This presentation is part of: Poster Session - Public Health; Methodological Advances
COST-EFFECTIVENESS OF TYPE 2 DIABETES PREVENTION IN THE GENERAL POPULATION IN A US SETTING
Michael E. Minshall, MD, MPH1, Andrew J. Palmer, BSc, MBBS2, Stephane Roze, MS3, and William J. Valentine, PhD2. (1) CORE-USA, Center for Outcomes Research USA, Indianapolis, IN, (2) CORE Center for Outcomes Research, Health Economics and Modeling, Binningen, Switzerland, (3) CORE Center for Outcomes Research, Department of Statistics, Binningen, Switzerland
PURPOSE: An epidemic increase in the prevalence of type 2 diabetes (T2D) is predicted as the population ages, and western lifestyles become increasingly unhealthy. A number of recent clinical studies have demonstrated that the onset of T2D is delayed or prevented in high risk groups by lifestyle changes and/or medications. T2D patients have more than double the mortality rates and higher annual treatment costs of equally-matched non-diabetes patients. A simulation model was developed to assess acceptable cost limits for a broad-based general population-targeted program aimed at reducing the incidence of T2D by (mean±standard deviation) 30±10%. METHODS: A Markov/second order Monte Carlo model used age- and gender-specific incidence of- and increased direct medical costs and mortality associated with- T2D in a US setting to forecast life expectancy (LE) and costs with or without intervention. Data were derived from published sources. The maximum costs/person a payer could outlay to achieve a 30±10%.reduction in T2D incidence a) without increasing the healthcare budget, and b) remaining within an attractive incremental cost-effectiveness ratio (ICER) limit (i.e. <$50,000/life year gained versus no intervention). Costs and LE were discounted at 3% annually. A health insurance perspective was taken. Sensitivity analysis was performed to identify parameters with important impacts on outcomes. RESULTS: A hypothetical diabetes prevention intervention aimed at a general population with mean age 51 years would improve undiscounted LE by 0.33±0.16 years/person (0.14±0.07 discounted 3% p.a.). Up to a cost of $68/year/person, the program be cost saving overall. The ICER versus no intervention would be <$50,000 at a cost of $520/person/year. Sensitivity analysis revealed that the age of the target population, effectiveness of the program, incidence of diabetes, and increase in mortality associated with diabetes have large impacts on the results. CONCLUSIONS: Diabetes prevention programs aimed at a general population could be cost saving or cost-effective if the costs of the program do not exceed the limits identified in this analysis. In higher-risk populations where the incidence of diabetes and effects of intervention are greater, these cost limits may be higher.
