Evaluation of the Million Hearts CVD Risk Reduction Model (MH)

The Million Hearts Cardiovascular Disease (CVD) Risk Reduction Model, run by the Centers for Medicare & Medicaid Services (CMS), seeks to improve cardiovascular care by providing incentives and supports for health care practitioners to engage in patient CVD risk calculation and population-level CVD risk management. CMS enrolled organizations throughout the United States, randomly assigning half to the intervention and half to a control group. This study is an evaluation of the model and will assess the model impacts on patient outcomes, changes in CVD care processes, and implementation challenges and successes.

Study Overview

• Status: Unknown
• Conditions: Cardiovascular Diseases
• Intervention / Treatment: Other: Enrollment in the Million Hearts CVD Risk Reduction Model; Other: Enrollment in control provider organizations

Detailed Description

In January 2017, the Centers for Medicare & Medicaid Services (CMS) launched the Million Hearts Cardiovascular Disease (CVD) Risk Reduction model, designed to reduce heart attacks and strokes among Medicare fee for-service (FFS) beneficiaries. CMS is testing the Million Hearts CVD model over five years among more than 400 participating organizations, with half randomly assigned to the intervention and half to a control group. These organizations include primary care practices, specialty/multispecialty practices, health centers, and hospital outpatient departments. The intervention organizations are expected to:

• Risk stratify all of their eligible Medicare FFS beneficiaries, using the American College of Cardiology/American Heart Association (ACC/AHA) calculator to estimate each eligible beneficiary's risk of having a heart attack or stroke over the next 10 years. Beneficiaries are eligible if they are ages 40-79 as of enrollment in the program, have not had a heart attack or stroke, are enrolled in Medicare Part A and B, do not have end-stage renal disease, and are not receiving hospice benefits. Beneficiaries with a CVD risk exceeding 30 percent are considered high risk, whereas those with a risk from 15-30 percent are medium risk. All others are low risk.
• Provide cardiovascular care management to high-risk beneficiaries--which includes discussing with patients different options for reducing CVD risk, developing a care plan, and following up with patients at least twice a year (any mode) to assess and encourage progress on the care plan, and annual in-person visits to reassess risk and revise care plans.
• Collect and report clinical data to CMS via the Million Hearts Model Data Registry.
• Participate in learning system activities, including webinars and videoconferences, designed to spread effective strategies for implementing the model.

CMS supports the intervention organizations with payments for risk stratification, cardiovascular management, and risk reduction. Participating organizations receive payments for each eligible beneficiary they risk stratify. In the first model year, the cardiovascular management fees are fixed per beneficiary per month (PBPM) for each high-risk enrollee. In model year 2 and later, CMS is replacing the cardiovascular management fees with risk reduction payments that are scaled to the organization's performance in reducing 10-year predicted risk among their beneficiaries who were high-risk at initial enrollment. To support the model's evaluation, CMS is also paying control organizations to collect and report clinical data on their eligible Medicare FFS beneficiaries, but these organizations are not asked to calculate CVD risk scores or otherwise change their clinical care.

The primary goal of the investigators is to evaluate the impact of this model on first-time heart attacks and strokes and CVD-related spending among high CVD risk Medicare FFS beneficiaries, comparing beneficiaries in intervention practices with beneficiaries in control practices. Using Medicare Part A, B, and D administrative claims data, investigators also plan to evaluate the impact of the model on beneficiary mortality, CVD service utilization, and CVD-related medication use. The investigators will also use survey data from providers to identify changes in CVD-related knowledge, behaviors, and care delivery.

