Abstract
Video-assisted dispatcher support in cardiopulmonary resuscitation (VA-CPR) has shown beneficial effects on CPR quality. This study examines the influence of professional background on the ability to identify typical errors in CPR performance using video-based assessments. Within this simulation-based observational study, 61 participants (31 EMS personnel, 30 emergency physicians) evaluated nine video sequences showing simulated CPR or ventilation. Participants were grouped by profession, not randomly. The primary endpoint was the correct identification of expert-defined errors in the presented videos, analyzed in relation to professional background. Evaluation accuracy for CPR and ventilation videos was examined in relation to participant characteristics. Overall, n = 427 CPR videos were correctly classified in 73.3% of cases by EMS personnel and 75.7% by emergency physicians (β = 0.370, SE = 0.297, 95% CI: -0.21 to 0.95, p = 0.213). Ventilation scenarios (n = 122) were correctly classified in 93.5% (EMS) and 98.3%, (EP) (β = 4.50, SE = 6.73, 95% CI: -8.82 to 17.82, p = 0.505). The models assessing classification accuracy for CPR and ventilation did not reach statistical significance (p(CPR)=0.869 and p(ventilation)=0.183), and none of the tested predictors were significantly associated with evaluation accuracy. No significant differences in evaluation accuracy for CPR and ventilation videos were observed between professional groups or across tested participant characteristics.