Reasons for ‘failing’ automated external defibrillators operated by lay rescuers

Abstract

Introduction: In the Netherlands, lay rescuer automated external defibrillators (AEDs) use in out-of-hospital cardiac arrest has almost tripled between 2006–2012 from 21% to 59%. Reports to our study center about failing AEDs have increased: AEDs not giving an advice to shock. The aim of this study was to assess to what extent AEDs are ‘failing’ and if this is due to device-dependent or to operator-dependent factors. Methods: From Jan–Dec 2012 we studied all analysis periods from the collected AED electrocardiograms (ECGs) in two Dutch regions. From each analysis period we determined the correctness of the (non) shock advice and reasons for an incorrect (non) shock advice. Furthermore, if a shock advice was given, we assessed if the shock was delivered. If no shock was delivered, we determined the reason for no shock delivery. Results:We analyzed 653 AED ECGswith 1834 rhythm analysis periods (594 shock advices; 1240 non-shock advices). Sensitivity and specificity for shockable rhythm detection was 96.6% and 98.1% respectively. The AED gave an incorrect shock advice in 1.9% (23/1243) of the analyses (false positive) due to devicedependent errors (n=9) and operator-dependent errors (n=14). In 3.4% (20/591), the AED incorrectly gave a non-shock advice (false negative) due to device-dependent errors (n=12) and operatordependent errors (n=8) such as movement. In 5.1% (30/594), a shock advicewas given but no shockwas delivered: operator failed to deliver shock (n=18), electrodes removed before shock delivery (n=7) and operator pressed ‘off’ button instead of ‘shock’ button (n=5). Conclusion: Our results demonstrate that errors associated with lay rescuer AED use are low. Most perceived AED failures are due to operator-dependent causes. Paying attention to these errors in CPR training programs might even lower the amount of ‘failing’ AEDs.