Abstract 123: Field Accuracy of Paramedic Shock Delivery Using a Filtered ECG Rhythm
Introduction: Recognition and rapid defibrillation of shockable rhythms is strongly associated with increased survival from out-of-hospital cardiac arrest (OHCA). The Salt Lake City Fire Department (SLCFD) adopted ECG rhythm filtering technology in 2011, along with a protocol to rapidly defibrillate shockable rhythms without awaiting the end of the 2-minute CPR epoch. Paramedics were also trained to empirically shock asystole, as studies have shown poor agreement in cases of fine and moderate amplitude Ventricular Fibrillation (VF). Hypothesis: We hypothesized that the mandate to shock perceived asystole plus the use of filtering technology would result in high case sensitivity for shockable rhythms at the expense of an unknown frequency of shock delivery to organized rhythms. Methods: Prospectively collected defibrillator data from cardiac arrest cases treated by SLCFD between Dec 2011 and June 2019 were analyzed. Timing of rhythm changes and defibrillation events was manually abstracted using the manufacturer’s review software. The gold standard for rhythm interpretation was post-incident physician interpretation. Results: Paramedics attempted resuscitation in 942 OHCAs. We excluded 41 pediatric cases, 140 cases of BLS or bystander-only AED resuscitation, and 65 cases in which the defibrillator file was unavailable. Overall, 696 adult cardiac arrests with 1,389 shocks delivered were available for analysis. Shocks were delivered to 958 (69%) shockable, 261 (19%) asystole, 158 (11%) PEA, 4 (0.3%) SVT, and 8 (0.6%) unknown underlying rhythms. In 280 cases no shock was delivered despite an initial shockable rhythm in 3 of these cases. Shock delivery case sensitivity was 180/183 (0.98, 95% confidence interval [CI]:0.97-1.0) with false positive proportion of delivered shocks of 158/1,389 (0.11, 95% CI:0.10-0.13) for PEA only and 419/1,389 (0.30, 95% CI:0.28-0.33) for combined PEA and asystole. Neurologically intact (CPC 1-2) overall and Utstein survival rates were 15% and 46% respectively. Conclusions: Using ECG rhythm-filtering technology and an aggressive protocol to defibrillate VF and empirically shock asystole, we demonstrated high case sensitivity for VF at the expense of an 11% rate of shock delivery to underlying PEA.