The Use of Propensity Scores (PS) in Observational Research
View Presentation View Presentation
*Vinay Mehta, Merck 

Keywords: propensity scores

The increasing scrutiny by regulatory agencies on product safety has led to the emerging use of innovative methods to conduct observational research. Observational research is often required by regulatory authorities in the form of post-marketing safety studies.

A major challenge with observational research is that patients who may have a "sicker" profile are often channeled to one product over another, and comparator studies may be biased. PS is a method to minimize this "channeling" bias and balance the treatment groups, estimated through the probability of treatment conditional on a set of covariates. The main advantage of PS is that it allows for the control of large number of confounders by collapsing into a single variable. Balance between groups can be achieved through stratification or matching on PS, or analyses can adjust for PS in multivariate models.