American Statistical Association
In this talk, I will discuss recent developments in the estimation of an optimal treatment strategy or regime from longitudinal data collected in an observational study. Estimation of the optimal time for an asymptomatic HIV infected subject to start highly active retroviral therapy (HAART) will serve as a paradigmatic example. I review assumptions under which it is possible to use observational data to estimate the optimal treatment regime in a class of logistically feasible dynamic regimes and describe an analytic approach based on dynamic marginal structural models (MSMs) that recovers the optimal regime in the class. I compare the MSM approach with an alternative approach based on g-estimation of optimal regime structural nested models (SNMMs). G-estimation of optimal regime structural nested models (SNMMs) is a statistically robust twist on the classic method of dynamic programming (backward induction) for sequential decision making under uncertainty.
|Date:||Wednesday, February 6, 2008|
|Time:||4:00 P.M. - 5:00 P.M.|
Memorial Sloan-Kettering Cancer Center
Department of Epidemiology and Biostatistics
307 East 63rd Street
(between First and Second Avenues)
3rd Floor Conference Room
New York, New York
Note: To gain access to the building, please follow the directions by the telephone in the foyer.