Medical therapy often consists of multiple stages, with a treatment chosen by the physician at each stage based on the patient’s history of treatments and clinical outcomes. These decisions can be formalized as a dynamic treatment regime. This article describes a new approach for optimizing dynamic treatment regimes, which bridges the gap between Bayesian inference and existing approaches, like Q-learning. The proposed approach fits a series of Bayesian regression models, one for each stage, in reverse sequential order. Each model uses as a response variable the remaining payoff assuming optimal actions are taken at subsequent stages, and as covariates the current history and relevant actions at that stage. The key difficulty is that the optimal decision rules at subsequent stages are unknown, and even if these decision rules were known the relevant response variables may be counterfactual. However, posterior distributions can be derived from the previously fitted regression models for the optimal decision rules and the counterfactual response variables under a particular set of rules. The proposed approach averages over these posterior distributions when fitting each regression model. An efficient sampling algorithm for estimation is presented, along with simulation studies that compare the proposed approach with Q-learning. Supplementary materials for this article are available online.
Bibliographical noteFunding Information:
Yuan’s and Murray’s research was partially supported by Award Number R01-CA154591 from the National Cancer Institute. The work of the first three authors was partially funded by NIH/NCI grant 5-R01-CA083932.
- Approximate dynamic programming
- Backward induction
- Bayesian additive regression trees
- Gibbs sampling
- Potential outcomes