Association of the Hospital Readmissions Reduction Program With Mortality During and After Hospitalization for Acute Myocardial Infarction, Heart Failure, and Pneumonia

Abstract

Key Points Question Was the announcement or implementation of the Hospital Readmissions Reduction Program (HRRP) associated with an increase in mortality following hospitalization for acute myocardial infarction, heart failure, or pneumonia among Medicare beneficiaries? Findings In this cohort study, between 2006 and 2014, in-hospital mortality decreased for the 3 conditions while 30-day postdischarge mortality decreased for acute myocardial infarction but increased for heart failure and pneumonia. Before the announcement of the HRRP, postdischarge mortality was stable for acute myocardial infarction and increasing for heart failure and pneumonia, and there were no inflections in slope around the announcement or implementation of the HRRP. Meaning There was no evidence for increase in in-hospital or postdischarge mortality associated with the HRRP announcement or implementation—a period with substantial reductions in readmissions.

The association of hospital teaching intensity with 30‐day postdischarge heart failure readmission and mortality rates

“…However, since the implementation of the HRRP, the inverse association between postdischarge mortality and readmission rates has reversed direction in some studies. Readmission rates have fallen but mortality rates, which some studies suggest had previously been decreasing, have been found by some studies to be stalled or increasing …”

“…Using linked Get with the Guidelines‐Heart Failure Registry and Medicare data from 416 hospitals, Gupta and colleagues found that thirty‐day postdischarge risk‐adjusted readmission rates decreased from 20.0 percent to 18.4 percent after HRRP implementation, whereas 30‐day postdischarge risk‐adjusted mortality rate increased from 7.2 percent to 8.6 percent in the HRRP penalty phase, although these findings have been challenged by Khera et al Using CMS Hospital Compare data from 2009 to 2015, Chatterjee and Joynt Maddox found that between 2009 and 2015, heart failure, risk‐adjusted 30‐day mortality fell from 13.5 percent to 13 percent for 600 initially poor‐performing hospitals but increased from 10.9 percent to 12 percent (0.17 percent per year, P < .001) for the remaining 3196 hospitals. Wadhera et al found that thirty‐day postdischarge mortality for heart failure patients increased before HRRP announcement, but this trend accelerated significantly after HRRP announcement, driven largely by patients who were not readmitted (and perhaps who should have been).…”

“…Readmission rates have fallen but mortality rates, which some studies suggest had previously been decreasing, 14 have been found by some studies to be stalled or increasing. 14,30,[32][33][34]36,37,39 Dharmarajan et al 33 (P < .001). However, the slope of the increase in postdischarge mortality did not increase after HRRP announcement or implementation, and the authors speculate that patient severity rather than the HRRP could be responsible for these findings.…”

“…The most significant concern is the possible negative impact of the HRRP program on mortality, specifically for heart failure. Depending on the data source, statistical methodology for studying longitudinal trends and their association with the HRRP, and level of analysis (hospital vs aggregate national results), different studies have reached discordant conclusions regarding this crucial question …”

“…Gupta and colleagues 39 found that thirty-day postdischarge risk-adjusted readmission rates decreased from 20.0 percent to 18.4 percent after HRRP implementation, whereas 30-day postdischarge risk-adjusted mortality rate increased from 7.2 percent to 8.6 percent in the HRRP penalty phase, although these findings have been challenged by Khera et al32,36 Using CMS Hospital Compare data from 2009 to 2015,…”