Mayo Clinic and Bayesian Health announced on 20 May 2026 a co-developed AI system designed to identify hospitalised patients with serious illness who may benefit from earlier palliative-care consultation. In a randomised clinical trial conducted at Mayo Clinic, an earlier version of the system was associated with a 44% increase in timely palliative-care referrals, a 25% reduction in 60-day readmissions, and a 28% reduction in 90-day readmissions, alongside improvements in patient-reported quality of life.

The motivation cited by the developers is the gap between the prevalence of serious illness in the readmitted-patient population and the rate of palliative consultation. Approximately one-third of all hospital readmissions involve patients with serious illness; fewer than half of those patients receive a palliative-care consultation during their initial admission. The tool runs continuously against the electronic health record and surfaces candidate patients with the contextual information clinicians need to initiate the goals-of-care conversation, rather than displacing the conversation itself.

Two methodological points are worth noting. First, the RCT findings are from an earlier version of the system, not the version Mayo and Bayesian are now jointly deploying; the read-across to the productised system depends on assumptions about feature parity. Second, "timely" referral is the developer-defined endpoint, and whether earlier referrals lead to clinically meaningful improvements in symptom control, hospice transition, and patient-family experience is the longer-horizon question that the announcement does not address.