Missing Medicare data alters hospital penalties, U-M study finds

For more than a decade, hospitals have worked to help older adults avoid repeated inpatient stays, incentivized by a federal program that cuts Medicare reimbursements if hospitals have higher-than-expected rates of readmissions for people with certain conditions.

The Hospital Readmissions Reduction Program has helped spur innovation, including initiatives to better prepare patients and their families to manage care after hospitalization, and to support them virtually at home.

But a new University of Michigan study finds that these financial penalties have hit some hospitals harder than they should, even if those hospitals have done a reasonable job at keeping people with heart failure, pneumonia and other serious conditions from ending up back in a hospital within a month of leaving one.

Such hospitals have been paying inflated readmission penalties for a seemingly unrelated reason: They happen to serve higher percentages of older adults who have chosen to enroll in Medicare Advantage plans run by private insurance companies.

So what’s the connection? 

Currently, the federal government only grades hospitals on their readmission performance for older adults with traditional Medicare, which is run by the federal government. Data from Medicare Advantage are not currently included in the calculations that determine these penalties. 

This is a problem, because Medicare Advantage enrollees tend to be healthier than traditional Medicare beneficiaries. But the readmission penalty program’s risk-adjustment is unable to capture these differences. 

So, the researchers find, hospital performance looks worse for hospitals treating fewer traditional Medicare beneficiaries and more Medicare Advantage enrollees, even if those hospitals take the same actions to prevent readmissions as other hospitals. 

The study, published in JAMA Network Open, suggests that not incorporating Medicare Advantage data results in an unwarranted redistribution of nearly $300 million a year in readmission penalties across hospitals nationwide. That’s more than half the total amount of readmissions penalties incurred each year across all hospitals. 

Penalties are publicly reported every year and covered by the news media.

The Centers for Medicare and Medicaid Services has issued a rule to begin using Medicare Advantage data in the program. But even when it takes effect later this year, it will not affect hospital penalties for several years.

An unintended consequence with major consequences

The analysis shows an unintended consequence of the intersection between two major health policies of the last 15 years: the rapid rise in Medicare plan enrollment, and the HRRP, says senior author Geoffrey Hoffman, Ph.D., a professor at the U-M School of Nursing and member of the U-M Institute for Healthcare Policy and Innovation.

“Medicare Advantage has experienced extraordinary growth in the past decade, yet policymakers haven’t caught up with implications of this growth for Medicare payment policy that is based purely on traditional Medicare enrollment,” says Hoffman. “The omission of Medicare Advantage data highlights the continuing issue of inadequate measurement of patient risk in the Hospital Readmission Reduction Program, with important implications for the penalties that hospitals face.”

When the HRRP started calculating readmission rates in 2012, only 29% of older adults and people with disabilities chose Medicare Advantage plans.

Today, it’s 54%. But the distribution is not even across the country. 

Hoffman and lead author Zoey Chopra, M.A., mapped that distribution and divided more than 3,200 hospitals into five groups based on the MA enrollment levels in 1,486 counties. 
Chopra is working toward both a medical degree in the U-M Medical School and a Ph.D. in economics in the U-M College of Literature, Science and the Arts. He is a Medical Scientist Training Program fellow at the Medical School. 

Differences in hospitals and populations

The hospitals in the areas with the highest Medicare Advantage enrollment were much more likely to be larger, nonprofit, teaching-oriented and in urban areas than hospitals in the areas with the lowest levels of Medicare Advantage enrollment. 

Even when the researchers took into account an aspect of the readmission penalty program that only judges hospitals against groups of their peer hospitals, they still found that the Medicare Advantage enrollment rate mattered. 

Past studies have shown that older adults who choose Medicare Advantage plans tend to be in the younger age range of eligibility, and to have fewer serious health conditions. 

Hoffman has studied patterns of migration back to traditional Medicare by adults who had previously chosen Medicare Advantage plans, and factors that cause older adults to leave a Medicare Advantage plan for another Medicare Advantage plan or traditional Medicare.

Even though the readmission penalty program adjusts penalties based on the health risks of hospitals’ patients with the conditions that are included in the program, the lack of Medicare Advantage data could be a source of bias, he said. 

Areas with more Medicare Advantage enrollees may end up with higher-risk traditional Medicare beneficiaries, because healthier enrollees migrate to Medicare Advantage. But those risk differences can’t be captured in the data models used by CMS. Therefore, by basing the program’s penalties only on traditional Medicare enrollees, hospitals with more Medicare Advantage patients are at greater, unwarranted risk of readmission penalties.

“Our study observes an inadvertent consequence of restricting the readmissions program to traditional Medicare participants,” said Chopra. “At the time of the HRRP’s rollout, this made sense, given lower enrollment and concerns about accuracy of the Medicare Advantage data. However, including Medicare Advantage data now appears imperative to avoid unnecessary penalties for hospitals treating relatively more private pay patients.” 

Potential solutions

The new policy taking effect later this year rule will add Medicare Advantage data to HRRP calculations of readmissions, but Hoffman and Chopra have concerns about the completeness of what may be available. Risk-coding and policy differences across Medicare Advantage and traditional Medicare could also complicate comparisons of hospital performance.

It will also base penalties on the last two years’ worth of readmission data, instead of three. While the inclusion of Medicare Advantage data nearly doubles the amount of data used for comparisons, given concerns about data completeness and accuracy, this change may nonetheless make it harder statistically to see how hospitals have really done on keeping readmissions as low as possible, Hoffman said.

In addition to the rule, they suggest, CMS could consider factoring the percentage of Medicare Advantage enrollees in a hospital’s area or patient base into the calculations for its readmission rate and potential penalty. 

The study was funded by the National Institute on Aging of the National Institutes of Health (RO1AG074944, T32AG000221). Chopra’s funding is through the Population Studies Center at the U-M Institute for Social Research. Andrew Ryan, Ph.D., of Brown University is a co-author of the study.

Hospital Readmission Reduction Program Penalties for Hospitals with High Medicare Advantage Penetration, JAMA Network Open, DOI:10.1001/jamanetworkopen.2025.54972