Background: Timely integration of palliative care (PC) in the management of patients with advanced cancer is a quality benchmark in oncology. PC is often underutilized due to delayed identification and referral of appropriate patients. Oncology nurses and patient care navigators (ONPCNs) can play a critical role in the timely referral of patients with advanced cancer to PC services. Jvion Inc has developed a prescriptive analytics solution, the Jvion CORE, that combines AI algorithms with machine learning techniques to identify patients with a tendency for poor outcomes. Based on its analyses, the Jvion CORE generates patient-specific, dynamic, and actionable insights without the need for additional documentation within the electronic health record (EHR). The appropriate utilization and initiation of the downstream workflow are critical to the success of these insights. Rainier Hematology/Oncology of Northwest Medical Specialties (NWMS) utilized ONPCNs to collect and disseminate this information at their practice.
Objectives: To apply the Jvion CORE insights to identify patients at high risk for mortality within the next 30 days to increase PC utilization, as measured by increasing referrals to PC services and enrollments to hospice.
Methods: Patients are scored weekly using the Jvion CORE PC product. The AI tool risk-stratifies patients and generates recommendations for the care team. Patients identified at risk for mortality in the next 30 days are flagged. The care coordinator reviews flagged patients, EHR charts, risk factors, and recommended interventions for high/medium-risk patients and considers additional needs or barriers for the patient. Barriers are documented in the patient’s care plan, and a supportive care visit is scheduled with an advanced practice provider (APP). The APP consults with the care team, including the ONPCNs, and proceeds with PC referral as deemed clinically appropriate. The ONPCNs will address any practical barriers to care as appropriate. The average monthly rates of PC consults and hospice referrals were calculated for 5 months before and 17 months after the Jvion CORE deployment at NWMS.
Results: NWMS has 21 providers managing 4329 unique patients per month; additionally, 1 breast cancer navigator, general navigator, and patient care coordinator oversee the Jvion function. The rate of PC consults increased by almost 70% in the index period following AI tool deployment. There was an 8-fold increase in hospice referrals. ONPCNs found the workload of retrieving the information from the AI portal feasible and user-friendly. There was good uptake with practice providers. After eliminating the first 6 months of Jvion CORE deployment to account for the user learning curve, the mean rates of monthly PC consults nearly doubled from baseline, and hospice referrals rose 12-fold in months 7 to 17 after Jvion CORE deployment.
Conclusions: The deployment of this novel AI solution was demonstrated to be feasible and effective at generating actionable insights. Incorporating these AI-driven insights into the practice’s workflow by ONPCNs improved the care team’s decision-making for whether and when a patient should be referred to PC or hospice services for end-of-life care. Larger studies are needed to confirm the value of AI for the management of cancer patients at end of life.