Dalla Lana School of Public Health

The Biostatistics Seminar Series presents:

“Propensity Score Analysis, Challenges, and Practical Recommendations: Evaluation through Simulations and Case Studies” by Dr. Sareh Keshavarzi, Princess Margaret Cancer Center

Abstract: Observational studies are sometimes preferred over randomized trials because of having lower cost, greater timeliness, and more generalizability. The issue with the observational studies is that the causal inference cannot be simply made due to potential confounding effects. Propensity scores (PS) analysis is a well-known approach that can be used to reduce the confounders’ effect on the treatments, exposures, or intervention estimates in observational studies. Few guidelines are available regarding the choice of PS approaches for the best performance in a given set of data. We conducted extensive scenarios of Monte Carlo simulations comparing the performance of conventional covariate adjustment and eight other common PS approaches to estimate the average treatment effect for the overall population on time-to-event outcomes. We also implemented the aforementioned PS approaches to compare the effect of receiving postoperative radiation therapy and having an engraftable tumor on different time-to-event clinical outcomes in Oral Squamous Cell Carcinoma Cancer patients. There are some challenges in PS analysis such as the problem of confounder selection and handling partially observed confounders by multiple imputation (MI). It is not clear how to best apply MI-based strategies and how to select the best list of confounders in the context of PS analysis. Currently we are exploring the differences in the effect estimates for the PS methods using different confounder selection methods and various MI-based strategies using a case study on non-alcoholic fatty liver disease patients. I will be presenting some preliminary results on that and provide some practical recommendations.

For Dr. Keshavarzi’s biosketch, please see https://www.dlsph.utoronto.ca/faculty-profile/keshavarzi-sareh/

Register in advance for this seminar via
https://phesc.zoom.us/meeting/register/tZMpcu2uqzMoH9IIU0yOH4Gnwv4wXZpGKc8B