Racial Disparities in End-of-Life Care Planning and Hospital Utilization: A Single Center Retrospective Study

Abstract

Introduction-Some prior studies suggest patients with hematologic malignancies receive more aggressive end-of-life care when compared to patients with solid tumor malignancies. Hematologic malignancies differ from solid tumors because of the continued potential for cure even in advanced disease, and potential difficulty identifying the terminal portion of a patient's illness. Racial/ethnic minorities are reported to have lower rates of hospice care, advanced directive use, and palliative care utilization. We studied differences in hospital utilization patterns and documentation of advance care planning between solid tumor and hematologic malignancy patients. In the subgroup of patients with hematologic malignancy, we also examined differences in these outcomes associated with racial/ethnic minority status.Methods-We conducted a retrospective cohort study of 9,469 patients with a diagnosis of cancer who received care at University of Washington (UW) Medicine and died between 2010 and 2015. Administrative data were available for the following events during the last 30 days of life: emergency department use, hospitalizations, and intensive care unit (ICU) utilization. We also examined death in a hospital and any documentation of advance directives (AD) in the electronic health record. We regressed each outcome on the binary predictor, adjusting for confounders (taken from a pool of potential confounders: age at death, patient gender, racial/ethnic minority status, level of education, insurance type, attributed facility, and number of Dartmouth Atlas chronic conditions). A variable was considered a confounder if its addition to the bivariate model changed the coefficient for the predictor by at least 10%. Binary outcomes were modeled with logistic regression. For count outcomes, we included only patients who had 1 or more days of the relevant type of care and modeled the remaining cases with negative binomial regression. All estimation was done with restricted maximum likelihood. Statistical significance was p <0.05.Results-In the last 30 days of life, decedents with hematologic cancer were significantly more likely to have aggressive hospital-based care, as measured by receipt of inpatient care, hospitalization for 14 or more days, multiple hospital admissions, and more days of hospital care, once admitted, than were those with solid tumor malignancies. They were also significantly more likely to have received ICU care, and to have spent more time in the ICU, once admitted, in the last 30 days of life. Finally, patients with hematologic malignancies were more likely to have died in a hospital rather than in other locations and more likely to have had documentation of AD in their electronic record. Among patients with hematologic malignancies, racial/ethnic minorities were less likely than white non-Hispanics to have documentation of AD and more likely to have 2+ emergency department visits or 14+ days of inpatient care, and had more days of inpatient care and ICU care, once admitted, in the last 30 days of lifeConclusions-Patients with hematologic malignancies received more aggressive care at the end of life as measured by hospital utilization, despite having more documentation of AD than patients with solid tumor malignancy. Racial/ethnic minorities with hematologic malignancies had lower rates of AD documentation and received even more aggressive care than their white counterparts. Although these are administrative data, they suggest opportunities to improve end-of-life care of patients with hematologic malignancies, particularly racial/ethnic minorities. [Display omitted] DisclosuresLee:Mallinckrodt: Honoraria; Amgen: Other: One-time advisory board member; Bristol-Myers-Squibb: Other: One-time advisory board member; Kadmon: Other: One-time advisory board member.