Black and Hispanic patients with dementia were less likely to use hospice care at the end of life, but more likely to use emergency department (ED) and inpatient services, according to a cohort study of Medicare beneficiaries.

Among over 5,000 decedents, 38.2% of Black patients and 42.9% of Hispanic patients used hospice care in their last 180 days of life compared with 50.5% of white patients (P<0.001), reported Pei-Jung Lin, PhD, of Tufts Medical Center in Boston, and colleagues at the Alzheimer’s Association International Conference and in JAMA Network Open.

For ED visits and inpatient services, these rates were (both P<0.001):

• Black decedents: 79.7% and 77.3%
• Hispanic decedents: 76.8% and 77%
• White decedents: 70.3% and 67.5%

Black and Hispanic decedents also incurred approximately 60% higher inpatient expenses at the end of life compared with white patients, with an estimated mean of $23,279 and $23,471 versus $14,609.

According to Lin, this study is unique because it is so wide-ranging. “Previous studies were largely limited to nursing home residents and therefore omitted a substantial proportion of dementia patients living at home,” she told MedPage Today.

While the research doesn’t pinpoint why there’s a racial gap, “prior studies have suggested several possible explanations, including culture, religious or spiritual values, and other reasons,” she noted. “Mistrust of the healthcare system due to medical racism and health inequalities in the U.S. may lead more non-white patients to perceive hospice care as ‘giving up’ and motivate them to request more aggressive, life-sustaining interventions.”

Only 20.7% and 21.4% of Black and Hispanic beneficiaries completed advance care planning compared with 57.1% of white beneficiaries. Black and Hispanic decedents were more likely to have written instructions choosing all care possible to prolong life, while more white decedents preferred to limit care in certain situations, withhold treatments, and forgo extensive life-prolonging measures.

In an interview with MedPage Today, Patricia Stone, PhD, RN, director of the Center for Health Policy at the Columbia School of Nursing in New York City, said the inclusion of information about end-of-life preferences is especially useful since it’s often not included in datasets.

She agreed with Lin that it’s still not clear why minority groups tend to resist hospice. “Is it a cultural reason, lack of trust? Or have they not been properly counseled by their care providers? I think it’s the latter.”

Lin pointed out that the study didn’t explore the roles of family members in decision making or what kind of counseling the patients received.

In an accompanying commentary, Norma B. Coe, PhD, and Courtney Lee, MD, MPH, both of the University of Pennsylvania in Philadelphia, noted that it’s possible that minority groups are getting the end-of-life care they want: “if Black and Hispanic patients prefer all interventions at the end of life, then not having advance care planning documents, while perhaps not optimal, may allow patients to receive their preferred default level of care and avoid miscommunication with doctors and staff. If lower levels of hospice use and higher levels of care received in acute care settings are based on patient preferences, then policies need to match treatment to patient preferences, perhaps allowing these differences to continue, and perhaps even to expand in order to best support patients and their families during the end of life.”

For this study, Lin and team included 5,058 Medicare fee-for-service beneficiaries diagnosed with dementia who died between 2000 and 2016. Mean age was 85.5 years, 60% were women, 16% were Black, 7% were Hispanic, and 76.9% were white.

In an adjusted analysis, the researchers found that the following groups were less likely to use hospice: Black decedents (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).

On the other hand, those more likely to use hospice included older decedents (ages 75-84 vs 65-74 years: OR 1.39, 95% CI 1.12-1.72; ages ≥85 vs 65-74 years: OR 1.39, 95% CI 1.13-1.71), women (OR 1.19, 95% CI 1.05-1.35), 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), those with more severe cognitive impairment (OR 1.51, 95% CI 1.02-2.23), and those with more limitations to instrumental activities of daily living (OR 1.07, 95% CI 1.01-1.12).

Lin and colleagues acknowledged that they did not determine causes of death for their cohort, which could affect end-of-life care. Additionally, their study population did not include Asians or other racial/ethnic groups.