NYU Langone AI tool PREVIA predicts PTSD risk from what trauma patients say in the emergency room

An AI tool developed by researchers at NYU Langone Health can predict which emergency-room trauma patients will later develop post-traumatic stress disorder by analysing how they describe what happened to them, according to a study published on November 9 in Nature Communications. The tool, called PREVIA, correctly identified more than 95.4% of at-risk patients across the narratives of 145 patients.

In a head-to-head test, PREVIA flagged all 30 cases from the same group that went on to develop PTSD, while an emergency clinician specialising in PTSD identified 89%. "This is the first study to show that a patient's spontaneous answers to a single, routine care question, 'What brought you to the ED?', carries enough of a psychological signal to forecast their PTSD risk a month later," said senior author Katharina Schultebraucks.

About 43 million injured patients are admitted to US emergency departments each year, and more than a quarter develop PTSD, but only about 28% of emergency departments routinely screen for it. PREVIA combines several AI models, using Llama to emphasise signs of risk and Mistral to highlight protective factors, and is designed to add no time to intake.