Did the HRRP lower readmissions at the cost of patient safety? The debate continues

The Hospital Readmissions Reduction Program (HRRP) has successfully slashed readmissions for heart failure, acute MI (AMI) and pneumonia without causing mortality increases, according to an analysis of Medicare data published in JAMA Network Open. But the author of an accompanying editorial isn’t convinced the results are so positive.

This topic has been a concern for researchers who worried the HRRP incentivized “inappropriate care strategies” for hospitals attempting to “game the system” and avoid paying penalties for 30-day readmission rates above the national average. Following the publication of one study suggesting heart failure mortality had increased after the policy was implemented, JACC: Heart Failure devoted a pair of point/counterpoint articles to researchers who argued whether the HRRP had indeed resulted in greater mortality for heart failure patients or whether the conclusions stemmed from incomplete or misconstrued evidence.

The new study looked at all three conditions initially included in the HRRP, encompassing 1.7 million AMI hospitalizations, 4 million heart failure hospitalizations and 3.5 million pneumonia hospitalizations among Medicare fee-for-service beneficiaries from 2006 through 2014. That timeframe spanned several years before the policy was announced in 2010 as well as a couple of years after the penalties for high readmissions went into effect in October 2012.

From 2006 to 2014, in-hospital mortality decreased for all three conditions:

  • AMI: 10.4 percent to 9.7 percent
  • Heart failure: 4.3 percent to 3.5 percent
  • Pneumonia: 5.3 percent to 4.0 percent

Thirty-day mortality decreased for AMI (7.4 percent to 7.0 percent) but increased for heart failure (7.4 percent to 9.2 percent) and pneumonia (7.6 percent to 8.6 percent). However, the researchers noted the rates of monthly post-discharge mortality were already increasing for those two conditions before the HRRP was announced. The already-positive slopes of those mortality graphs didn’t significantly increase when the policy was announced or implemented, suggesting the steady increase in mortality wasn’t linked to the HRRP.

“Among Medicare beneficiaries, there was no evidence for an increase in in-hospital or post-discharge mortality associated with HRRP announcement or implementation—a period with substantial reductions in readmissions,” wrote lead author Rohan Khera, MD, with the University of Texas Southwestern Medical Center, and colleagues. “The improvement in readmission was therefore not associated with any increase in in-hospital or 30-day post-discharge mortality.”

Khera and colleagues said the previous study showing an increase in heart failure mortality associated with the HRRP may not be generalizable because it only included data from a select group of hospitals participating in the Get With the Guidelines-Heart Failure registry.

“Findings from our study are particularly reassuring as the CMS considers expanding the HRRP to include a hospital-wide readmission measure with the goal of reducing readmissions for the entire spectrum of patients,” they wrote.

Ashish K. Jha, MD, MPH, isn’t convinced the data are entirely positive for the HRRP. Writing in a related editorial, he pointed out before the HRRP, inpatient mortality rates were declining for both AMI and heart failure. But after the policy was announced, that improvement flattened out.

These relative increases in inpatient mortality would logically decrease the rates of 30-day mortality, Jha noted, because the sickest patients are dying in the hospital rather than being discharged where they would be eligible for readmission.

“This makes the fact that post-discharge mortality appears to have not increased far less reassuring than it would be otherwise,” he said.

“Right now, a high-readmission, low-mortality hospital will be penalized at 6 to 10 times the rate of a low-readmission, high-mortality hospital. The signal from policymakers is clear—readmissions matter a lot more than mortality—and this signal needs to stop. If patients understood what policymakers were doing, they would be outraged.”

Jha said the cost-benefit balance of the HRRP remains unclear eight years after the announcement of the program, something he deems unacceptable considering its potential to significantly impact clinical practice.

“If thousands of hospitals and tens of thousands of clinicians are going to be asked to change the way they provide care, adequate resources and a robust set of efforts are needed to fully evaluate the program from a variety of perspectives to ensure that we continually improve the program,” Jha wrote. “All of us as clinicians need to know that if policy makers are going to ask us to change how we practice medicine, it is helping us deliver better care to our patients.”