Racial Differences in Health Care Transitions and Hospice Use at the End of Life.

Abstract

<b><i>Background:</i></b> Although the fragmentation of end-of-life care has been well documented, previous research has not examined racial and ethnic differences in transitions in care and hospice use at the end of life. <b><i>Design and Subjects:</i></b> Retrospective cohort study among 649,477 Medicare beneficiaries who died between July 2011 and December 2011. <b><i>Measurements:</i></b> Sankey diagrams and heatmaps to visualize the health care transitions across race/ethnic groups. Among hospice enrollees, we examined racial/ethnic differences in hospice use patterns, including length of hospice enrollment and disenrollment rate. <b><i>Results:</i></b> The mean number of care transitions within the last six months of life was 2.9 transitions (standard deviation [SD] = 2.7) for whites, 3.4 transitions (SD = 3.2) for African Americans, 2.8 transitions (SD = 3.0) for Hispanics, and 2.4 transitions (SD = 2.7) for Asian Americans. After adjusting for age and sex, having at least four transitions was significantly more common for African Americans (39.2%; 95% confidence interval [CI]: 38.8-39.6%) compared with whites (32.5%, 95% CI: 32.3-32.6%), and less common among Hispanics (31.2%, 95% CI: 30.4-32.0%), and Asian Americans (26.5%, 95% CI: 25.5-27.5%). Having no care transition was significantly more common for Asian Americans (33.0%, 95% CI: 32.0-34.1%) and Hispanics (28.8%, 95% CI: 28.0-29.6%), compared with African Americans (19.2%, 95% CI: 18.9-19.5%) and whites (18.9%, 95% CI: 18.8-19.0%). Among hospice users, whites, African Americans, and Hispanics had similar length of hospice enrollment, which was significantly longer than that of Asian Americans. Nonwhite patients were significantly more likely than white patients to experience hospice disenrollment. <b><i>Conclusions:</i></b> Racial/ethnic differences in patterns of end-of-life care are marked. Future studies to understand why such patterns exist are warranted.