Impacts of the Million Hearts® Cardiovascular Disease Risk Reduction Model on Medications, Heart Attacks and Strokes, and Medicare Spending after Two Years: A Cluster‐Randomized Trial

Abstract

Research ObjectiveIn 2017, the Centers for Medicare & Medicaid Services launched the Million Hearts® Cardiovascular Disease (CVD) Risk Reduction Model with the aim to reduce first‐time heart attacks and strokes among high‐risk Medicare fee‐for‐services (FFS) beneficiaries within 5 years. This large‐scale randomized trial pays providers to assess each of their Medicare patient’s risk of a heart attack or stroke and then makes additional payment if they reduce CVD risk among their high‐risk cohort. We assess whether, during its first two years, the model increased use of CVD medications, reduced incidence of heart attacks and strokes, and reduced Medicare spending.Study DesignA pragmatic, cluster‐randomized trial. CMS enrolled 516 organizations (primary and specialty practices, health centers, and hospitals) throughout the country and assigned half to the intervention group. For the 313 organizations that enrolled beneficiaries in 2017, we linked enrollment and clinical data with Medicare claims and estimated impacts as intervention‐control differences (regression adjusted) in outcomes.Population Studied300,550 Medicare FFS beneficiaries, 18% of whom were high risk (≥30% or higher risk of a heart attack or stroke in the next 10 years) and 40% were medium risk (15‐30% risk) in both groups. Intervention and control group beneficiaries were similar at enrollment in demographics, CVD risk factors, and recent utilization and spending.Principal FindingsAmong the high‐risk group, almost 40% of CVD risk at baseline was modifiable, mainly from elevated blood pressure and cholesterol. High‐risk intervention group enrollees were 4 percentage points more likely than the control group (28 vs 24%, P < .001) to initiate or intensify statins or antihypertensive medications factors within 6 months of enrollment. When including the larger medium‐risk group—for whom CMS does not separately pay for CVD risk reduction—rates of initiation or intensification were 3 percentage points higher in the intervention versus control groups (23 vs 20%, P < .001). Over a mean follow‐up of 17 months, the incidence of first‐time heart attack and stroke was similar in the intervention versus the control groups for the high‐risk group (90% confidence interval for hazard ratio: 0.93‐1.14) and for the medium‐ and high‐risk groups combined (90% CI: 0.93, 1.06). Medicare Part A and B spending was also similar in the intervention and control groups.ConclusionsThus far, the Million Hearts model has improved initiation and intensification of medications to treat modifiable risk factors among the high‐risk group, with positive spillover to the much larger medium‐risk group. However, the model has not reduced heart attacks and strokes or generated any savings to offset model costs.Implications for Policy or PracticeWhile CVD risk factors have declined substantially over the past 50 years, progress has slowed recently and even reversed in some groups. Value‐based purchasing efforts to date have focused on achieving control for individual CVD risk factors. This trial shows that a new “pay for prevention” approach that rewards assessing and reducing total CVD risk (combining individual risk factors into a single score readily interpretable by patients and providers) can improve use of CVD medications. More time is needed to assess whether the model will eventually reduce incidence of heart attacks and strokes.Primary Funding SourceThe study was funded by the Centers for Medicare and Medicaid Services.