• Study Type: Observational
• Enrollment (Anticipated): 210000

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 40 years to 79 years (ADULT, OLDER_ADULT)
• Accepts Healthy Volunteers: Yes
• Genders Eligible for Study: All

• Sampling Method: Probability Sample

Study Population

The study population is a census of Medicare fee-for-service beneficiaries who were enrolled by providers (physicians, nurse practitioners, or physician assistants) in the intervention and control organizations. Providers enroll a beneficiary when they have an office visit with a patient and submit to CMS all of the CVD risk data needed to generate a risk score that is current as of that visit date. Providers submit these data to CMS via a registry-called the Million Hearts Model Data Registry-designed specifically for this purpose. Beneficiaries will be included in the study population if they meet the eligibility criteria. Following an intent-to-treat principle, we will keep beneficiaries in the sample for as long as they are alive and observable in Medicare FFS claims data (the primary source for study outcomes), whether or not they continue to receive clinical services from the participating organizations.

Description

Inclusion Criteria:

• Ages 40-79 as of enrollment in the program
• Enrolled in Medicare Part A and B
• Some analyses will include only those with a 10-year predicted CVD risk exceeding 30 percent (high risk). Others will include those with a 10-year predicted CVD risk exceeding 15 percent (high and medium risk).
• Some secondary analyses related to medication use will also be restricted to those enrolled in Medicare Part D

Exclusion Criteria:

• Have had a heart or stroke previously
• Have end-stage renal disease
• Currently enrolled in hospice care

Study Plan

How is the study designed?

Number of groups / cohorts

2

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
Enrollment in the Million Hearts CVD Risk Reduction Model; Eligible Medicare fee-for-service (FFS) beneficiaries enrolled in provider organizations that were randomly assigned to participate in the Million Hearts CVD Risk Reduction Model intervention.	Other: Enrollment in the Million Hearts CVD Risk Reduction Model; The intervention is at a provider organization level and all eligible beneficiaries enrolled by participants will be considered exposed. Intervention organizations receive payments to: Risk stratify eligible Medicare FFS beneficiaries; Provide CVD care management to high-risk beneficiaries; Collect and report clinical data to CMS via the Million Hearts Data Registry and participate in learning system activities Participating organizations receive payments for each eligible beneficiary they risk stratify. In model year 1, the organizations receive a fixed payment per beneficiary per month to provide cardiovascular management. In model years 2-5, the organizations receive a risk reduction payment that is scaled based on reductions in 10-year predicted risk scores among their cohort of high-risk beneficiaries.
Enrollment in control provider organizations; Eligible Medicare fee-for-service (FFS) beneficiaries enrolled in provider organizations that were randomly assigned to the control group.	Other: Enrollment in control provider organizations; Control organizations receive payments to collect and report clinical data on their eligible Medicare FFS beneficiaries, but are not asked to calculate CVD risk scores or otherwise change their clinical care.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Incidence of heart attacks and strokes among high risk	Incidence of first-time heart attack and stroke among high CVD risk enrollees	5 years
Incidence of heart attacks and strokes among high and medium risk	Incidence of first-time heart attack and stroke among high and medium CVD risk enrollees	5 years
CVD-related spending among high risk	Medicare Part A and B spending (in dollars per person per quarter) for (1) heart attack/stroke hospitalizations and related post-acute care, and (2) heart attack/stroke emergency department visits among high CVD risk enrollees	5 years
CVD-related spending among high and medium risk	Medicare Part A and B spending (in dollars per person per quarter) for (1) heart attack/stroke hospitalizations and related post-acute care, and (2) heart attack/stroke emergency department visits among high CVD risk enrollees	5 years

