By Melinda Young

Results of a comprehensive study, analyzing more than 6 million Medicare admissions, revealed declining 30-day hospital readmissions from 2009 to 2014. The readmissions rate fell from 19.65 per 100 beneficiaries in 2009 to 17.27 in 2014, a 12.1% decrease. Some policymakers have attributed the decline to the Hospital Readmissions Reduction Program (HRRP), introduced in 2010. But researchers found the declining readmissions also could be explained by declining hospital admission rates over the same period.1

“We noted a trend that has largely gone unnoticed — that the period of readmission reductions also has been one in which admission rates declined substantially,” says Michael McWilliams, MD, PhD, Warren Alpert Foundation professor of healthcare policy and professor of medicine at Harvard Medical School and Brigham and Women’s Hospital.

Declining admissions likely was due to several causes, including the Recovery Audit Contractor (RAC) program that cracked down on very short hospital stays and prompted a surge in observation stays, McWilliams says.

“Other factors might include our growing ability to handle more complex conditions on an outpatient basis or in the emergency department without admission, and a decline in the number of hospital beds in the United States,” he explains.

Bed Decline Already Underway

The hospital bed decline has been happening for some time, he adds. “This [readmissions] decline cannot be due to the Hospital Readmission Reduction Program because it is evident across all payers,” McWilliams says. “Hospitals would not sacrifice 12% of their revenue in response to much smaller penalties applicable only to Medicare patients.”

Also, the admission trend was underway before the HRRP was announced and levied its first penalties to hospitals with too many 30-day readmissions.

“We know that most readmissions are not related to the care provided in prior admissions, but are independent events — that is, other admissions that happen to occur within 30 days of another,” McWilliams says.

Other Factors Affect Readmissions

Studies show that about 20-30% of readmissions are due to potentially preventable deficits in care during the prior admission. “An even smaller percentage are actually due to such deficits,” McWilliams explains. “Because the probability of two independent admissions falling within 30 days of each other is lower when there are fewer admissions per patient, we would expect that the declining admission rate would reduce readmission rates, too.”

Two other studies revealed the Medicare readmissions reduction was overstated because of a difference in diagnosis coding used to risk-adjust readmission rates.1,2,3

“As an analogy, you could think of a crowded dart board. As more darts are thrown, the probability of your next throw hitting a dart and falling to the floor goes up,” McWilliams says. “But that doesn’t mean that you are any worse at throwing darts.”

When investigators conducted a computer simulation to determine the readmission reduction expected from the declining admission rate, they found the expected decline in readmissions was as large as the observed reduction.

“This suggests the reduction attributed to the HRRP was likely the byproduct of a broader decline in admissions,” McWilliams says.

“The takeaway message from the study is either hospital responses to HRRP incentives were weak or ineffective, or they may have improved quality of care, but both resulted in prevented readmissions and induced readmissions,” he says. “This is because when we reach out to patients and follow patients, we often recognize unmet needs. If it’s the latter, it suggests readmissions are not a valid measure of quality and presents the question, ‘What is the point of a pay-for-performance program that does not reward improvements in the quality it elicits?” he adds. “Such a program would provide no incentive for providers to improve.”

The findings more likely suggest the HRRP’s efforts were ineffective, McWilliams says.

“In general, I think the national conversation would be better served by a focus on what improves quality rather than a continued focus on how to fix pay-for-performance programs that have intractable problems and cause unintended consequences,” he says.


  1. McWilliams JM, Barnett ML, Roberts ET, et al. Datawatch: Did hospital readmissions fall because per capita admission rates fell? Health Aff (Millwood) 2019;38:1840-1844.
  2. Ibrahim AM, Dimick JB, Sinha SS, et al. Association of coded severity with readmission reduction after Hospital Readmissions Reduction Program. JAMA Intern Med 2018;178:290-292.
  3. Ody C, Msall L, Dafny LS, et al. Decreases in readmissions credited to Medicare’s program to reduce hospital readmissions have been overstated. Health Aff (Millwood) 2019;38:36-43.