Algorithm-Based Palliative Care in Patients With Cancer: A Cluster Randomized Clinical Trial

Area of Science:

• Oncology
• Palliative Care
• Health Informatics

Background:

• Guideline-recommended early specialty palliative care (PC) access is limited in community oncology settings for patients with advanced solid tumors.
• Effective strategies are needed to integrate PC into routine cancer care.

Purpose of the Study:

• To evaluate if algorithm-based defaults within the electronic health record (EHR), incorporating opt-out options and accountable justification, can increase completed PC visits.
• To assess the impact of this intervention on patient-centered outcomes and end-of-life care.

Main Methods:

• A 2-arm cluster randomized clinical trial involving 15 community oncology clinics in Tennessee.
• Patients with advanced lung or gastrointestinal cancer were identified via EHR algorithm.
• Intervention sites used EHR defaults for PC orders with accountable justification for opting out; control sites managed referrals at clinician discretion.

Main Results:

• The intervention group showed a significantly higher rate of completed PC visits (43.9%) compared to the control group (8.3%) (adjusted odds ratio, 8.9).
• Patients in the intervention group received less systemic therapy within 14 days of death (6.5% vs. 16.1%).
• No significant differences were observed in quality of life or feeling heard and understood.

Conclusions:

• Algorithm-based EHR defaults with accountable justification represent a scalable strategy to enhance PC referrals in community oncology.
• This approach effectively increases PC consultations and may reduce intensive end-of-life treatment.