Estimating Optimal Dynamic Treatment Regimes with Survival Outcomes: An Application to the Treatment of Type 2 Diabetes

• Activity Number: 543 - Making Sense of Complex Featured Data with Statistical Methods
• Type: Invited
• Date/Time: Wednesday, July 31, 2019 : 2:00 PM to 3:50 PM
• Sponsor: SSC
• Author(s): Gabrielle Simoneau* and Erica Moodie and Robert Platt and Laurent Azoulay
• Companies: McGill University and McGill university and McGill University and McGill University
• Keywords: accelerated failure time; causal inference; censored data; large observational data; precision medicine; type 2 diabetes

Abstract

The statistical study of precision medicine is concerned with dynamic treatment regimes (DTRs) in which treatment decisions are tailored to evolving patient-level information. Of interest is to learn about an optimal DTR, that is, the sequence of treatment decisions that yields the best outcome. Statistical methods for identifying optimal DTRs from observational data are theoretically complex and hardly accessible to researchers, especially when the outcome is a survival time subject to right censoring. We propose a doubly-robust method, called dynamic weighted survival modeling, for estimating optimal DTRs for such endpoints. We give an overview of the method and illustrate its usefulness in an application to the treatment of type 2 diabetes using a large observational database. This talk was co-sponsored by SSC and CANSSI.