NYU Langone’s PREVIA Flags PTSD Risk From Emergency Narratives

NYU Langone Health announced on November 9, 2026, that its PREVIA AI ensemble analyzed emergency-department trauma narratives to estimate one-month post-traumatic stress disorder (PTSD) risk, reporting 95.4% sensitivity for the best configuration in a 145-patient study. The experimental screener is designed to identify at-risk patients without adding a lengthy mental-health questionnaire to emergency intake.

The Nature Communications paper, also published November 9, reports 78.5% AUROC and 46.8% precision across three cross-validation folds and calls for validation in larger, independent cohorts. Sensitivity measures the share of later PTSD cases detected, precision measures the share of flagged patients who actually developed PTSD, and AUROC summarizes how well the scores separate the two outcomes. That precision means many positive screens would be false positives, potentially adding follow-up staff time, patient burden and resource-allocation costs, although the next step would typically be a brief assessment rather than invasive intervention.

PREVIA is designed around a practical gap: only about 28% of US emergency departments routinely screen for PTSD, and only two-thirds of those use validated questionnaires, according to the study. Standard assessments compete with immediate medical needs when clinician time is typically around 15–20 minutes per patient. The first month after trauma is a critical identification window because many patients are no longer in contact with the hospital by then.

Senior author Katharina Schultebraucks, co-director of computational psychiatry at NYU Grossman School of Medicine, said: “PREVIA can serve as an AI screener that identifies at-risk patients and explains why, without adding any time to intake in busy emergency rooms.”

From a Bedside Narrative to a Risk Score

Research coordinators asked patients at two New York City level-1 trauma centers to describe the event that brought them to the emergency department. Participants were enrolled within 24 hours of trauma and given up to three minutes to respond while an iPad recorded video. Narratives averaged about 374 words. Whisper transcribed patient speech, and native English-speaking research staff reviewed the transcripts. One-month PTSD outcomes were determined using the Clinician-Administered PTSD Scale for DSM-5, a clinical diagnostic interview.

PREVIA divides each transcript across three layers. In the first, Llama and Mistral independently assess early post-traumatic stress symptoms, reactions during the traumatic event, risk factors and protective factors. DeepSeek reconciles their eight assessments into structured summaries, while the two models’ numerical scores are averaged with equal weighting. A final assessor combines the summaries with age, sex, trauma type and PTSD prevalence information to produce a risk score from 0 to 100 and a short explanation.

The pipeline uses zero-shot prompts, with no parameter training or fine-tuning on the study sample. Cross-validation was used to optimize and evaluate the classification threshold.

Performance Against Simpler Models and One Specialist

The paper reports broadly comparable overall performance between PREVIA and single zero-shot configurations of O3, Llama and Mistral on many metrics, with PREVIA’s clearest advantage in sensitivity. For subthreshold PTSD, meaning clinically meaningful symptoms below the full diagnostic threshold, PREVIA reached 89.7% sensitivity and a 69.4% F1 score, a combined measure of precision and sensitivity.

The study reports a separate comparison of 30 randomly selected narratives, stratified to include nine PTSD cases, in which PREVIA recorded 100% sensitivity and 84.4% AUROC. One emergency clinician recorded 88.9% sensitivity and 83.9% AUROC; both recorded 50% precision. The clinician had more than 24 years of emergency-medicine experience and specialized PTSD research experience. The single-rater design leaves open whether the comparison generalizes across experience levels and clinical settings.

Validation and Deployment Requirements

Participants were English-speaking, medically stable adults aged 18–80, and women accounted for 31% of the sample. Medically unstable patients were excluded, potentially omitting some patients at highest risk of chronic PTSD. The authors identify testing in other languages, longer-term PTSD trajectories, comorbidities such as depression, and integration with clinician judgment and emergency workflows as future work.

The authors propose a clinical workflow that pairs a brief audio recording with those demographic and trauma variables from the electronic health record, then returns a risk score and explanation without additional clinician input. The study did not establish that such screening improves patient outcomes. PREVIA ran in a HIPAA-compliant high-performance-computing environment on NVIDIA A100 GPUs. Self-hosting supports control over patient data, but adoption depends on hospitals having suitable computing and governance infrastructure.

The public PREVIA repository provides the complete pipeline and prompts for open-weight models that run locally, plus five fabricated narratives for an end-to-end demonstration. Identifiable clinical narratives and psychiatric assessments remain restricted for privacy and consent reasons, so the demo can test operation but cannot reproduce the paper’s quantitative results.

PREVIA’s immediate contribution is an independently testable screening approach built from a routine emergency encounter and clinician-assessed outcomes. Its clinical value now hinges on independent validation: whether larger cohorts preserve its sensitivity, whether services can manage the false positives, and whether hospitals can incorporate recording, follow-up and secure computing without disrupting emergency care.