An introduction to propensity score matching methods

Abstract

A propensity score is the probability that an individual will be assigned to a condition or group, given a set of covariates when the assignment is made. For example, the type of drug treatment given to a patient in a real-world setting may be non-randomly based on the patient's age, gender, geographic location, overall health, and/or socioeconomic status when the drug is prescribed. Propensity scores are used in observational studies to reduce selection bias by matching different groups based on these propensity score probabilities, rather than matching patients on the values of the individual covariates. Although the underlying statistical theory behind propensity score matching is complex, implementing propensity score matching with SAS ® is relatively straightforward. An output data set of each patient's propensity score can be generated with SAS using PROC LOGISTIC, and a generalized SAS macro can do optimized N:1 propensity score matching of patients assigned to different groups. This paper gives the general PROC LOGISTIC syntax to generate propensity scores, and provides the SAS macro for optimized propensity score matching. A published example of the effect of comparing unmatched and propensity score matched patient groups using the SAS programming techniques described in this paper is presented.