Wide complex tachycardia discrimination tool improves physicians' diagnostic accuracy.

Area of Science:

• Cardiology
• Medical Diagnostics
• Artificial Intelligence in Medicine

Background:

• Accurate differentiation of wide complex tachycardias (WCTs) into ventricular tachycardia (VT) or supraventricular WCT (SWCT) is crucial for patient management.
• A previously developed VT Prediction Model estimates VT probability using 12-lead ECGs, but its impact on physician diagnostic accuracy was unevaluated.

Purpose of the Study:

• To determine if the VT Prediction Model enhances physicians' accuracy in differentiating WCTs.
• To assess the model's contribution to diagnostic performance compared to standard ECG interpretation methods.

Main Methods:

• Nine physicians interpreted 50 WCT ECGs (25 VT, 25 SWCT) over four days.
• Interpretation methods progressed from WCT ECG alone, to WCT + baseline ECG, to WCT ECG + VT Prediction Model probability, and finally to WCT ECG + baseline ECG + VT Prediction Model probability.

Main Results:

• The VT Prediction Model significantly increased diagnostic accuracy (Day 3: 84.2%) compared to WCT ECG alone (Day 1: 68.7%) and WCT + baseline ECG (Day 2: 76.4%).
• No additional accuracy improvement was observed when baseline ECG was added to the VT Prediction Model (Day 3 vs. Day 4: 84.2% vs. 84.0%).
• The model significantly improved VT sensitivity (78.2%) and specificity (90.2%) compared to WCT ECG alone.

Conclusions:

• The VT Prediction Model demonstrably improves physician diagnostic accuracy for WCT discrimination.
• The model offers a valuable tool for enhancing the interpretation of complex cardiac arrhythmias.