A novel referral system to optimize timing to specialized palliative care in patients with advanced cancer: an interventional study.

Abstract

Background: Early integration of specialist palliative care (PC) improves outcomes for patients with advanced cancer, yet implementation of needs-based referral remains inconsistent in routine oncology practice. The Palliative Care Referral System (PCRS) is a digital clinical decision-support tool designed to support referral using patient-reported symptoms, prognostic indicators, and clinical complexity. We evaluated the impact of PCRS implementation on referral patterns, patient-centered outcomes, healthcare utilization, and end-of-life (EoL) care.

Patients and methods: In this single-center prospective quasi-experimental pre-post implementation study, 380 patients with advanced solid tumors were enrolled before (Pre-PCRS, n = 190) and after (Post-PCRS, n = 190) PCRS implementation. During the Post-PCRS phase, oncologists completed the PCRS at each outpatient visit. Patient and caregiver satisfaction, quality of life, healthcare utilization, and EoL indicators were assessed monthly for six months. Time to first PC referral was analyzed using competing-risk methods.

Results: Overall 24-week referral rates were similar in the Post-PCRS and Pre-PCRS cohorts (21% vs 20%). However, referral timing and patterns changed across tumor types, with earlier referrals in breast and prostate cancer after PCRS implementation. The proportion of patients with at least one hospitalization was numerically lower after implementation (16% vs 27%; OR, 0.53; 95% CI, 0.32-0.88; p = 0.017), but the difference was not significant after correction for multiple comparisons (p = 0.085). No significant differences were observed in patient or caregiver satisfaction, quality of life, emergency department visits, oncology visits, or PC visits. Indicators of aggressive EoL care were not significantly different between cohorts.

Conclusions: PCRS implementation did not increase the overall referral rate but was associated with changes in referral timing and patterns across tumor types. Whether these changes reflect more appropriate, needs-based care and translate into lower healthcare utilization requires confirmation in larger studies.