But recent Mathematica evaluation of CMS' Million Hearts program shows no cost saving
CMS’ randomized trial of a program that pays physicians for assessing and reducing cardiovascular (CVD) disease risk is resulting in greater use of CVD medications, according to results presented at the American Heart Association’s Quality of Care and Outcomes Research (QCOR) meeting this weekend. But the difference is not huge, partly because many patients are already taking medications.
During the first two years of the program, which is called the Million Hearts Cardiovascular Disease Risk Reduction Model, patients at high risk for CVD in the intervention group were four percentage points more likely (28% vs. 24%) than the high-risk patients in the control group to start or intensify their use of statins or anti-hypertensive medication, according to results presented by G. Greg Peterson of Mathematica Policy Research. The vast majority (90%) of the high-risk patients were already taking a statin or an antihypertensive medication so much of the change was in intensifying the medication.
When Peterson and his colleagues presented this data in a November 2019 Mathematica evaluation report, they characterized the 28% vs. 24% difference as modest, although they noted that there was a spillover effect into the medium-risk patients.
The program uses a calculator to estimate the likelihood of someone having a heart attack or stroke in the next 10 years. For the purposes of this program, high risk is if the risk of heart attack or stroke is 30% or higher in the next 10 years; medium is between 15% and 30%, and low is less than 15%.
Peterson and his co-investigators also surveyed practices in the Million Hearts program about their assessment of patients because the program includes a financial incentive for assessment. They reported at the AHA meeting that the intervention group providers were twice (71% vs. 39%) as likely to have assessed the CVD risk of Medicare beneficiaries as the control group.
In the 217-page evaluation report that Mathematica published late last year, Peterson and his colleagues said that difference is indication that the “intervention organizations appear to have substantially improved the delivery of CVD preventive care in ways CMS envisioned.”
Million Hearts program is an ambitious attempt to measure the effect of relatively small financial incentives on the management of CVD diseases. Launched in 2017, the program randomly assigned half of the 516 participating healthcare organizations (primary and specialty physician practices, health centers, and hospital) into the group that received incentives and the other half into a group that didn’t. The incentives include a one-time payment of $10 for each Medicaid beneficiary whose CVD risk is assessed. Starting in the second year of the five-year trial, CMS is making monthly payments to the intervention group organizations that hinge on much the CVD risk reduction scores of their high-risk patients go down. The payments range from nothing up $10 per high-risk beneficiary; the greater the risk reduction, the higher the payment. The Mathematica evaluation report says that CMS made a total of roughly $5.6 million in payments to the intervention group organizations during the first 18 months of Million Hearts.
The evaluation report, which covers the first two years of the program, has a good news/bad news dichotomy. The intervention and its financial incentive, says the report, appear to have reduced the likelihood of dying by about 7% among the high- and medium-risk beneficiaries. And the data suggest that the CVD risk scores for the high-risk beneficiaries improved by 20%, primarily because of improvements in blood pressure.
The bad news: The Medicare Part A and Part B spending of the intervention and control groups was about the same, so as a payer, Medicare didn’t save any money to offset the $5.6 million in payments. Mathematica’s evaluation also says CVD hospitalizations increased among the high-risk beneficiaries in intervention group and emergency department visits went up among the intervention group’s high- and medium-risk beneficiaries.