Concordance between an artificial intelligence self-triage programme and physical triage - Summary - MDSpire
Coming Soon: Introducing MDSpire News. Learn more
Conexiant’s news site is now MDSpire News. Learn more

Comparison of an AI-Based Self-Triage System with Traditional Physical Triage Methods

  • By

  • Maaike Wempe

  • Frits Holleman

  • Michiel Schinkel

  • Michiel Gorzeman

  • October 1, 2026

Share

Objective:

To assess whether AI self-triage can correspond with urgency scores assigned by a conventional triage method in emergency department triage.

Approach:
  • Classic Physical Triage: Utilized the Dutch NTS for urgency classifications, ranging from U0 (resuscitation required) to U5 (no treatment needed on the same day).
  • AI Triage Programme: Employed Symptomate, a medical symptom checker that uses AI and machine learning to assess urgency and recommend care pathways.
  • Study Design: Conducted a prospective, observational, comparative study at OLVG hospital, including adults who underwent physical triage.
Key Findings:
  • AI self-triage is currently unsuitable for widespread implementation.
  • AI triage cannot yet substitute for physical triage methods.
Limitations:
  • Limited number of studies demonstrating conclusive benefits from AI triage.
  • Potential biases in patient selection and data collection.

Original Source(s)

Related Content