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Rate of all-cause mortality	Mortality rate from any cause, analyzed separately for just high CVD risk enrollees and for high and medium CVD risk enrollees combined	5 years
Change in 10-year predicted CVD risk	Change in 10-year predicted risk of hearts attack or stroke between baseline and reassessment visits, with predicted risk estimated in percentage points using the Million Hearts Longitudinal Atherosclerotic CVD Risk Assessment Tool. Predicted CVD risk ranges from 0 to 100 percent and larger predicted CVD risk represent worse outcomes. Measure will be analyzed separately for just high CVD risk enrollees and for high and medium CVD risk enrollees combined.	3 years
Spending, without model payments	Medicare Part A and B spending (in dollars per person per quarter, without including additional payments associated with the Million Hearts CVD Risk Reduction Model), analyzed separately for just high CVD risk enrollees and for high and medium CVD risk enrollees combined	5 years
Spending, with model payments	Medicare Part A and B spending and additional payments (in dollars per person per quarter) associated with the Million Hearts CVD Risk Reduction Model, among high and medium CVD risk enrollees	5 years
Number of CVD-related hospitalizations	Hospitalizations for heart attack, stroke, and other cardiovascular disease (in number per 1,000 people per quarter), analyzed separately for just high CVD risk enrollees and for high and medium CVD risk enrollees combined	5 year
Number of CVD-related emergency department visits	Number of outpatient emergency department visits for heart attack, stroke, and other cardiovascular disease (in number per 1,000 people per quarter), analyzed separately for just high CVD risk enrollees and for high and medium CVD risk enrollees combined	5 years
Number of Million Hearts office visits	Number of office visits with a Million Hearts-participating provider (in number per 1,000 people per quarter), analyzed separately for just high CVD risk enrollees and for high and medium CVD risk enrollees combined	5 years
Percent of eligible beneficiaries using statins	Percent of beneficiaries with Medicare Part D coverage and elevated LDL cholesterol at baseline who initiated or intensified statins to lower cholesterol within one year of baseline. This will also be analyzed separately for just high CVD risk enrollees and for high and medium CVD risk enrollees combined	1 year
Percent of eligible beneficiaries using anti-hypertensive medications	Percent of beneficiaries with Medicare Part D coverage and elevated blood pressure at baseline who Initiated or intensified medications to lower blood pressure within one year of baseline. This will also be analyzed separately for just high CVD risk enrollees and for high and medium CVD risk enrollees combined	1 year
Percent of eligible beneficiaries using either statins or anti-hypertensive medications	Percent of beneficiaries with Medicare Part D coverage and either LDL cholesterol or elevated blood pressure at baseline who initiated or intensified statins to lower cholesterol or medications to lower blood pressure within one year of baseline. This will also be analyzed separately for just high CVD risk enrollees and for high and medium CVD risk enrollees combined	1 year
Proportion of providers reporting they calculate CVD risk scores for at least half of their Medicare beneficiaries	Proportion of providers who self-report that they calculate a cardiovascular risk score for at least 50% of their Medicare beneficiary panel. Based on responses to the Million Hearts Provider Survey. A greater percent of providers calculating risk scores is a better outcome.	5 years
Proportion of providers reporting they review CVD risk scores more consistently	Proportion of providers who self-report that they review CVD risk scores more consistently now than before the start of the Million Hearts model. Based on responses to the Million Hearts Provider Survey. A greater percent of providers reporting that they review risk scores more consistently is a better outcome.	5 years
Proportion of providers reporting follow-up with high-risk beneficiaries through any mode to monitor plans to reduce risk at least every three months	Proportion of providers who self-report that once they have identified Medicare beneficiaries as having high CVD risk, that their practice follows up with the beneficiaries through any mode (e.g., office visits, telephone calls, emails, or letters) to monitor plans to reduce risk. Based on responses to the Million Hearts Provider Survey. A greater percent of providers reporting that they follow up with high-risk beneficiaries is a better outcome.	5 years

Collaborators and Investigators

• Sponsor: Mathematica Policy Research, Inc.
• Collaborators: RAND; Centers for Medicare and Medicaid Services

Study record dates

Study Major Dates

• Study Start (ACTUAL): January 3, 2017
• Primary Completion (ANTICIPATED): December 31, 2021
• Study Completion (ANTICIPATED): December 31, 2022

Study Registration Dates

• First Submitted: March 11, 2019
• First Submitted That Met QC Criteria: August 2, 2019
• First Posted (ACTUAL): August 6, 2019

Study Record Updates

• Last Update Posted (ACTUAL): August 6, 2019
• Last Update Submitted That Met QC Criteria: August 2, 2019
• Last Verified: August 1, 2019

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Cardiovascular Diseases
• Other Study ID Numbers: MPR50496

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No