Multi-level preventive care for Type 2 diabetes

Abstract

Type 2 Diabetes Mellitus (T2DM) accounts for 4.6 million deaths globally and for 11% of the global health expenditure (IDF, 2012). Several different primary, secondary, and tertiary preventive interventions promise better health outcomes and cost savings. Such interventions are typically studied in isolation. This paper proposes a compartmental mathematical model for T2DM that comprehends the interactions of multiple preventive interventions for various stages of T2DM, population dynamics, and the ensuing levels of clinical indicators, costs and utilities of disease states. We use the model to optimize portfolios of interventions for a multi-level preventive care program (using data from a population with high T2DM prevalence such as the UAE) and give insights about different ways in which interventions can be beneficial (such as for screening or for averting new cases). We demonstrate that the cost effectiveness with a classical discounted net present value perspective does not imply cost effectiveness for long-run planning, and that joint optimization of a portfolio of interventions can have benefits relative to the sequential optimization of interventions individually. Thus, accounting for long-run demographics and the interaction of interventions may be a useful extension to traditional cost-utility analyses when designing preventive care policies.