Using SAS® Software to Perform a Case-Control Match on Propensity Score in an Observational Study

In large observational studies there are often significant differences between characteristics of a treatment group and a no treatment group. Such differences should not exist in a randomized trial. These differences must be adjusted for in order to reduce treatment selection bias and determine treatment effect. There are several methods to reduce the bias of these differences and make the two groups more similar. One method is to perform analyses after matching cases (members of the treatment group) to controls (members of the no treatment group) based on a number of individual characteristics. A refinement of this method is to create a propensity score to represent the relationship between multiple characteristics and an outcome as a single score, and then match on that single score. This paper will show SAS users how to create a propensity score using the LOGISTIC procedure and then match cases to controls based on this score with a user-written SAS macro program. The results of using the presented code, run on a large observational database of myocardial infarction patients, will be given as an example.