Racial and Ethnic Differences in Hospice Use and Hospitalizations at End-of-Life Among Medicare Beneficiaries With Dementia
Pei-Jung Lin,Yingying Zhu,Natalia Olchanski,Joshua T Cohen,Peter J Neumann,Jessica D Faul,Howard M Fillit,Karen M Freund
DOI: https://doi.org/10.1001/jamanetworkopen.2022.16260
2022-06-01
Abstract:Importance: The pool of studies examining ethnic and racial differences in hospice use and end-of-life hospitalizations among patients with dementia is limited and results are conflicting, making it difficult to assess health care needs of underresourced racial and ethnic groups. Objective: To explore differences in end-of-life utilization of hospice and hospital services among patients with dementia by race and ethnicity. Design, setting, and participants: This cohort study used national survey data from the Health and Retirement Study linked with Medicare and Medicaid claims that reflected a range of socioeconomic, health, and psychosocial characteristics. Eligible participants were Medicare fee-for-service beneficiaries aged 65 years or older diagnosed with dementia who died between 2000 and 2016. Analyses were performed from June to December 2021. Exposures: Race and ethnicity. Main outcomes and measures: We examined the frequency and costs of hospice care, emergency department (ED) visits, and hospitalizations during the last 180 days of life among Medicare decedents with dementia. We analyzed the proportion of dementia decedents with advance care planning and their end-of-life care preferences. Results: The cohort sample included 5058 beneficiaries with dementia (mean [SD] age, 85.5 [8.0] years; 3038 women [60.1%]; 809 [16.0%] non-Hispanic Black, 357 [7.1%] Hispanic, and 3892 non-Hispanic White respondents [76.9%]). In adjusted analysis, non-Hispanic Black decedents (odds ratio [OR], 0.65; 95% CI, 0.55-0.78), nursing home residents (OR, 0.81; 95% CI, 0.71-0.93), and survey respondents represented by a proxy (OR, 0.84; 95% CI, 0.71-0.99) were less likely to use hospice, whereas older decedents (age 75-84 vs 65-74 years: OR, 1.39; 95% CI, 1.12-1.72; age ≥85 vs 65-74 years: OR, 1.39; 95% CI, 1.13-1.71), women (OR, 1.19; 95% CI, 1.05-1.35), and decedents with higher education (high school vs less than high school: OR, 1.17; 95% CI, 1.01-1.36; more than high school vs less than high school: OR, 1.32; 95% CI, 1.13-1.54), more severe cognitive impairment (OR, 1.51; 95% CI, 1.02-2.23), and more instrumental activities of daily living limitations (OR, 1.07; 95% CI, 1.01-1.12) were associated with higher hospice enrollment. A higher proportion of Black and Hispanic decedents with dementia used ED (645 of 809 [79.7%] and 274 of 357 [76.8%] vs 2753 of 3892 [70.7%]; P < .001) and inpatient services (625 of 809 [77.3%] and 275 of 357 [77.0%] vs 2630 of 3892 [67.5%]; P < .001) and incurred roughly 60% higher inpatient expenditures at the end of life compared with White decedents (estimated mean: Black, $23 279; 95% CI, $20 690-$25 868; Hispanic, $23 471; 95% CI, $19 532-$27 410 vs White, $14 609; 95% CI, $13 800-$15 418). A higher proportion of Black and Hispanic than White beneficiaries with dementia who were enrolled in hospice were subsequently admitted to the ED (56 of 309 [18.1%] and 22 of 153 [14.4%] vs 191 of 1967 [9.7%]; P < .001) or hospital (48 of 309 [15.5%] and 17 of 153 [11.1%] vs 119 of 1967 [6.0%]; P < .001) before death. The proportion of dementia beneficiaries completing advance care planning was lower among Black (146 of 704 [20.7%]) and Hispanic (66 of 308 [21.4%]) beneficiaries compared with White beneficiaries (1871 of 3274 [57.1%]). A higher proportion of Black and Hispanic decedents with dementia had written instructions choosing all care possible to prolong life (30 of 144 [20.8%] and 12 of 65 [18.4%] vs 72 of 1852 [3.9%]), whereas a higher proportion of White decedents preferred to limit care in certain situations (1708 of 1840 [92.8%] vs 114 of 141 [80.9%] and 51 of 64 [79.7%]), withhold treatments (1448 of 1799 [80.5%] vs 87 of 140 [62.1%] and 41 of 62 [66.1%]), and forgo extensive life-prolonging measures (1712 of 1838 [93.1%] vs 120 of 138 [87.0%] and 54 of 65 [83.1%]). Conclusions and relevance: The results of this cohort study highlight unique end-of-life care utilization and treatment preferences across racial and ethnic groups among patients with dementia. Medicare should consider alternative payment models to promote culturally competent end-of-life care and reduce low-value interventions and costs among the population with dementia.