Identifying Optimally Cost-Effective and Interpretable Dynamic Treatment Regimes

• Activity Number: 247 - Causal Inference and Statistical Learning of Intervention and Policy Effects
• Type: Contributed
• Date/Time: Wednesday, August 11, 2021 : 10:00 AM to 11:50 AM
• Sponsor: Health Policy Statistics Section
• Author(s): Nicholas Illenberger* and Nandita Mitra and Andrew Justin Spieker
• Companies: University of Pennsylvania and University of Pennsylvania and Vanderbilt University Medical Center
• Keywords: cost-effectiveness; dynamic treatment regimes; health policy; Q-learning; time-varying confounding

Abstract

Health policy decisions regarding patient treatment strategies require consideration of both treatment effectiveness and cost. Optimizing treatment rules with respect to effectiveness may result in prohibitively expensive strategies; on the other hand, optimizing with respect to costs may result in poor patient outcomes. We propose a two-step approach for identifying an optimally cost-effective and interpretable dynamic treatment regime. First, we develop a combined Q-learning and policy search-based approach to estimate an optimal list-based regime under a constraint on expected treatment costs. Second, we propose a procedure to select an optimally cost-effective regime from a set of candidate regimes corresponding to different cost-constraints. Our approach can estimate optimal regimes in the presence of time-varying confounding and correlated outcomes, both of which are common in cost-effectiveness analyses. Through simulation studies, we examine the operating characteristics of estimated optimal treatment rules. We apply our approach to identify an optimally cost-effective regime for assigning adjuvant radiation and chemotherapy to endometrial cancer patients.