics of an RCT. We will also describe methods for assessing whether the propensity score model has been adequately specified using the observed data.