The impact of a universal inpatient palliative care consultation protocol for patients with gynecologic cancer: a quality improvement project

Objectives: The American Society of Clinical Oncology (ASCO) recommends early palliative care assessment (PCA) for patients with advanced or recurrent malignancy. We report initial findings of a protocol designed to meet this standard. Methods: We used key stakeholders from a multi-disciplinary team of gynecologic oncologists and palliative care providers to inform an initiative to increase compliance with the ASCO recommendation. As a result, in April of 2020, a quality improvement protocol was initiated whereby a palliative care consult was ordered for all patients with a gynecologic cancer diagnosis admitted to the gynecologic oncology service. Admissions for 23-hour observation and those who did not receive outpatient care from the gynecology oncology team were excluded. Admissions meeting the inclusion criteria 3 months prior to protocol initiation (PPI) were compared to admissions in months 4-6 after protocol initiation (API) to allow for a 3 month wash out period. A retrospective chart review was completed to identify clinicodemographic characteristics for each admission. The primary outcome was completion of inpatient PCA calculated per admission. Secondary outcomes were palliative care and hospice interventions, and advanced care documentation (code status, power of attorney or living will). T-test, Fisher's exact test, relative risk (RR) and exact binomial confidence intervals (CIs) were used. Results: There were 152 admissions, 82 PPI and 70 API. The mean age was 61 years (±13), the majority were White (49%) or Black (39%), had a primary diagnosis of ovarian/primary peritoneal (37%) or uterine cancer (32%) and had advanced/recurrent disease (93%). The majority were medical admissions (n=91, 62%) with 10 (7%) for radiation and 44 (30%) for postoperative care. The demographics of admissions PPI and API are summarized in Table 1. PCA was significantly higher API versus PPI (63% vs. 38%, RR 1.95, 95% CI 1.26-3.02). Additionally, API admissions were more likely to receive follow-up outpatient PCA referrals (20% vs 5%, p=0.005). Death (inpatient or outpatient) was more commonly observed in Black patients compared to White (61% vs 39%, p=0.049). Of the 24 patients who died during the study period, 7 died inpatient. Notably all patients who died inpatient were Black (0% White vs 0% other vs 100% Black p=0.03). Documentation of goals of care discussions, code status, or power of attorney did not change after protocol implementation (RR 1.14, CI 0.90-1.41). Conclusions: This universal palliative care consultation protocol improved the frequency of PCA on an inpatient gynecologic oncology service and increased outpatient PCA referrals, but did not achieve complete compliance. Early observations have also identified a potential disparity in frequency and location of death that warrants further exploration. The next phase of this QI initiative will explore and address barriers to achieving universal PCA.