Bystander AED effectiveness despite low utilisation in OHCA

OBJECTIVES Observational estimates of bystander automated external defibrillator (AED) effectiveness are inconsistent in settings with mature public-access defibrillation infrastructure but low utilisation, partly because the conventional reference group pools emergency medical services (EMS)-shocked with non-defibrillated patients. We estimated bystander AED effectiveness against a partitioned no-defibrillation reference, tested effect modification by location, and quantified the deployment–utilisation gap.METHODS Retrospective analysis of a prospectively collected multicentre out-of-hospital cardiac arrest (OHCA) registry (Korea, October 2015–June 2025; 23,918 adults of presumed medical aetiology). A three-tier prehospital defibrillation-provider variable was defined: bystander AED applied (n = 383); EMS-delivered shock without preceding bystander AED (n = 5,975); and no prehospital defibrillation (reference; n = 17,560). Primary outcomes were survival to discharge and good neurological outcome (Cerebral Performance Category 1–2), analysed by multivariable logistic regression with propensity-score, inverse-probability-weighting and location-stratified secondary analyses.RESULTS Bystander AED application was associated with improved survival (adjusted odds ratio [aOR] 1.60; 95% CI 1.19–2.14) and good neurological outcome (2.31; 1.65–3.23), comparable to that of EMS-delivered shock; the pooled-control specification attenuated estimates to non-significance (survival 1.09; neurological 1.14). Shock-delivered applications significantly exceeded EMS-delivered shock for neurological outcome (aOR 1.43; P = 0.03). Effectiveness did not differ by location (interaction P ≥ 0.15). Bystander AED was applied in 8.5% of Public-area versus 1.9% of Non-public AED-opportunity cases — a 4.5-fold gap.CONCLUSIONS Bystander AED was associated with improved outcomes when applied, comparable to or exceeding EMS-delivered shock; the dominant constraint was the deployment–utilisation gap, greatest in residential settings. Residential rapid-activation infrastructure warrants policy priority alongside, not instead of, continued device placement.