C-learning: a New Classification Framework to Estimate Optimal Dynamic Treatment Regimes

Personalizing treatment to accommodate patient heterogeneity and the evolving nature of a disease over time has received considerable attention lately. A dynamic treatment regime is a set of decision rules, each corresponding to a decision point, that determine that next treatment based on each individual’s own available characteristics and treatment history up to that point. We show that identifying the optimal dynamic treatment regime can be recast as a sequential classification problem and is equivalent to sequentially minimizing a weighted expected misclassification error. This general classification perspective targets the exact goal of optimally individualizing treatments and is new and fundamentally different from existing methods. Based on this fresh classification perspective, we propose a novel, powerful and flexible C-learning algorithm to learn the optimal dynamic treatment regimes backward sequentially from the last stage till the first stage. C-learning is a direct optimization method that directly targets optimizing decision rules by exploiting powerful optimization/classification techniques and it allows incorporation of patient’s characteristics and treatment history to dramatically improves performance, hence enjoying the advantages of both the traditional outcome regression based methods (Q-and A- learning) and the more recent direct optimization methods. The superior performance and flexibility of the proposed methods are illustrated through extensive simulation studies