Broader readmissions reduction program could penalize hospitals serving low-income patients

If the current Hospital Readmissions Reduction Program (HRRP) were adjusted to include a wider range of conditions, “safety-net” hospitals could be penalized an additional $198,000 per year when compared to those with fewer low-income patients, according to an analysis published in The New England Journal of Medicine.

“Although there is now little difference in readmission rates between safety-net hospitals and other hospitals, the hospital-wide measure would increase the disparity in penalties,” wrote Rachael B. Zuckerman, PhD, and colleagues, all of whom are employed by the Department of Health and Human Services, which funded the study.

The concept of a hospital-wide readmission measure has been discussed in recent years as concerns have been raised about the HRRP established as part of the Affordable Care Act in 2010. Beginning in the 2013 fiscal year, hospitals were penalized if they had higher-than-expected 30-day readmission rates after hospitalizations for acute MI, heart failure or pneumonia. Total hip and knee replacement and chronic obstructive pulmonary disease were added to the program in fiscal year 2015.

But Zuckerman and co-authors pointed out the following issues with HRRP:

  • Decreases in readmissions were greater for the targeted conditions versus nontargeted ones, suggesting it could be beneficial to incentivize reducing readmissions for a wider range of conditions.
  • A “substantial number” of hospitals are ineligible to receive a penalty because of small sample size, limiting the reach of the program. Hospitals must have had 25 hospitalizations for the conditions over a three-year period to be eligible.
  • Safety-net hospitals faced larger penalties during the early years in the program.

To project how a hospital-wide measure of readmissions might impact penalties, the researchers compared one year of hospital admissions data for a range of conditions to three years of data for the five conditions targeted by the HRRP. More than 6.8 million admissions were included in the sample for the hospital-wide model and nearly 4.4 million admissions were included for the five condition-specific measures.

Moving to a hospital-wide model only increased the number of hospitals eligible for penalty by 4 percent. The authors projected including more conditions would cause fewer hospitals to be penalized (43 percent versus 80 percent). However, those hospitals would pay heftier penalties—$393,000 on average instead of $134,000.

In addition, penalized safety-net hospitals would owe an average of $551,000 compared to $353,000 for other hospitals.

“Under the hospital-wide measure, all index stays contribute to a single, hospital-wide readmission rate, and hospitals have only one chance to be penalized, which concentrates the penalties while reducing the proportion of hospitals penalized,” Zuckerman and colleagues wrote. “Since safety-net hospitals tend to perform slightly worse on the hospital-wide measure, they are more likely to receive a penalty, which would increase the disparity in penalties between the two groups.”

The authors attempted two separate adjustment techniques to balance the burden. First, they calculated penalties with adjustment for Medicaid enrollment, but that method didn’t erase the disparity between safety-net and other hospitals.

However, when penalties were assigned within deciles measuring the proportion of Medicaid and low-income Medicare patients a hospital treats, safety-net hospitals paid only $21,000 more on average.

“Transition to a hospital-wide measure would require an adjustment in the penalty formula to keep penalties in the same range for most hospitals and without a change in procedures would have a deleterious effect on safety-net hospitals,” the researchers wrote.