JAMA Netw Open. 2026 Mar 2;9(3):e260755. doi: 10.1001/jamanetworkopen.2026.0755.
ABSTRACT
IMPORTANCE: Hospice is central to end-of-life (EOL) care for patients with advanced cancers and is an excluded benefit under Medicare Advantage (MA), with coverage instead provided by traditional Medicare (TM). With growing MA penetration, more beneficiaries also switch between MA and TM for financial protection and physician access considerations, although less is known about how different Medicare programs and plan switching behaviors affect EOL care for patients with advanced cancers.
OBJECTIVE: To evaluate hospice utilization and places of hospice care by Medicare plan switching patterns.
DESIGN, SETTING, AND PARTICIPANTS: This retrospective cohort study used Surveillance, Epidemiology, and End Results (SEER) Medicare data to identify beneficiaries aged ≥66 years diagnosed with distant-stage female breast, colorectal, lung, pancreatic, or prostate cancers from 2010 to 2019 who died by 2020. Beneficiaries were followed for up to 1 year before death. Data were analyzed from August 1, 2024, to December 14, 2025.
EXPOSURE: Plan switching patterns classified as continuous MA, continuous TM, MA to TM, TM to MA, and other (ie, multiple switches).
MAIN OUTCOMES AND MEASURES: Main outcomes were hospice enrollment in the last year of life and within 3 days of death, total hospice length of stay, and place of last hospice stay (home, nursing home, hospice facility, inpatient facility, or other) using multivariable regressions.
RESULTS: The sample included 196 536 decedents (46.5% female, 49.2% aged 66-74 years). Plan switching was infrequent (1.5% TM to MA; 1.8% MA to TM). Those who switched plans were more likely to be members of racial and ethnic minority groups and dual Medicare-Medicaid enrollees. Hospice enrollment was highest for those with continuous MA (74.8%), followed by those who switched from TM to MA (69.0%), those with continuous TM (68.5%), and those who switched from MA to TM (66.4%). Continuous MA beneficiaries had longer hospice stays than continuous TM beneficiaries (48.3 vs 43.8 days). Compared with continuous TM, continuous MA beneficiaries were more likely to receive hospice at home (1.93 percentage points [pp]; 95% CI, 1.40-2.45 pp; P < .001), while those who switched from MA to TM were more likely to receive hospice in nursing homes (2.45 pp; 95% CI, 1.26-3.63 pp; P < .001), particularly among dual Medicare-Medicaid enrollees (6.01 pp; 95% CI, 2.80 to 9.21 pp; P < .001).
CONCLUSIONS AND RELEVANCE: In this cohort study of Medicare decedents with advanced cancers, continuous MA enrollees were most likely to receive hospice at home, while those who switched from MA to TM more frequently received hospice care in nursing homes. Plan switching near the EOL may reflect access barriers, highlighting the importance of addressing care coordination to improve EOL care.
PMID:41874509 | DOI:10.1001/jamanetworkopen.2026.0755
