Hospice Use and Proxy-Reported Last Month of Life Care Quality Among Older Medicare Beneficiaries: A Pooled NHATS Analysis

Background/Objectives: Hospice care is intended to improve comfort, symptom management, communication, and person-centered care. National evidence using consistent last month of life measures across multiple survey rounds remains limited. This study examined the association between hospice use and proxy-reported quality of care during the last month of life among older Medicare beneficiaries. Methods: A pooled repeated cross-sectional analysis was conducted using 4666 completed Last Month of Life interviews from the National Health and Aging Trends Study (NHATS), Rounds 3–14 (survey years 2013–2024). Hospice use was identified using a composite of five service- and setting-based indicators to account for questionnaire routing. The primary outcome was an overall care rating of excellent or very good. Secondary outcomes assessed personal care, respect, communication, decision-making, symptom assistance, and adequacy of assistance. Analyses used round- and cohort-appropriate analytic weights, 56 Fay-balanced repeated replication weights, equal-round normalization, and survey-weighted logistic regression. Results: Hospice status was classifiable for 4583 respondents, of whom 2658 received hospice care. Weighted hospice prevalence was 57.5% (95% CI, 55.6–59.5%). Among the 4445 respondents with a valid overall care rating, excellent or very good care was reported for 81.9% of hospice recipients and 74.7% of nonrecipients. In the demographic-adjusted model, hospice use was associated with higher odds of an excellent or very good care rating (OR, 1.48; 95% CI, 1.20–1.83); the estimate was 1.83 (95% CI, 1.46–2.31) after an additional adjustment for contemporaneous severity-related measures. Standardized probabilities were 82.6% with hospice and 73.1% without hospice care, an adjusted difference of 9.5 percentage points (95% CI, 5.7–13.3). Hospice care was also associated with greater receipt and adequacy of assistance for pain, breathing difficulties, and anxiety or sadness. Conclusions: Hospice use was associated with better proxy-reported overall care and more complete symptom assistance during the last month of life. These findings are consistent with a potential contribution of interdisciplinary hospice care, including symptom assessment, caregiver education, and communication. Residual confounding, selective hospice enrollment, and retrospective proxy reporting preclude causal interpretation.