Abstract 16585: A Novel Algorithm Makes Accurate Shock or No-Shock Decisions During Ongoing Chest Compressions
Introduction: During cardiac arrest, prolonged interruptions in chest compressions (CCs) are associated with lowered chances of survival. Because CCs cause substantial artifact on the ECG, automated external defibrillators (AEDs) today require rescuers to pause CCs in order for the device to make a ‘shock’ or ‘no shock’ decision. A new algorithm, cprINSIGHTTM Analysis Technology, uses ECG and transthoracic impedance signals during ongoing CCs to decide ‘shock,’ ‘no shock,’ or ‘pause CCs for further analysis.’ If it can avoid pausing in most cases, while correctly choosing who does and does not need a shock, the algorithm may improve patient outcomes.
Purpose: In this study, we analyzed the performance of the algorithm on a new set of out-of-hospital cardiac arrest data. The goal for cprINSIGHT was to come to a ‘shock’ or ‘no shock’ decision more than 65% of the time without requiring a pause, with sensitivity and specificity over 92%.
Methods: We included 2,941 30-second segments of ECG and impedance data during ongoing CCs in 520 patients treated by an EMS system in the United States. To determine the patients’ true heart rhythms, three physician experts used “clean” segments adjacent to those artifacted segments, and came to consensus for each segment on whether the rhythm was shockable or non-shockable. cprINSIGHT was independently run on the artifacted segments, and compared to this gold standard.
Results: For 71.0% of the segments [95% Lower Single-Sided Confidence Level (LSSCL) > 68.9%], cprINSIGHT came to a shock or no shock decision without requiring a pause in CCs. In these segments, the algorithm achieved a sensitivity of 96.1% (446 of 464 shockable segments) [LSSCL > 94.4%], and a specificity of 96.6% (1570 of 1625 nonshockable segments) [LSSCL > 95.7%].
Conclusions: The results indicate that this algorithm makes accurate decisions and can therefore eliminate many of the pauses required by conventional AEDs to make shock / no shock decisions. By eliminating these pauses, cprINSIGHT has the potential to reduce interruptions in CPR-generated blood flow, and improve patient outcomes from cardiac arrest.
The authors wish to thank Prince George’s County Fire/Emergency Medical Services Department for collection of data making this study possible.
Author Disclosures: D.W. Piraino: Employment; Significant; Physio-Control, Inc. A. Esibov: Employment; Significant; Physio-Control, Inc. F.W. Chapman: Employment; Significant; Physio-Control, Inc..
- © 2016 by American Heart Association, Inc.