Special Seminar

Robert Platt, PhD

Professor, Departments of Pediatrics and of Epidemiology, Biostatistics, and Occupational Health, McGill University

Targeted Learning in Pharmacoepidemiology

ALL ARE WELCOME

Abstract:

Targeted learning, or targeted maximum likelihood estimation (TMLE) is a framework for estimation that allows for specification and estimation of policy-relevant target parameters.  I will introduce TMLE, and discuss why targeted estimation in general, in which the statistical method is tailored to the parameter of interest rather than vice versa, should be used.  I will briefly outline TMLE for estimation of marginal parameters in pharmacoepidemiologic research with large administrative datasets.  I will demonstrate TMLE in a study of the use of statins and all-cause mortality, and via simulation.

Bio:

Robert Platt is Professor in the departments of Pediatrics and of Epidemiology, Biostatistics, and Occupational Health at McGill University.  He received a BSc in Mathematics from McGill, an MSc in Statistics from the University of Manitoba, and MS and PhD in Biostatistics from the University of Washington, and joined McGill in 1996.  Dr. Platt was director of the Biostatistics program from 2006-2012, and is currently the Director of Graduate Programs in the Department of Epidemiology, Biostatistics, and Occupational Health.  He holds a CIHR Foundation Scheme grant, and is co-PI and leader of the methods team for the Canadian Network for Observational Drug Effect Studies.  His research interests centre on causal inference in epidemiologic studies, with emphasis on pharmacoepidemiology and pediatric and perinatal epidemiology.