Conference Coverage

Heart failure readmission penalties linked with rise in deaths

 

Key clinical point: During the first 2 years of Medicare’s Hospital Readmissions Reduction Program hospital readmissions after a heart failure hospitalization declined but mortality simultaneously increased.

Major finding: Risk-adjusted 30-day readmissions fell by a relative 9% during the reduction program but relative mortality rose by 18%.

Data source: Review of 115,245 Medicare beneficiaries with heart failure treated at hospitals in the Get With the Guidelines--Heart Failure program.

Disclosures: Dr. Fonarow has been a consultant to Amgen, Janssen, Medtronic, Novartis, and St. Jude. Dr. Stevenson has been a consultant to Abbott, has received travel support from St. Jude, and has received research funding and food from Novartis.


 

AT THE AHA SCIENTIFIC SESSIONS

– Evidence continues to mount that Medicare’s penalization of hospitals with excess heart failure readmissions has cut readmissions but at the apparent price of more deaths.

During the penalty phase of the Hospital Readmission Reduction Program (HRRP), which started in Oct. 2012, 30-day all-cause mortality following a heart failure hospitalization was 18% higher compared with the adjusted rate during 2006-2010, based on Medicare data from 2006-2014 that underwent “extensive” risk adjustment using prospectively-collected clinical data, Gregg C. Fonarow, MD, and his associates reported in a poster at the American Heart Association scientific sessions. During the same 2012-2014 period with imposed penalties, 30-day all-cause readmissions following an index heart failure hospitalization fell by a risk-adjusted 9% compared to the era just before the HRRP. Both the drop in readmissions and rise in deaths were statistically significant.

Mitchel L. Zoler/Frontline Medical News
Dr. Gregg C. Fonarow


A similar pattern existed for the risk-adjusted readmissions and mortality rates during the year following the index hospitalization: readmissions fell by 8% compared with the time before the program but deaths rose by a relative 10%, also statistically significant differences.

“This is urgent and alarming. The Centers for Medicare & Medicaid Services needs to revamp the program to exclude heart failure patients and take steps to mitigate the damage,” Dr. Fonarow said in an interview. He estimated that the uptick in mortality following heart failure hospitalizations is causing 5,000-10,000 excess annual deaths among U.S. heart failure patients that are directly attributable to the HRRP. Similar effects have not been seen for patients with an index hospitalization of pneumonia or acute MI, two other targets of the HRRP, he noted.

The HRRP “currently has penalties for readmissions that are 15-fold higher than for mortality. They need to penalize equally, and they need to get at the gaming that hospitals are doing” to shift outcomes away from readmissions even if it means more patients will die. Heart failure patients “who need hospitalization are being denied admission by hospitals out of fear of the readmissions penalty,” said Dr. Fonarow, professor and co-chief of cardiology at the University of California, Los Angeles. “Seeing increased mortality linked with implementation of the penalty is completely unacceptable.”

Although a prior report used similar Medicare data from 2008-2014 to initially find this inverse association, that analysis relied entirely on administrative data collected in Medicare records to perform risk adjustments (JAMA. 2017 July 17;318[3]:270-8). The new analysis reported by Dr. Fonarow and his associates combined the Medicare data with detailed clinical records for the same patients collected by the Get With the Guidelines--Heart Failure program. The extensive clinical data that the researchers used for risk-adjustment allowed for a more reliable attribution to the HRRP of readmission and mortality differences between the two time periods. Despite the extensive risk adjustment “we see exactly the same result” as initially reported, Dr. Fonarow said.

The findings “remind us that it is very important to look at the unintended consequences” of interventions that might initially seem reasonable, commented Lynne Warner Stevenson, MD, professor and director of cardiomyopathy at Vanderbilt University in Nashville, Tenn.

Mitchel L. Zoler/Frontline Medical News
Dr. Lynne Warner Stevenson


Concurrent with the presentation at the meeting the results also appeared in an article published online (JAMA Cardiol. 2017 Nov 12;doi:10.1001/jamacardio.2017.4265).

A separate analysis of data collected in the Get With the Guidelines--Heart Failure during 2005-2009 showed that within the past decade the 5-year survival of U.S. hospitalized heart failure patients has remained dismally low, and similar regardless of whether patients had heart failure with reduced ejection fraction (HFrEF, 46% of all heart failure patients in the analysis), heart failure with preserved ejection fraction (HFpEF, also 46% of patients), or the in-between patients who had heart failure with borderline ejection fraction (HFbEF, an ejection fraction of 41%-49%, in 8% of patients).

The results, from 39,982 patients, showed a 75% mortality rate during 5-years of follow-up, with similar mortality rates regardless of the patient’s ejection-fraction level, reported Dr. Fonarow and his associates in a separate poster. In every age group examined, patients with heart failure had dramatically reduced life expectancies compared with the general population. For example, among heart failure patients aged 65-69 years in the study, median survival was less than 4 years compared with a 19-year expected median survival for people in the general U.S. population in the same age range.

These very low survival rates of heart failure patients initially hospitalized for heart failure during the relatively recent era of 2005-2009 “is a call to action to prevent heart failure,” said Dr. Fonarow.

The poor prognosis most heart failure patients face should also spur aggressive treatment of HFrEF patients with all proven treatments, Dr. Fonarow said. It should also spur more effort to find effective treatments for HFpEF, which currently has no clearly-proven effective treatment.

These results also appeared in a report simultaneously published online (J Amer Coll Cardiol. 2017 Nov 12;doi: 10.1016/j.jacc.2017.08.074).

On Twitter @mitchelzoler

Related Articles