Using microsimulation to inform targeted cardiovascular disease prevention policy [presentation] Presentation uri icon

abstract

  • Background/Aims: The U.S. Preventive Services Task Force (USPSTF) makes four recommendations for highly effective clinical services in CVD prevention: 1) screening for hypertension, 2) screening for high cholesterol, 3) aspirin counseling, and 4) screening for diabetes in asymptomatic adults with hypertension. This study asks which prevention activities are the most cost-effective and which can prevent the greatest amount of CVD health burden at the margin. Moreover, to whom can preventive services be targeted to realize the greatest impacts at the lowest costs? Amidst the increasing push towards individualized medicine, this study adds valuable economic information to help guide health policy. Methods: This study introduces the Cardiovascular Disease Prevention Policy Model (CVD-PPM) as a tool for evaluating preventive health interventions. The CVD-PPM is implemented as a Markov-based microsimulation model, developed in Microsoft Excel using Visual Basic for Applications. The underlying disease model allows for cardiovascular risk factors—such as BMI, cholesterol, and blood pressure—to transition according to individual characteristics during the course of a simulated person’s lifetime. Risk of CVD-related events or death is estimated using customized risk equations derived from Framingham Heart Study data. Acute and ongoing disease costs are estimated using the Medical Expenditure Panel Survey (MEPS). Health policy or specific clinical interventions can be evaluated by differences in costs, event rates, life expectancy, and quality-adjusted life years (QALYs) and compared across age groups, race/ethnicities, and between sexes. Results: This analysis finds aspirin to be the most cost-effective clinical service in the prevention of CVD (cost-saving, $110 per person) and lipid screening has the most potential health impact: 850,000 QALYs saved on a simulated 4 million person birth cohort. Hypertension is found to be the least cost-effective ($60,000 per QALY saved), and diabetes screening is found to have the smallest health impact (40,000 QALYs saved). Analyses of sub-population groups reveal more nuanced results. Conclusions: By and large, clinical prevention strategies to reduce CVD are cost-effective and can have substantial health impact. However, evidence-based micro-level simulation models can better inform targeted prevention strategies to reach underserved and highly cost-effective population groups.