Estimating the Impact of Hospice Care on Key Patient-Family Care Outcomes Using a Nationwide US Probability Sample

Hospice is touted as an exemplary model for end-of-life care. However, there is little generalizable evidence estimating benefits of hospice at the national level. Using a national population-based probability sample of U.S. adults with weights applied (data collected Spring/Summer 2015), we conducted six logistic regression models to evaluate linkages between hospice involvement and end-of-life outcomes (pain, home death, presence of family, access to life-prolonging care, respect for spiritual/religious beliefs, financial burden) from N = 235 informal caregivers of decedents prior to death. Respondents were M = 55 years old (SD = 17), 55% female, 77% White, 91% non-Hispanic, and 71% had cared for a hospice recipient. In fully adjusted analyses, hospice users had better reported outcomes observed in two models: (a.) presence of family/friends at death (OR = 2.82, SE = .48, P < .05), and (b.) respect for spiritual/religious beliefs (OR = 9.52, SE = .73, P < .01). Hospice utilization was not statistically significant (P > .05) in all other adjusted models. Although more research is needed, findings support claims that hospice remains a high quality option for end-of-life care in the U.S.