An augmented intelligence (AI) tool can identify patients with advanced cancer who have a high or medium risk of short-term mortality, according to research published in JCO Oncology Practice.

The tool prompted significant increases in palliative care and hospice referrals at a large hematology-oncology practice, according to researchers.

The tool, Jvion CORE, uses a “continuously learning eigen-based n-dimensional space environment to determine the most likely trajectory for an individual,” the researchers explained. “That trajectory is used to determine an individual’s risk for mortality by understanding the likelihood that the trajectory intersects with high-risk areas within the eigen space.”

Continue Reading

In a retrospective study, the researchers evaluated how deploying the tool impacted palliative care and hospice referrals at a community hematology-oncology practice in the Pacific Northwest. The practice had 6 locations and 21 providers managing an average of 4329 unique patients per month.  

Between June 2018 and October 2019, the tool was used to screen 28,246 patients. It identified 886 patients as having a medium or high risk for 30-day mortality.

Among the at-risk patients, the most common cancer was lung (24.9%), followed by breast (22.1%) and small intestine or colorectal cancer (11.9%).

The average monthly rates of palliative care and hospice referrals were calculated 5 months before tool deployment, which served as a baseline, and 17 months after the tool was deployed. 

The mean rate of palliative care consults was 17.3 per 1000 patients per month (PPM) prior to deployment, and this increased to 29.1 per 1000 PPM post-deployment. The mean rate of hospice referrals increased from 0.2 to 1.6 per 1000 PPM from pre- to post-deployment.

When the researchers eliminated the first 6 months after tool deployment to account for a user learning curve, the mean rate of palliative care consults increased to 33.0 per 1000 PPM post-deployment, and the mean rate of hospice referrals increased to 2.4 per 1000 PPM.

The researchers found the 30-day mortality algorithm to be accurate, with area under the receiver operator characteristic curve values of 0.93 at 30 days and 0.92 at 90 days. Among patients identified as medium- or high-risk, 10.3% had died at 30 days, and 16.4% had died at 90 days.

“Deployment of an AI tool at a hematology-oncology practice was found to be feasible for identifying patients at high or medium risk for short-term mortality,” the researchers concluded. “Insights generated by the tool drove clinical practice changes, resulting in significant increases in PC [palliative care] and hospice referrals.”

Disclosures: Some study authors are employed by biotech, pharmaceutical, and/or device companies. Please see the original reference for the authors’ affiliations.


Gajra A, Zettler ME, Miller KA. Impact of augmented intelligence on utilization of palliative care services in a real-world oncology setting. JCO Oncol Pract. Published online September 10, 2021. doi:10.1200/OP.21.00179