The feasibility of an inpatient intervention to improve end-of-life care in gynecologic oncology patients.
Puechl, A; Lim, S; Truong, T; Havrilesky, L; Davidson, BA
Published in: Journal of Clinical Oncology
69 Background: Unscheduled hospitalizations in patients with advanced cancer predicts < 6 month median survival. Patients with cancer hospitalized near the end of life (EOL) often receive aggressive treatments without survival advantages and have poor quality-of-life. EOL discussions and palliative care (PC) services reduce aggressive care. Methods: An intervention was designed to decrease aggressive EOL care in Gynecologic Oncology (GO) inpatients at a high risk for death within 6 months. Inclusion criteria were advanced gynecologic cancer and 1 of the following: bowel obstruction, failure to thrive, dehydration, cancer-related pain, malignant pleural effusion or ascites, or PCN management. Patients were identified by the care team on admission. An independent auditor assessed if patients were appropriately identified. Eligible patients received a 3-part high risk Gynecologic Oncology (HR-GO) bundle: 1) PC consult, 2) goals of care (GOC) conversation ≤ 48 hours of admission, and 3) PC follow-up within 2 weeks of discharge. Data was collected through chart review. Barriers to bundle implementation were identified. Results: 83 eligible patients were admitted between 8/2017-4/2018. 38 patients (46%) had multiple hospital admissions. 41 patients (49%) received the entire bundle during at least one hospital admission. During their 1st hospital admission, 30% of patients received all 3 parts of the HR-GO bundle. Among those who did not receive the entire bundle, 43% had two components missing. PC follow-up was the most commonly missed component, followed by a GOC conversation within 48 hours of admission. As identified by the auditor, twenty percent of eligible patients were not placed into the study cohort during their first hospital admission. Two barriers to bundle implementation were identified:1) the dependence on care team members to identify high-risk patients and 2) the avoidance of needed GOC discussions when patient’s 1° attending was unavailable. Conclusions: The implementation of an inpatient intervention to decrease aggressive EOL care in GO patients is feasible however barriers must be addressed. Currently work on an automated system to identify patients for intervention is ongoing.
