Wilson AI palliative referral

Effect of an Artificial Intelligence Decision Support Tool on Palliative Care Referral in Hospitalized Patients: A Randomized Clinical Trial

Patient / Population Intervention / Exposition Comparaison Critère de jugement

In 1,717 hospitalised patients analysed (of 2,544 randomised) across 12 nursing units at two hospitals, an AI/ML decision-support tool predicting need for palliative care and triggering a consultation recommendation to the medical service increased palliative care consultation compared with usual care.

N
1 717
Schéma
Pragmatic stepped-wedge cluster RCT, 12 nursing units at 2 hospitals, 15 months
Critère
Palliative care consultation note
Pertinence
3
RésultatPalliative care consultation IRR 1.44 (95% CI 1.11–1.92). Exploratory: 60-day readmission OR 0.75 (95% CI 0.57–0.97) and 90-day readmission OR 0.72 (95% CI 0.55–0.93).
Wilson PM, et al. J Pain Symptom Manage. 2023;66(1):24-32. 10.1016/j.jpainsymman.2023.02.317
Discussion et critique

An early pragmatic trial moving ML prediction of palliative need from validation into inpatient workflow, with a readmission signal worth following up. Only 67% of randomised patients were retained for analysis, the readmission findings are exploratory, the primary endpoint is a consultation note rather than patient-centred care, and it ran at two hospitals — tier 3 for these reasons.