Residents' End-of-Life Hospital Use Is Lower in Nursing Homes with Substantial Palliative Care Knowledge (FR431-B)

Abstract

1.Describe how greater palliative care knowledge in nursing homes is associated with end-of-life hospital use.2.Discuss how the research findings in the context of the current nursing home/hospice policy environment and how findings support the diverse efforts underway to introduce palliative care expertise into nursing homes. Little is known about how a nursing home's (NH) level of palliative care (PC) knowledge and practice is associated with end-of-life care To understand how higher NH PC knowledge and practice is associated with residents' end-of-life use of hospitals In 2009-2010, questions from the validated NH “Palliative Care Survey” were administered to a stratified, proportionate random sample of 4,149 U.S. NH directors of nursing (DoNs); contact was achieved at 3,695 of those and 61.5% of DoNs responded. Based on responses, PC scores were created; the knowledge score ranged from 0 to 4 and practice from 0 to 9. These scores were merged with 2009-2010 NH resident assessment (MDS) data and Medicare claims data for the 46,518 decedents from study NHs to obtain resident-level covariates and study outcomes: any hospitalization or ER visit in the last 30 days of life and terminal hospitalization. NH and county-level covariates were obtained from OSCAR survey data, the Area Resource File, and from aggregated MDS and claims data (ie, percent of total NH days on hospice). Multivariate logistic regressions with generalized estimating equations were used to test associations. Controlling for an NH's hospice use, for each 1-point increase in an NH's PC knowledge score, residents had a 10% lower likelihood of having an end-of-life hospitalization (AOR, 0.90; 95% CI, 0.87, 0.93); a 6% lower likelihood of an ER visit (AOR, 0.94; 95% CI, 0.91, 0.97); and an 8% lower likelihood of a terminal hospitalization (AOR, 0.92; 95% CI, 0.89, 0.95). Similar associations were not observed for residents dying in NHs with greater PC practice scores. Greater NH PC knowledge appears important in reducing terminal hospital and ER use.