Dynamic Policy Modeling for Chronic Diseases: Metaheuristic-Based Identification of Pareto-Optimal Screening Strategies

We present a risk-group oriented chronic disease progression model embedded within a metaheuristic-based optimization of the policy variables. Policy-makers are provided with Pareto-optimal screening schedules for risk groups by considering cost and effectiveness outcomes as well as budget constraints. The quality of the screening technology depends on risk group, disease stage, and time. As the metaheuristic solution technique, we use the Pareto ant colony optimization (P-ACO) algorithm for multiobjective combinatorial optimization problems, which is based on the ant colony optimization paradigm. Our approach is illustrated by a numerical example for breast cancer. For a 10-year time horizon, we provide cost-effective screening schedules for selected annual and total budgets. We then discuss policy implications of 16 mammography screening scenarios varying the screening schedule (annual, biennial, triennial, quadrennial) and the rate of women tested (25%, 50%, 75%, 100%). Due to the model's flexible structure, interventions for multiple chronic diseases can be considered simultaneously.