Robust Outcome Analysis for Observational Studies Designed Using Propensity Score Matching

• Activity Number: 526
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Bradley E Huitema*+ and Scott Kosten and Joseph W McKean
• Companies: Western Michigan University and Western Michigan University and Western Michigan University
• Keywords: Propensity Analysis ; ANCOVA ; Robust Estimation ; Observational Studies ; Outcome Analysis

Abstract

Estimating the treatment effect in an observational study is a challenging problem because there are likely to be differences between the treatment and control groups on a set of baseline covariates. If one or more of these baseline covariates is correlated with the response variable the difference in sample outcome means is likely to be a biased estimate of the true treatment effect. Propensity score matching can be used to redesign the study in order to provide meaningful comparison groups. After these comparison groups are formed a choice must be made for the outcome analysis. Our research focused on the development of a new procedure that combines propensity score matching and a rank-based analysis of the general linear model. Our fully adjusted procedure was compared with many other approaches, including those described in Hill and Reiter (2006), in a Monte Carlo simulation study. Overall, our procedure produced highly efficient and robust confidence intervals for the treatment effect. An application of the procedure to a large observational study is described.