Individualized treatment rules under stochastic treatment cost constraints

Abstract Estimation and evaluation of individualized treatment rules have been studied extensively, but real-world resource constraints received limited attention in existing methods. We investigate a setting which is intervened upon based on covariates to optimize the mean counterfactual outcome under cost when random. In particularly interesting special case, an instrumental variable corresponding encouragement with proportion receiving treatment. For such settings, we first develop method estimate optimal rules. further construct asymptotically efficient plug-in estimator average effect relative given reference rule.