According to the Medicare Payment Advisory Commission (MedPAC), hospital readmission rates have fallen following implementation of the Hospital Readmissions Reduction Program (HRRP), and HRRP did not have a negative impact on mortality rates.
In early January, Craig Lisk, MS, and Jeff Stensland, PhD, both MedPAC analysts, presented
findings of their analysis on HRRP, the report for which is due before Congress later in June.
The idea for a program that would penalize hospitals with high readmission rates was floated back in 2008, when the MedPAC commission was concerned about the lack of care coordination and deficits in care transitions between acute and postacute settings spiked the readmission rate, Lisk said
during his presentation. However, the absence of financial incentives prevented hospitals from taking steps to improve care that patients received outside, he added.
After CMS started public reporting of hospital readmission rates in 2009, Congress enacted HRRP in 2010. Consequently, hospitals with high readmission rates (above an average threshold) during a 2-year period between 2010 and 2012 saw a reduction in their hospital inpatient payments in 2013.
Lisk pointed out that several studies conducted to assess the impact of HRRP observed a drop in readmission rates on both a raw and risk-adjusted basis, with a greater reduction documented for hospitals covered by the program compared with critical access hospitals that were not, giving the program at least partial credit for the reduced readmissions. This body of evidence included a study
published in Annals of Internal Medicine
, which found that low-performing hospitals saw a significant reduction in readmissions.
However, questions lingered, including:
- Were reductions real or were patients being classified as observation only?
- Did observation stays and emergency department (ED) visits increase post HRRP?
- Were intensive coding and higher risk scores responsible for reduced risk-adjusted readmissions?
- Did HRRP increase mortality?
Under a mandate put in place by the 21st Century Cures Act, MedPAC was directed by Congress to understand the relation between reduced readmissions and changes in outpatient and emergency services, Lisk said. MedPAC subsequently conducted an analysis that examined the association between readmission rates and changes in ED visits, observation stays, and 30-day mortality rates post discharge.
The study found that the raw readmission rates declined for all 3 measures: all-cause, unplanned, and potentially preventable readmissions. Readmission rates also decreased for each condition covered by HRRP: heart failure, chronic obstructive pulmonary disease (COPD), pneumonia, hip and knee replacement, and acute myocardial infarction (AMI). The trend persisted for risk-adjusted readmissions for the HRRP-covered conditions as well.
When the study evaluated whether the reductions in risk-adjusted readmissions were real or the result of coding discrepancies, a 17% decrease in per capita admissions was noted between 2010 and 2016. The authors attributed this decrease to a combination of outpatient care, reduced number of 1-day stays, and incentives provided under the Recovery Audit Contractor program and Accountable Care Organizations. “Part of the change in reported complexity could have been due to changes in coding,” Lisk told the commissioners during the meeting.
Although inpatient use and postdischarge readmissions declined, the analysis found a steady rise of both ED visits and observation use, meaning that ED visits and observation visits could substitute for readmissions, according to Lisk. However, the coefficient of correlation between the variables was very weak, accounting for only 3% of variation, which the authors found could be explained by comparing conditions covered by HRRP and those that were not. Interestingly, the unplanned readmission rate saw a 2.9% drop for the 5 conditions covered by HRRP, compared with a 1.3% drop for those that were not covered. Observation visits increased at the same rate (1%) for the 2 groups, as did ED visits (2.0% and 2.2%, respectively).
Further, evaluation of mortality rates found that both raw and risk-adjusted mortality decreased for AMI, pneumonia, COPD, and hip and knee replacement between 2008 and 2016 after HRRP was implemented. For heart failure, however, the raw mortality rate seemed to increase, but the risk-adjusted mortality rate decreased during the same time period.
Overall, the analysis found that Medicare spent $2.04 billion less on readmissions in 2016 as a result of the reduced readmission rates following HRRP initiation, after accounting for slight increases in observation stays and ED visits.
“The program is not perfect,” Lisk said, and he presented certain policy recommendations to improve HRRP, including revising the penalty formula and expanding the program beyond the current 5 health conditions.