Nursing Home Ratings and Characteristics Predict Hospice Use Among Decedents With Serious Illnesses

Approximately one-third of older Americans experience a nursing home (NH) stay within 3 months of death, but it is unclear how NH characteristics influence end-of-life care. We investigated associations between NH characteristics and hospice use.

Retrospective cohort study analyzing Medicaid and Traditional Medicare claims, Minimum Data Set, Centers for Medicare and Medicaid Services (CMS) Care Compare ratings, and other NH characteristics.

Connecticut Medicaid-insured individuals with a serious illness and an NH stay within 6 months of death, deceased January 1, 2017-September 30, 2024.

Multivariable logistic regression analyzing associations between NH characteristics and (1) hospice use within 6 months of death and (2) short hospice use (≤7 days). Covariates included sociodemographics and clinical characteristics.

Of 25,302 individuals, 51% were ≥85 years, 64.8% were women, and 79.3% were non-Hispanic White. Overall, 12,453 (49.2%) received hospice care, 4768 of 12,453 (38.3%) with short hospice use (≤7 days). Short-term (vs long-term) NH stays were associated with increased odds of hospice use and short hospice use. Many NH characteristics were associated with hospice use. Individuals with long-term stays had lower odds of hospice use with stays at NHs with the highest (vs lowest) CMS ratings for quality measures and staffing [odds ratio (OR), 0.55; 95% confidence interval (CI), 0.44-0.68, OR, 0.78; 95% CI, 0.66-0.93, respectively]. Those with short-term stays had lower odds of hospice use with stays at NHs with the highest CMS Health Inspection ratings. People with long-term stays at NHs that were part of a chain, had Alzheimer's care units, or had more beds had increased odds of hospice care.

Among Connecticut Medicaid-insured decedents with NH stays, people with long-term stays and stays in NHs with better CMS ratings had lower odds of hospice use. The inverse relationship between hospice care and NH quality warrants further research and consideration of incorporating end-of-life quality measurement into CMS quality ratings.