Detection of pulseless electrical activity by a public access defibrillator using ECG and ICG

Purpose: Emergency pulse checks are challenging in the out of hospital resuscitation setting even when carried out by trained rescuers. As a consequence, current European Resuscitation Council (ERC) guidelines have eliminated pulse checks for lay responders or even minimally trained operators. A hemodynamic sensing technique, capable of automatically diagnosing cardiac arrest, together with current electrocardiogram (ECG) algorithms embedded in a Public Access Defibrillator (PAD), would aid in the management of collapsed patients. An impedance cardiogram (ICG) recorded via defibrillator pads could be used and may provide opportunities for improvement over ECG alone: an ICG+ECG algorithm could be more accurate for the detection of Pulseless Electrical Activity (PEA) and provide advice about cardiopulmonary resuscitation (CPR). Algorithms reported in the literature offer impressive results by coupling the ECG and ICG. However, the required analysis may not be feasible in an emergency setting, when limited by the low processing power in any compact and low cost PAD.Methods: A retrospective analysis of ECG+ICG recorded in cardiac arrest patients and controls was used to train an algorithm to detect PEA. Data were collected following ethical approval and were marked and documented by trained physicians. Segments where CPR was administered were excluded. ECG+ICG were recorded in 132 cardiac arrest patients (53 training, 79 validation) and 97 controls (47 training, 50 validation).The detection of QRS complexes in the ECG, using a modified Pan-Tompkins approach, triggers the analysis of the ICG signal in order to detect the changes in impedance which could be masked by artifacts originating from gasping and ventilation. A threshold for the changes in the high pass filtered ICG (fc=1.5Hz) was used as a discriminator.Results: The diagnostic algorithm indicated PEA with sensitivities and specificities (95% confidence intervals) of 89.4% (88.4 –90.5) and 94.5% (94.2 –94.8) for the validation set.Conclusions: An algorithm to detect PEA, embedded in a compact PAD which simultaneously assesses ECG+ICG in real time offers encouraging results.