The association between hospital end-of-life care quality and the care received among patients with heart failure.

CONTEXT Improving end-of-life care (EOLC) quality among heart failure patients is imperative. Data are limited as to the hospital processes of care that facilitate this goal. OBJECTIVES To determine associations between hospital-level EOLC quality ratings and the EOLC delivered to heart failure patients. METHODS Retrospective analysis of the Veterans Health Administration (VA) and the Bereaved Family Survey data of heart failure patients from 2013-2015 who died in 107 VA hospitals. We calculated hospital-level observed-to-expected casemix-adjusted ratios of family-reported "excellent" EOLC, dividing hospitals into quintiles. Using logistic regression, we examined associations between quintiles and palliative care consultation (PCC), receipt of chaplain and bereavement services, inpatient hospice, and intensive care unit (ICU) death. RESULTS Of 6,256 patients, mean age was 77.4 (standard deviation=11.1), 98.3% were male, 75.7% white and 18.2% were black. Median hospital scores of "excellent" EOLC ranged from 41.3% (Interquartile range (IQR) 37.0-44.8%) in the lowest quintile to 76.4% (IQR 72.9-80.3%) in the highest. Patients who died in hospitals in the highest quintile, relative to the lowest, were slightly though not significantly more likely to receive a PCC (adjusted proportions 57.6% vs. 51.2%; p=0.32); but were more likely to receive chaplaincy (92.6% vs. 81.2%), bereavement (86.0% vs. 72.2%) and hospice (59.7% vs. 35.9%) and were less likely to die in the ICU (15.9% vs. 31.0%; p<0.05 for all). CONCLUSION Patients with heart failure who die in VA hospitals with higher overall EOLC quality receive more supportive EOLC. Research is needed that integrates care processes and develops scalable best-practices in EOLC across healthcare systems.