Differential Diagnosis of Wide Complex Arrhythmias using Multivariate Analysis of RR Intervals and QRS Complex Durations

Mikhail Chmelevsky Department of Electrocardiology, Almazov National Medical Research Centre, Saint-Petersburg, Russia

Introduction. Most of current algorithms of differential diagnosis of wide complex arrhythmias are based on the detailed analysis of QRS morphology using standard 12-lead ECG. We examined a diagnostic value of comprehensive multivariate analysis of RR intervals and QRS duration in differential diagnosis of wide complex arrhythmias.

Methods. The study enrolled 108 patients (male – 64 (59,3%), age Me (min; max) – 56,5 (8; 85)) with 279 wide QRS morphologies (min; Me; max – 120; 144; 222 ms) during sinus rhythm. An average (min, max) of 3 (1;8) different wide QRS morphologies were selected from Holter 24-h ECG data (KT Result 3, INCART, RF). Beforehand, a number of 230 (82,4%) and 49 (17,6%) of all wide QRS morphologies were validated as SVA with aberrant conduction and VA, respectively. Five consecutive RR intervals before and after examined wide QRS complex were selected as well as nine QRS morphologies. The durations of all these intervals (ms) were exported to perform a statistical uni- and multivariate analysis followed by evaluation of specifity (Sp) and sensivity (Sn) with 95% CI level (Statistica v.12, Statsoft Inc., USA).

Results. The duration of SVA (min; Me; max – 120; 132; 179 ms) and VA (min; Me; max – 120; 148; 222 ms) was different (p <0,001). ROC analysis showed Sn (95% CI) 59,6% (52,9 – 66,0%) and Sp 79,6% (65,7 – 89,8%) for QRS duration with 140 ms criterion. The following linear discriminant function analysis of all RR intervals and QRS durations classified SVA and VA with Sn 78.3% (72.4 – 83.4%) and Sp 83.7% (70.3 – 92, 7%) with p <0.001.

Conclusions. Our study showed a significant prospect to improve differential diagnosis of wide complex arrhythmias using comprehensive multivariate analysis of RR intervals and QRS durations without other morphological QRS complex characteristics.









Powered by Eventact EMS