This study introduces Recurrent Individual Treatment Assignment (RITA), a novel approach to address heterogeneous treatment effects in longitudinal settings. Unlike traditional methods that rely on average treatment effects (ATEs) from randomized controlled trials (RCTs), RITA optimizes treatment assignment by focusing on individual-level treatment response. Through simulation, RITA demonstrates superior performance compared to RCT-based strategies, particularly when unobserved heterogeneity is present, by adapting individual assignments over time, balancing exploration and exploitation.

Ilja Cornelisz, Chris van Klaveren