AI emergency triage systems may replicate clinical stigmas, deprioritizing high-risk patients
Source Summary
Researchers at UT Southwestern Medical Center found that AI-driven emergency triage tools may be vulnerable to the same clinical stigmas present in human assessments, potentially causing harmful biases and incorrect patient prioritization. The study, published in npj Digital Medicine, demonstrates that these algorithmic systems could deprioritize high-risk patients based on stigmatizing information.
Why it matters
AI bias in emergency triage could directly harm patient outcomes by delaying critical care for vulnerable populations already subject to clinical stigmatization.
