Abstract:
|
The vision for precision medicine is a healthcare system that adapts to the uniquely evolving health status of each individual patient. Such a system can deliver interventions only if, when, and in the amount they are needed thereby improving patient outcomes while reducing resource expenditures. Treatment regimes operationalize precision medicine as a sequence of decision rules, one per stage of intervention, that map up-to-date patient information to a recommended treatment. An optimal treatment regime maximizes the mean of some cumulative clinical outcome if applied to select treatments in a population of interest. In this course, we provide a rigorous but accessible introduction to estimation of and inference for optimal treatment regimes using data from observational or randomized studies. Topics covered include: formalizing an optimal treatment regime using potential outcomes; regression-based and policy-search estimators of an optimal treatment regime; design and analysis of sequential multiple assignment randomized trials; asymptotic methods for statistical inference; and software for constructing and evaluating estimated optimal treatment regimes. No prior exposure to treatment regimes is assumed
|