Rochester, Minn.-based Mayo Clinic is using AI to help identify hospitalized patients who may benefit from palliative care earlier in their stay — part of a broader effort to expand access to supportive care while reducing avoidable readmissions.
The initiative comes as health systems continue facing growing demand for palliative care alongside workforce shortages and limited specialist capacity. Roughly one-third of readmissions involve patients with serious illness, yet fewer than half receive palliative care consultations.
In a randomized clinical trial at Mayo Clinic, the tool showed what is possible when that gap closes: a 44% increase in timely palliative care referrals, a 25% reduction in 60-day readmissions and a 28% reduction in 90-day readmissions, along with improved patient quality of life.
Jacob Strand, MD, chair of palliative care at Mayo Clinic, said access barriers have historically stemmed from two intertwined problems: front-line clinicians struggling to identify which patients would benefit most, and a national palliative care workforce too small to serve every eligible patient.
“It’s not because clinicians aren’t being attentive to patient needs,” Dr. Strand told Becker’s. “It relates to busy front-line clinicians trying to manage a lot of things in the hospital. Sometimes those palliative care needs aren’t always readily visible.”
The tool, built on Bayesian Health’s clinical intelligence platform and integrated into Mayo Clinic’s EHR, gives palliative care teams a real-time, hospital-wide view of patients flagged for unmet needs — what Dr. Strand described as a “heat map” of where patients are triggering high-risk signals. Crucially, the model flags based on patient-centered signals — pain, caregiver stress, psychosocial needs, goals-of-care conversations — rather than disease-specific criteria alone, allowing clinicians to “get the right resource to the right patient at the right time in their hospital course.”
Dr. Strand was direct about what separates a useful AI tool from one that gets ignored: workflow integration. To address this, the team built a human-in-the-loop layer so a palliative care team member monitors the model’s output in real time, adjusting for staffing and census pressures. On the bedside end, the tool surfaces relevant patient information and pre-stages a referral order, reducing clicks without removing physician decision authority.
The trial also surfaced something Dr. Strand did not fully anticipate: how much relationship building between palliative care and front-line teams mattered to the tool’s effectiveness. When the model surfaced a recommendation without that groundwork, clinicians were skeptical.
“They like to know: What goes into this, and why is it going to be beneficial to my patient?” he said.
When that trust is in place, the impact reaches patients directly. Dr. Strand said the tool has helped re-establish continuity for patients with existing outpatient palliative relationships who are readmitted, allowing teams to reconnect quickly. “They know that their team who followed them before is going to stay right there with them,” he said.
Dr. Strand also sees the tool as a potential lever for health systems with smaller specialty teams, combining governance and oversight with data that makes the case for program growth.
“Palliative care teams are going to need this type of information to show decision-makers where the unmet need exists in their own system, and then develop solutions to fit that,” he said. “A lot of us in palliative care worry about losing the humanity in medicine the more we automate. What’s been powerful about this work is that it feels like one way to bring a lot of that humanity back to the bedside — where it belongs.”
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
