From Training to Reality: System-Level Barriers and Behavioral Gaps in Prehospital Cardiopulmonary Resuscitation...

Despite regular cardiopulmonary resuscitation (CPR) training, emergency medical technicians (EMTs) often struggle to transfer learned skills into real-world performance. This study explores how system-level and contextual factors affect learning transfer in Taiwan, using the Systems Engineering Initiative for Patient Safety and Kirkpatrick frameworks.

We employed a mixed-methods design involving 123 EMTs. Data sources included quality cardiopulmonary resuscitation simulation scores, 125 video-recorded out-of-hospital cardiac arrest events, and EMTs' open-ended reflections. Quantitative analyses examined associations between CPR performance and training or demographic variables. Qualitative data were analyzed thematically, guided by Systems Engineering Initiative for Patient Safety domains and Kirkpatrick levels 2 to 3.

Automated external defibrillator voice prompts were significantly associated with improved compression rates ( P = .03 in regression analysis), whereas no demographic factor predicted compression depth or recoil, and the regression model explained only a small fraction of performance variance (adjusted R2 = 0.04). Video data revealed frequent field errors such as equipment misplacement, delayed automated external defibrillator use, and poor team coordination. Reflections highlighted environmental barriers and lack of structured debriefing.

Findings suggest that real-world CPR performance is more influenced by systemic and contextual obstacles than by individual competencies. Enhancing psychological realism, team-based simulations, and debriefing practices may improve learning transfer from training to field performance. The integration of simulation, field video, and EMT reflections underscores that training-performance gaps must be addressed through system-level reforms rather than isolated technical retraining.