Estimating the optimal individualized treatment rule from a cost-effectiveness perspective

Area of Science:

• Health Economics
• Biostatistics
• Precision Medicine

Background:

• Individualized treatment rules (ITRs) aim to maximize clinical benefit using patient characteristics.
• Current ITR research often overlooks the economic costs of interventions.
• Health policy requires balancing treatment effectiveness with resource allocation and costs.

Purpose of the Study:

• To develop and evaluate a method for estimating optimal, cost-effective individualized treatment rules (ITRs).
• To integrate both treatment effectiveness and economic costs into decision-making for personalized medicine.
• To propose a statistical learning algorithm for nonparametric estimation of ITRs under a composite outcome.

Main Methods:

• Proposed a decision-tree-based statistical learning algorithm.
• Utilized a net-monetary-benefit-based reward for direct optimization of ITRs.
• Conducted simulation studies to compare different reward estimation approaches.

Main Results:

• The proposed algorithm provides nonparametric estimations of optimal ITRs.
• Simulation studies identified the strengths and weaknesses of various reward estimation methods.
• The top-performing method was applied to the Systolic Blood Pressure Intervention Trial (SPRINT) data.

Conclusions:

• The study presents a novel approach to identify cost-effective ITRs by directly optimizing net monetary benefit.
• The decision-tree algorithm effectively accounts for individual heterogeneity in treatment decisions.
• The findings have implications for resource allocation and personalized healthcare policy.