Prehospital Facial Recognition Technology: Randomized Controlled Simulation Study.
The western journal of emergency medicine · 2026-10-29 · Randomised trial
Abstract
INTRODUCTION: The risk of medical errors increases for the approximately 689,000 unidentified, noncommunicative patients who present to emergency departments annually. The concomitant rise in hospital admissions, lengths of stay, and mortality will likely persist unless approaches to identifying noncommunicative patients are developed and adopted. Facial recognition technology could provide a solution where scanning the face reliably provides access to identification and medical history of patients who self-enroll in the application database. Our objective was to compare paramedics' ability to identify noncommunicative patients and make related clinical decisions using a cloud-based facial-recognition application vs standard practice. METHODS: We conducted a double-blind, randomized controlled study with 90 participating firefighterparamedics in a simulated patient encounter. Our primary outcome was patient identification. Secondary outcomes included identification of key medical information, appropriateness of treatment decisions, and transport to the patient's preferred hospital. We uploaded facial photographs and simulated medical profiles to a secure cloud-based app. Participants used a tablet to scan the patient's face (standardized patient or actor). Successful matches returned identifying information, medical history, emergency contacts, preferred hospital, and primary care physician contact information. The 48 experimental participants had access to this patient information, while the 42 controls did not, even if they scanned the face. Standardized patients (SP) portrayed a noncommunicative urosepsis patient. Participants reported their findings to the facilitators and evaluated the technology on a 5-point Likert scale. RESULTS: We analyzed 90 study participants (48 experimental, 42 control), of whom 40/48 (83.3%) experimental group identified the SP compared with 9/42 (21.4%) control group (OR [odds ratio], 13.87; 95% CI, 5.03-38.25; P < .001). Identification of important medical information of the SP by the study participants occurred in 30/48 (62.5%) in the experimental group versus 4/42 (9.5%) in control (OR, 15.83; 95% CI, 4.84-51.76; P < .001). Correct transport decision to the preferred hospital occurred in 27/48 (56.3%) experimental participants compared with 10/42 (23.8%) controls (OR, 4.11; 95% CI, 1.65-10.23; P = .02). Overall, 80% of participants reported that the technology would benefit their day-to-day clinical work. CONCLUSION: The use of a cloud-based facial-recognition application significantly improved firefighter-paramedics' ability to identify noncommunicative patients and access critical medical information, with an appropriate transport decision to the patient's preferred hospital.
Abstract © the authors or publisher, via PubMed. Reuse and removal
Tags
Fire service & police · Decision support tools · Paramedics & EMTs
Related articles
- Implementation and Evaluation of the Stroke Triage App Among Paramedics in Prehospital Emergency Care: Sequential Mixed Methods Study. (J Med Internet Res, 2026)
- Implementation and use of computerised clinical decision support (CCDS) in emergency pre-hospital care: a qualitative study of paramedic views and experience using Strong Structuration Theory. (Implement Sci, 2018)
- "Oh, Another Overdose, for the Love of Pete": First Responder Perspectives on Overdose Response Technology. (West J Emerg Med, 2025)
- The Effects of Integrated IT Support on the Prehospital Stroke Process: Results from a Realistic Experiment. (J Healthc Inform Res, 2019)
- Accuracy of Commercial Large Language Model (ChatGPT) to Predict the Diagnosis for Prehospital Patients Suitable for Ambulance Transport Decisions: Diagnostic Accuracy Study. (Prehosp Emerg Care, 2025)
- Intuitive use and usability of ventricular assist device peripheral components in simulated emergency conditions. (Artif Organs, 2011)