VA Efforts to Cut HF Readmissions are Working, With No Uptick in Deaths

The findings contrast with a Medicare study showing a trend toward increased death after a similar program was instituted.

VA Efforts to Cut HF Readmissions are Working, With No Uptick in Deaths

Over a 10-year period in which the Veterans Affairs healthcare system worked to reduce readmissions among patients with heart failure, a steady decline in readmissions was seen, with no accompanying increase in mortality, a new study suggests.

The findings are in contrast to a 2018 analysis of Medicare patients, which showed an increase in deaths within 30 days of discharge among patients hospitalized for heart failure or pneumonia and hinted that those deaths may have been an unintended negative consequence of the Hospital Readmissions Reduction Program (HRRP), enacted by the 2010 Affordable Care Act.

“I'd say the jury is still out on the impact of the HRRP,” said Justin T. Parizo, MD ( Stanford University School of Medicine, CA), who led the VA study. “Given our data, we can definitively say that the mortality is not worsening in the overall heart failure population. But beyond that, I don't think that we can say exactly why.”

Efforts by the VA to reduce heart failure readmissions during the study period included public reporting, incorporating 30-day readmission rate for heart failure into each hospital’s overall star ratings, and encouraging hospitals to participate in quality improvement programs such as the American College of Cardiology’s Hospital to Home Initiative.

Speaking with TCTMD, Parizo said while the VA and Medicare populations are close in age and are both US-based, the VA’s structure and systems of care may explain some of the differences.

Ten-Year Trends

Published online June 17, 2020, ahead of print in JAMA Cardiology, Parizo’s study of 164,566 patients with heart failure treated in VA hospitals between January 2007 and September 2017 found a 2% decline in 30-day readmissions over the study period and a 1.1% decline in 1-year readmissions. Mortality rates at 30 days decreased by 0.5% over the study period, while mortality rates at 1-year increased by 1.3% (P < 0.001 for trend for both comparisons).

The VA study also examined 30-day readmission rates by ejection fraction, noting similar declines in patients with an LVEF of ≥ 40% or < 40%. There also was no significant difference in 30-day mortality rates by ejection fraction strata. Analysis by region where VA treatment occurred showed significant readmission reductions in all geographic areas except the Western part of the United States. There were no interactions by geographic region for the mortality increase.

Parizo said the robustness of the VA’s database and its use of a single electronic health record also enabled additional patient-level factors to be included in its analyses of readmissions and mortality that were not analyzed in the 2018 Medicare study. Those additional factors, he said, provide needed insight into the health status of the population.

“There were some concerns that potentially the Medicare population over time was having a worsening of illness severity that couldn't be completely adjusted for,” he said. “In our study, we were able to adjust for things like systolic and diastolic blood pressure, weight, and various laboratory values. It gave us more confidence that we were adequately adjusting for trends of illness severity.”

Work Still to Do

Unlike the HRRP, which financially penalizes hospitals based on readmissions, the VA system does not.  Parizo and colleagues say the VA may be more successful at implementing coordinated interventions at an individual patient level and making certain resources easily available to its hospitals. Importantly, the VA has extensive programs focused on outpatient monitoring, which the researchers say may help identify issues early, particularly of a noncardiac nature.

Still, Parizo said more needs to be done to transition away from the current “10,000-foot view” of the situation to a more granular view that can identify drivers of readmission and create better interventions that keep patients out of the hospital. For every 20 VA patients who were readmitted, one died.

“We need prospective studies, and hopefully randomized studies, looking at the effects of those interventions, or at least interventions implemented in series across various different hospitals or hospital systems,” he concluded.

Disclosures
  • Parizo reports no relevant conflicts of interest.

Comments