Direct Comparison of Methods to Differentiate Wide Complex Tachycardias: Novel Automated Algorithms Versus Manual ECG Interpretation Approaches
Sarah LoCoco,Anthony H. Kashou,Abhishek J. Deshmukh,Samuel J. Asirvatham,Christopher V. DeSimone,Krasimira M. Mikhova,Sandeep S. Sodhi,Phillip S. Cuculich,Rugheed Ghadban,Daniel H. Cooper,Thomas M. Maddox,Peter A. Noseworthy,Adam M. May
DOI: https://doi.org/10.1161/circep.123.012663
2024-07-26
Circulation Arrhythmia and Electrophysiology
Abstract:Circulation: Arrhythmia and Electrophysiology, Ahead of Print. BACKGROUND:Differentiating wide complex tachycardias (WCTs) into ventricular tachycardia (VT) and supraventricular wide tachycardia via 12-lead ECG interpretation is a crucial but difficult task. Automated algorithms show promise as alternatives to manual ECG interpretation, but direct comparison of their diagnostic performance has not been undertaken.METHODS:Two electrophysiologists applied 3 manual WCT differentiation approaches (ie, Brugada, Vereckei aVR, and VT score). Simultaneously, computerized data from paired WCT and baseline ECGs were processed by 5 automated WCT differentiation algorithms (WCT Formula, WCT Formula II, VT Prediction Model, Solo Model, and Paired Model). The diagnostic performance of automated algorithms was compared with manual ECG interpretation approaches.RESULTS:A total of 212 WCTs (111 VT and 101 supraventricular wide tachycardia) from 104 patients were analyzed. WCT Formula demonstrated superior accuracy (85.8%) and specificity (87.1%) compared with Brugada (75.2% and 57.4%, respectively) and Vereckei aVR (65.3% and 36.4%, respectively). WCT Formula II achieved higher accuracy (89.6%) and specificity (85.1%) against Brugada and Vereckei aVR. Performance metrics of the WCT Formula (accuracy 85.8%, sensitivity 84.7%, and specificity 87.1%) and WCT Formula II (accuracy 89.8%, sensitivity 89.6%, and specificity 85.1%) were similar to the VT score (accuracy 84.4%, sensitivity 93.8%, and specificity 74.2%). Paired Model was superior to Brugada in accuracy (89.6% versus 75.2%), specificity (97.0% versus 57.4%), and F1 score (0.89 versus 0.80). Paired Model surpassed Vereckei aVR in accuracy (89.6% versus 65.3%), specificity (97.0% versus 75.2%), and F1 score (0.89 versus 0.74). Paired Model demonstrated similar accuracy (89.6% versus 84.4%), inferior sensitivity (79.3% versus 93.8%), but superior specificity (97.0% versus 74.2%) to the VT score. Solo Model and VT Prediction Model accuracy (82.5% and 77.4%, respectively) was superior to the Vereckei aVR (65.3%) but similar to Brugada (75.2%) and the VT score (84.4%).CONCLUSIONS:Automated WCT differentiation algorithms demonstrated favorable diagnostic performance compared with traditional manual ECG interpretation approaches.
cardiac & cardiovascular systems