End-of-life outcomes after implementation of an embedded, criteria-based palliative care consultation service in inpatient oncology.

Abstract

12121 Background: Specialty palliative care (PC) improves outcomes for patients with advanced cancer, including quality of life and end-of-life (EOL) care utilization. In 2017, the American Society of Clinical Oncology recommended that patients with advanced cancer see specialty PC within 8 weeks of diagnosis. Despite this, inpatient PC referrals exceed outpatient referrals nationwide. Recognizing the importance of earlier PC access, our cancer center implemented a criteria-based PC consultation service in the inpatient oncology setting. We now report pre/post EOL outcomes from this novel care model. Care model: Our center embedded a “Palliative Oncology” consult team on the oncology floor consisting of 1 physician, 1 nurse practitioner (NP), 2 social workers, and 2 spiritual care providers, all with PC specialization. Admitted patients were screened by a floor nurse or PC NP for 2 referral criteria: 1) advanced/metastatic solid cancer, or 2) moderate/severe symptoms. The inpatient oncology team was informed of patients meeting criteria through daily prompting supported by ongoing education. Referrals remained at discretion of the oncology team rather than triggering automatically. Methods: We identified all cancer center decedents with advanced solid cancers who experienced an admission to the oncology floor before (10/1/19-6/30/20) or after (7/1/20-9/30/21) the new model. Six trained abstractors performed chart review. We compared outcomes between patients who died pre- and post-implementation using a t-test (continuous variables) and χ2 test (categorical variables). Results: Of 820 decedents, 186 died pre-intervention and 634 died post-intervention. During the post-intervention period, more decedents saw inpatient PC (72% vs. 59%, p < .001) and outpatient PC (34% vs. 23%, p < .01). Post-intervention decedents had more inpatient PC visits (mean: 11.1 vs. 8.9, p < .05) and had their first PC visit sooner before death (mean: 159 vs. 76 days, p < .001). During the post-intervention period, fewer decedents had ED (41% vs. 52%, p < 0.05), hospital (57% vs. 71%, p < .001), or ICU encounters (17% vs. 25%, p < .01) in the last 30 days of life. For post-intervention decedents, the proportion referred to hospice remained stable (67% vs. 60%, p = .09), but mean hospice length-of-stay increased (36 vs. 22 days, p < .05), as did the proportion who died at home (50% vs. 41%, p < .05). Fewer post-intervention decedents received systemic cancer therapy in the last 14 days of life (5% vs. 9%, p < .05). There were no between-group differences in age at cancer diagnosis, age at death, gender, race/ethnicity, insurance, or primary cancer. Conclusions: A criteria-based “Palliative Oncology” inpatient consult service was associated with earlier, more intense PC and improved EOL quality outcomes, representing a successful effort to enhance PC integration in the inpatient oncology setting.