Emergency room triage shows measurable bias by race, age, and insurance — first time tested on real hospital data

Researchers analyzed real hospital triage decisions across 100,000+ emergency room visits and found that wait times, retreatment rates, and clinical pathways differ significantly based on patient age, race, gender, language, and insurance status. This is the first time anyone has measured fairness in triage using actual hospital event logs instead of simulations or abstract fairness metrics.

For years, hospitals have assumed their triage systems are neutral — fast decisions under pressure, applied equally. This study breaks that assumption by showing that the decisions themselves vary systematically by patient characteristics in ways that correlate with justice theory concepts like procedural fairness and distribution. The practical consequence is immediate: hospitals now have a measurement method (process mining applied to event logs) that can show them exactly where their triage is unequal, not in theory but in their own data. Until now, fairness in emergency care was either not measured or measured using abstract algorithms divorced from what actually happens to patients.