Video-vs audio-instructed dispatcher-assisted CPR and outcomes after out-of-hospital cardiac arrest: a nationwide registry-based cohort study.
Resuscitation · 2026-02-06 · Registry study
Abstract
PURPOSE: Dispatcher-assisted cardiopulmonary resuscitation (DA-CPR) may enhance bystander cardiopulmonary resuscitation (CPR) effectiveness in out-of-hospital cardiac arrest (OHCA). We evaluated whether video-based DA-CPR (video group [VG]) improves neurological outcomes and CPR time metrics compared with audio-based DA-CPR (audio group [AG]). METHODS: We conducted a retrospective, nationwide cohort study of emergency medical service (EMS)-treated OHCA in Korea from 2019 to 2021 using the Korean Out-of-Hospital Cardiac Arrest Registry (KOHCAR). The setting included all 18 provincial EMS System. The exposure was the DA-CPR modality (VG vs. AG). The primary outcome was good neurological status at discharge (Cerebral Performance Category 1-2). Secondary outcomes were prehospital return of spontaneous circulation (ROSC), survival to discharge, and CPR time metrics (detection time interval [DTI] ≤90 s, instruction time interval [ITI] ≤90 s, CPR start time interval [CTI] ≤150 s). Associations were estimated using multivariate logistic regression and propensity score-matching (PSM) analyses. RESULTS: Among 35,471 patients (AG 32,973 [93.0%]; VG 2,498 [7.0%]), the VG showed higher good neurological outcome, prehospital ROSC, and survival to discharge than the AG, with adjusted odds ratios (aORs) (95% confidence intervals) of 1.64 (1.39-1.92), 1.29 (1.18-1.41), and 1.50 (1.33-1.69), respectively. However, the proportions meeting ITIs ≤90 s and CTIs ≤150 s were lower in the VG (aOR 0.91 [0.84-0.99] and 0.89 [0.81-0.97]). These findings remained robust in the PSM analyses. CONCLUSION: DA-CPR in the VG was associated with associated with higher odd of favorable neurological outcomes and survival in patients with OHCA, while instruction and CPR initiation times were modestly longer. Optimizing protocols to reduce video transition workflows and incorporating dispatch-center and technical variables should be priorities for future prospective studies.