Impacts of the Million Hearts® Cardiovascular Disease Risk Reduction Model on Provider‐Reported Preventive Care: A Pragmatic Randomized Trial

Research Objective The Million Hearts ® Cardiovascular Disease (CVD) Risk Reduction Model aims to reduce first‐time incidence of heart attacks and strokes among Medicare beneficiaries through improvements in CVD preventive care. The model pays providers (1) to assess each of their Medicare beneficiary’s risk of having a CVD event in 10 years and (2) to reduce risk among their high‐risk cohort. As risk stratification is central to clinical guidelines for CVD prevention, it is unclear whether the model’s modest incentives would change care beyond what practices would already provide. We assess whether the model has improved providers’ delivery of CVD preventive care. Study Design A survey of providers from organizations in a randomized trial. CMS enrolled 516 organizations (primary care and specialty practices, health centers, and hospitals) throughout the country and randomly assigned half to the intervention group. We surveyed randomly selected providers within each of 283 organizations participating after one and a half years of model implementation (45% of organizations from both groups either withdrew or did not enroll any beneficiaries, mainly due to burdensome reporting requirements). We estimated impacts as regression‐adjust differences in intervention and control providers’ responses to survey questions about CVD preventive care provided to their patients (survey response rate: 72%). Population Studied 261 providers in 245 organizations who enrolled beneficiaries and responded to the survey. Most respondents (77%) were physicians, 17% were nurse practitioners, and 6% were physician assistants. Despite high attrition rates, respondents in the intervention (N = 138) and control groups (N = 123) were well balanced on specialty and organization type. Principal Findings Relative to the control group, the model increased by 31% percentage points the share of providers reporting they risk stratify at least half of their Medicare beneficiaries (71% in the intervention group versus 39% in the control, P  < .001). Both intervention and control providers reported increasing their use of risk scores since the model began in 2017, but intervention providers reported substantially greater gains (25 to 71%) than control providers (31 to 39%). Most intervention group providers (73%) said that risk stratification helped them identify beneficiaries at risk of CVD events, and most (71%) reported that the model prompted them to more systematically apply the current standard of CVD care to their Medicare beneficiaries. Conclusions The Million Hearts® Model appears to have substantially improved CVD preventive care including enhanced identification and provider awareness of beneficiaries at risk for CVD events. Implications for Policy or Practice While clinical guidelines recommend using CVD risk stratification in primary CVD prevention, many of the participating organizations—which are likely a self‐selected group of higher performers—were not routinely risk stratifying patients at baseline. This trial shows that, among organizations that remained in the model 1.5 years after it started, commitment to the Million Hearts CVD preventive care model and modest financial incentives focused provider attention on CVD risk, leading to more routine use risk assessment, better detection of CVD risks in patient panels, and greater adherence to current standards of CVD preventive care. Together, these impacts have the potential to eventually reduce the risk of heart attacks and strokes. Primary Funding Source Centers for Medicare and Medicaid Services.