AI triage helps spot life-threatening illnesses before patients realise the danger.

When people experience early symptoms of serious illnesses, they often underestimate the risk. A recent study explored whether an AI-powered virtual triage platform—available online and via mobile app—could help identify life-threatening conditions and nudge users toward the right level of care. The analysis covered 3 million virtual triage sessions over 16 months, focusing on five acute conditions that demand emergency treatment: heart attack, stroke, asthma attack, pneumonia, and pulmonary embolism.

The findings were striking. Among 12,101 users whose symptoms warranted emergency care, 38.5% initially planned no professional care at all, and 61.5% had no intention of seeking emergency treatment. Even after adjusting for the system’s built-in safety over-triage, a third still planned no professional care and over half intended to avoid the emergency department. Younger adults (18–44) were least likely to plan emergency care, and women were slightly less likely than men to indicate urgent intent.

By analysing symptom patterns and prompting for immediate action, the virtual triage engine offers a way to bridge the gap between patient perception and clinical risk. If integrated into health-system “digital front doors,” AI triage could help accelerate emergency referrals, reduce preventable deaths, and improve use of hospital resources. The authors call for further clinical studies to confirm whether these early warnings can reliably change real-world patient behaviour.

Reference:
 Gellert GA, Kabat-Karabon A, Gellert GL, et al. The potential of virtual triage AI to improve early detection, care acuity alignment, and emergent care referral of life-threatening conditions. Front Public Health. 2024;12:1362246. https://doi.org/10.3389/fpubh.2024.1362246