An Analytic Framework for Effective Public Health Program Design Using Correctional Facilities

Abstract

We present a decision analytic framework that uses a mathematical model of Chlamydia trachomatis transmission dynamics in two interacting populations using ordinary differential equations. A public health survey informs model parametrization, and analytical findings guide the computational design of the decision-making process. The potential impact of jail-based screen-treat (S-T) programs on community health outcomes is presented. Numerical experiments are conducted for a case study population to quantify the effect and evaluate the cost-effectiveness of considered interventions. Numerical experiments show the effectiveness of increased jail S-T rates on community cases when resources for a community S-T program stays constant. Although this effect decreases when higher S-T rates are in place, jail-based S-T programs are cost-effective relative to community-based programs. Summary of Contribution: Public health programs have been developed to control community-wide infectious diseases and to reduce prevalence of sexually transmitted diseases (STD). These programs can consist of screening and treatment of diseases and behavioral interventions. Public correctional facilities play an important role in operational execution of these public health programs. However, because of lack of capacity and resources, public health programs using correctional facilities are questioned by policy-makers in terms of their costs and benefits. In this article, we present an analytical framework using a computational epidemiology model for supporting public health policy making. The system represents the dynamics of Chlamydia trachomatis transmission in two interacting populations, with an ordinary differential equations-based simulation model. The theoretical epidemic control conditions are derived and numerically tested, which guide the design of simulation experiments. Then cost-effectiveness of the potential policies is analyzed. We also present an extensive sensitivity analyses on model parameters. This study contributes to the computational epidemiology literature by presenting an analytical framework to guide effective simulation experimentation for policy decision making. The presented methodology can be applied to other complex policy and public health problems.