End-of-life (EOL) care in medical oncology fellows' clinics.

Abstract

75 Background: Medical oncology fellowship training presents a critical time for developing skills necessary for discussing end of life (EOL) care preferences with patients and delivering EOL care. There is little data exploring the delivery of EOL care by fellows in practice. We describe our fellows’ experience and identify areas to improve delivery of EOL care and communication. Methods: We retrospectively reviewed electronic medical records (EMR) of patients at our center with advanced (metastatic or recurrent) solid tumors who received care in fellows’ clinics and died between 7/1/2016 and 6/7/2017. We used Fisher’s exact test to determine associations between variables. Results: We included 103 patients from 46 distinct fellow clinics covered by 16 fellows. 54 (52%) were male; median age was 70 (37-94) years. 42 (40%) had a gastrointestinal malignancy, 31 (30%) had lung cancer. Median duration of advanced cancer was 12 (0.5-120) months. Patients received a median of 2 (1-10) lines of chemotherapy. 10 (10%) received chemotherapy within 14 days of death. 10 (10%) used outpatient palliative care (PC). 48 (47%) had EOL/hospice discussions in the outpatient clinic, while 10 (10%) had code status documented in a clinic encounter, and 20 (19%) had a copy of an advanced directive in the EMR. Median time from EOL discussion to death was 30 days (4-171). 69 (67%) enrolled in hospice and median length of stay on hospice was 18 days (1-142). 58 (56%) were hospitalized and 13 (13%) admitted to the ICU within 30 days of death. Of patients not enrolled in hospice, 8 (8%) died at home without hospice, 23 (22%) died in a hospital or ICU, 3 (3%) died in other or unknown places. Patients who had EOL discussions in clinic were more likely to enter hospice (p = .006) and less likely to be hospitalized (p = 0.016) within 30 days of death. Conclusions: Fellows at our center have significant exposure to outpatient EOL discussions and delivery of EOL care. EOL discussions in clinic were associated with decreased rates of health care utilization within 30 days of death. There were low rates of outpatient code status discussions and documentation of advanced directives, and few patients received outpatient PC, which represent potential areas to enhance fellow training and improve patient EOL care.