Diagnostic performance of a video-assisted cardiac arrest recognition protocol by emergency medical dispatcher in out-of-hospital cardiac arrest.
Resuscitation · 2025-12-22 · Method & validation
Abstract
INTRODUCTION: Cardiac arrest recognition by emergency medical dispatch (EMD) is essential for initiation of telephone CPR that could lead to bystander CPR before emergency medical service (EMS) arrival. Video-assisted cardiac arrest recognition protocol was developed and implemented for better recognition of cardiac arrest suspected calls that recognition was uncertain under conventional audio-call dispatch. We tested the performance of video-assisted protocol after pilot implementation in a metropolitan dispatch center. METHODS: All emergency medical dispatch calls received by Seoul emergency dispatch center that used video-assisted cardiac arrest recognition protocol from June 2020 to December 2020 were enrolled. The primary outcome was the presence of cardiac arrest upon EMS arrival evaluated by EMS providers arrived at the scene. The predictive performance measure of video-assisted protocol was calculated including sensitivity and specificity with 95 % confidence intervals (CIs). RESULT: During the study period, video-assisted protocol was used in 115 emergency calls that dispatchers were uncertain based on the initial audio call. In 15 cases the dispatcher additionally recognized the case as cardiac arrest under video-assistance. Of recognized cases, 14 (93.3 %) cases were confirmed as cardiac arrest upon EMS arrival. The sensitivity (95 % CI) and the specificity (95 % CI) of the video-assisted protocol were 66.7 % (43.0-85.4) and 98.9 % (94.2-100) respectively. CONCLUSION: Use of video-assisted cardiac arrest recognition protocol helped recognize additional cardiac arrest cases that could not be recognized with the audio-call in the dispatch center. Technological and operational improvement should be continuously studied for better cardiac arrest recognition using video-assisted protocol during emergency call taking.