Augmented Reality–Guided Decision Support in Simulated Pediatric Cardiac Arrest

Importance: Guideline-adherent management of pediatric in-hospital cardiac arrest (IHCA) remains challenging, and deviations from best practices are common. Augmented-reality (AR)-enabled, role-specific decision support may improve adherence to American Heart Association (AHA) Pediatric Advanced Life Support (PALS) guidance and key performance metrics. Objective: To determine whether an AR-enhanced, role-specific decision support system improves resuscitation performance and adherence to AHA PALS guidelines during simulated pediatric IHCA. Design, Setting, and Participants: This open-label, multicenter, simulation-based randomized clinical trial was conducted from April to May 2025 at 2 tertiary pediatric emergency centers (Geneva, Switzerland and Alberta, Canada). Participants included teams of pediatric nurses and physicians. Intervention: Teams managed a standardized scenario of a 12-minute IHCA due to hyperkalemia (progressing from nonshockable to shockable rhythms) using the AR-enhanced, role-specific decision support system (intervention) or AHA PALS pocket cards (control). Main Outcomes and Measures: The primary outcome was time from recognition of loss of pulse to first epinephrine. Secondary outcomes included adherence to 3- to 5-minute epinephrine dosing intervals, time to first defibrillation, adherence to 2-minute shock and rhythm-check cycles, chest compression fraction, peri-shock pause, medication-dosing accuracy, and user experience and technology acceptance. Results: A total of 54 participants were randomized into 18 teams (18 team leaders [12 female [71%] and 36 nurses [33 female [87%]), with 9 teams (27 participants) in each group. Mean (SD) time to first epinephrine was shorter in the intervention group (97.2 [38.5] vs 113.8 [44.5] seconds; mean difference, -16.6 seconds; 95% CI, -51.3 to 17.0 seconds; P = .40), but this difference was not significant. For subsequent epinephrine, the intervention group improved consistency: mean (SD) deviation from the 4-minute target was 17.2 (32.5) vs 49.7 (40.3) seconds (mean difference, -32.4 seconds; 95% CI, -58.8 to -5.8 seconds; P = .03), with fewer guideline violations (2 of 19 participants [11%] vs 9 of 21 participants [43%]; risk difference, -0.32; 95% CI, -0.55 to -0.05; risk ratio, 0.25; 95% CI, 0.06 to 0.996; P = .03). Time to first defibrillation and adherence to 2-minute cycles were similar between groups. Estimates for chest compression fraction, peri-shock pauses, and medication-dosing accuracy did not suggest meaningful between-group difference. User experience and technology acceptance were favorable. Conclusions and Relevance: In this randomized clinical trial, AR support did not clearly improve time to first epinephrine in simulated pediatric cardiac arrest, with estimates compatible with both benefit and little or no effect. It improved adherence to epinephrine dosing intervals without impairing other performance domains. Trial Registration: ClinicalTrials.gov Identifier: NCT06376643.