A tutorial on optimal dynamic treatment regimes

TL;DR

This tutorial provides a comprehensive introduction to optimal dynamic treatment regimes, covering their formal definition, causal inference framework, estimation methods, and applications in precision medicine.

Contribution

It offers a systematic, detailed, and accessible overview of the methodology and applications of optimal dynamic treatment regimes, including formal definitions and estimation techniques.

Findings

Overview of causal inference framework for treatment regimes

Description of statistical models and estimation methods

Application to simulated and real data

Abstract

A dynamic treatment regime is a sequence of treatment decision rules tailored to an individual's evolving status over time. In precision medicine, much focus has been placed on finding an optimal dynamic treatment regime which, if followed by everyone in the population, would yield the best outcome on average; and extensive investigation has been conducted from both methodological and applications standpoints. The aim of this tutorial is to provide readers who are interested in optimal dynamic treatment regimes with a systematic, detailed but accessible introduction, including the formal definition and formulation of this topic within the framework of causal inference, identification assumptions required to link the causal quantity of interest to the observed data, existing statistical models and estimation methods to learn the optimal regime from data, and application of these methods to both simulated and real data.