Adherence to automated external defibrillator instructions during real-life in-hospital cardiac arrest: an analysis of device data

BACKGROUND: High-quality chest compression (CC) and timely defibrillation are crucial to survive in-hospital cardiac arrest (IHCA). Automated external defibrillators (AED) are implemented in hospitals to minimize defibrillation delay. This study examined whether AED prompts were followed by professionals during IHCA, and analyzed whether adherence influenced CC quality. METHODS: A retrospective observational study was conducted on real AED data from a German hospital between 2007 and 2024. Data included time-stamped documentation of rhythm analysis cycles, voice prompts, and CC quality metrics. Adherence was defined as the timely implementation of prompts according to predefined criteria. Statistical analysis included descriptive statistics, non-parametric comparisons, Generalized Estimating Equations for adherence rates, and chi-square testing. RESULTS: Among 466 AED files, 419 cases were analyzed. Prompt adherence ranged from 15.3% to 81.1% (p < 0.001). The prompts "start CPR" and "press shock button" were followed in 81.1% and 77.7% of cases, while "push harder" was followed in 15.3%. Mean CC depth increased after the "push harder" prompt (from 3.26 cm to 3.66 cm, 3.61 cm, 3.51 cm; p < 0.001). Mean CC rate improved only within the first interval (from 89.8 cpm to 102.2 cpm; p < 0.001). Adherence to "push harder" was associated with high-quality CC in 54.6% of cases (p < 0.001). CONCLUSION: Adherence to core AED prompts was high, whereas interruptive and corrective instructions were disregarded more frequently. AEDs for professionals should incorporate pre-analysis countdowns or alert tones to enhance user compliance. To ensure sustained high-quality compressions, devices should provide motivational-corrective instructions and prompt timely rescuer switches.