How U.S. Doctors Die: A Cohort Study of Healthcare Use at the End of Life

• Published in: Journal of the American Geriatrics Society (JAGS) Vol. 64; no. 5; pp. 1061 - 1067
• Main Authors: Matlock, Daniel D., Yamashita, Traci E., Min, Sung-Joon, Smith, Alexander K., Kelley, Amy S., M. Fischer, Stacy
• Format: Journal Article
• Language: English
• Published: United States Blackwell Publishing Ltd 01.05.2016
• Subjects: Aged; end of life; Female; hospice; Hospices - statistics & numerical data; Hospitalization - statistics & numerical data; Humans; Intensive Care Units - statistics & numerical data; Male; Medicare; Medicare Part A; Physicians; Retrospective Studies; Socioeconomic Factors; Terminal Care - statistics & numerical data; United States - epidemiology; end of life; physicians; Medicare; hospice
• ISSN: 0002-8614; 1532-5415; 1532-5415
• DOI: 10.1111/jgs.14112

Objectives To compare healthcare use in the last months of life between physicians and nonphysicians in the United States. Design A retrospective observational cohort study. Setting United States. Participants Fee‐for‐service Medicare beneficiaries: decedent physicians (n = 9,947) and a random sample of Medicare decedents (n = 192,006). Measurements Medicare Part A claims data from 2008 to 2010 were used to measure days in the hospital and proportion using hospice in the last 6 months of life as primary outcome measures adjusted for sociodemographic characteristics and regional variations in health care. Results Inpatient hospital use in the last 6 months of life was no different between physicians and nonphysicians, although more physicians used hospice and for longer (using the hospital: odds ratio (OR) = 0.98, 95% confidence interval (CI) = 0.93–1.04; hospital days: mean difference 0.26, P = .14); dying in the hospital: OR = 0.99, 95% CI = 0.95–1.04; intensive care unit (ICU) or critical care unit (CCU) days: mean difference 0.35 more days for physicians, P < .001); using hospice: OR = 1.23, 95% CI = 1.18–1.29; number of days in hospice: mean difference 2.06, P < .001). Conclusion This retrospective, observational study is subject to unmeasured confounders and variation in coding practices, but it provides preliminary evidence of actual use. U.S. physicians were more likely to use hospice and ICU‐ or CCU‐level care. Hospitalization rates were similar.