Many More Palliative Care Consults With Predictive Analytics
Palliative Connect, a system powered by predictive analytics, increased palliative care consultations for seriously ill patients by 74%, according to the authors of a recent study.1
“We were not so surprised that nudging physicians to consider palliative care for seriously ill patients was effective at increasing the number of consults,” reports Katherine Courtright, MD, MSHP, the study’s lead author and assistant professor of medicine at University of Pennsylvania’s Perelman School of Medicine.
More exciting were the apparent positive effects “downstream” from the consult. These included more referrals to home-based palliative care for continuity after discharge, and better documentation of advance care plans.
Palliative Connect uses electronic health record data and machine learning to develop a score on a patient’s likely prognosis over six months. The researchers compared the number of palliative care consults in an eight-week period in a group of 134 admitted patients before and after Palliative Connect was used. The Palliative Connect group received far more consults (85 vs. 22 in the other group). Patients also were seen an average of a day and a half sooner.
“This approach attempts to level the playing field for all seriously ill patients in the hospital,” Courtright says. Ideally, all get the chance to discuss their values, goals, and preferences, not just the ones busy clinicians can identify. “This approach assesses all patients in the same way,” Courtright explains. Then, the system informs clinicians of the highest-risk patients who are likely to benefit from a palliative care consult. Ultimately, it leaves the decision up to their clinician.
Courtright sees two ethical concerns with machine-learning predictive models such as Palliative Connect:
- Models need to be assessed for possible unintended effects on healthcare disparities before they are implemented widely;
- The excitement about the potential of these tools could lead to hasty deployment, before researchers can fully evaluate their effectiveness and safety. “Predictive models on their own do not improve care,” Courtright notes.
The tools need to be linked to a clinical action or intervention, and then tested rigorously. “We need to be ready to say some of them just don’t move the needle on care the way we had hoped or intended, and move on,” Courtright says.
REFERENCE
- Courtright KR, Chivers C, Becker M, et al. Electronic health record mortality prediction model for targeted palliative care among hospitalized medical patients: A pilot quasi-experimental study. J Gen Intern Med 2019;34:1841-1847.
A system powered by predictive analytics increased palliative care consultations for seriously ill patients by 74%, according to the authors of a recent study.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.