Application of a New Algorithm in the Differential Diagnosis of Wide QRS Complex Tachycardia

Authors, Journal, Affiliations, Type, DOI

• Authors: András Vereckei, Gábor Duray, Gábor Szénási, Gregory T. Altemose, John M. Miller
• Journal: European Heart Journal, 2007;28(5):589–600
• Affiliations: Third Department of Medicine, Semmelweis University, Budapest, Hungary; National Healthcare Center, Cardiovascular Center, Budapest; EGIS Pharmaceuticals PLC, Budapest; Mayo Clinic, Scottsdale, AZ; Indiana University School of Medicine, Krannert Institute of Cardiology, Indianapolis, IN
• Type: Prospective original article (algorithm development and validation)
• DOI: https://doi.org/10.1093/eurheartj/ehl473

Overview

Vereckei et al. prospectively validated a new four-step WCT algorithm against the Brugada criteria in 453 EP-confirmed tachycardias (331 VT, 105 SVT, 17 pre-excited) at Indiana University. The algorithm replaces Brugada's complex morphological Step 4 with two new criteria: (Step 2) initial R wave in lead aVR and (Step 4) Vi/Vt ratio ≤1, where Vi and Vt are the voltages traversed in the initial and terminal 40 ms of the QRS respectively. The new algorithm achieved 90.3% overall accuracy vs 84.8% for Brugada (P=0.006), with superior sensitivity (95.7% vs 88.2%) and NPV for VT diagnosis (83.5% vs 65.3%); the gain was driven almost entirely by the Vi/Vt criterion outperforming Brugada's morphological Step 4. This paper was the first true prospective head-to-head comparison of a new WCT algorithm against the Brugada criteria, and led directly to the simplified aVR-only algorithm published by the same group in 2008.

Keywords

Wide QRS complex tachycardia, Brugada criteria, ventricular tachycardia, supraventricular tachycardia, aVR, Vi/Vt criterion, ECG algorithm, differential diagnosis

Key Takeaways

Background and Motivation

• Prior WCT algorithms — including the Brugada 4-step (1991) — still retained complex morphological criteria in the final step, which is both difficult to recall and the most error-prone component
• Although Brugada reported Sn 98.7%/Sp 96.5%, other groups found lower performance (Alberca 1997; Drew & Scheinman 1995); accurate diagnosis still fails in ~10% of WCTs using all published criteria
• Goal: devise a simplified algorithm eliminating most morphological criteria while maintaining or improving diagnostic accuracy; compare head-to-head with Brugada criteria

Methods

• Development cohort: 103 WCTs from Indiana University database (retrospective; used to derive algorithm)
• Validation cohort: 453 regular monomorphic WCTs (331 VT, 105 SVT, 17 PXT [pre-excited tachycardias]) recorded from 287 consecutive patients during EP study at Indiana University, June 1998 – November 2004
  • All diagnoses EP-confirmed; complete 12-lead ECG at 25 mm/sec provided for analysis
  • WCT defined: rate ≥100 bpm, QRS ≥120 ms
• Observers: two authors blinded to EP diagnosis and clinical data; interobserver variability was non-significant; results from Observer 1 reported
• BBB/fascicular block criteria: Willems et al. 1985 JACC criteria used for Step 3 classification
• Vi/Vt validation: tested in 111 sinus-rhythm ECGs with LBBB (n=25), RBBB (n=56), and non-specific IVCD (n=30) — confirmed Vi/Vt >1 in 88%, 98%, and 90% respectively

The New Vereckei 2007 Four-Step Algorithm

Applied in strict sequence; a positive finding at any step yields the diagnosis:

Step 1 — AV dissociation:

• If present → VT (identical to Step 3 of the Brugada algorithm)
• If absent → proceed to Step 2

Step 2 — Initial R wave in lead aVR:

• Definition: an R wave (or RS complex where R is the initial deflection) present as the FIRST deflection in lead aVR; an rS complex does NOT qualify — the initial deflection must be positive
• If present → VT
• Mechanism: VT originating in the ventricular myocardium commonly activates the ventricles from base to apex or in an inferior-to-superior direction, producing an initial positive (superiorly directed) force visible as an R wave in aVR; SVT with BBB/aberrancy activates the septum first via His-Purkinje → initial negative deflection in aVR (Q or q wave)
• Distinction from "northwest axis" criterion: the old axis criterion requires a predominantly positive QRS in the right-superior quadrant (axis −90° to +180°); the aVR criterion requires only an INITIAL positive deflection regardless of overall QRS polarity — SVTs with a predominantly positive QRS in aVR due to a right-superior axis can still have an initial negative deflection and thus correctly NOT trigger the criterion
• Key finding: NONE of the 17 PXTs had a positive aVR criterion → the criterion may help distinguish VT from both SVT and PXT; activation over an accessory pathway proceeds base-to-apex → negative initial deflection in aVR
• Normal variant: an initial R in aVR may occur with inferior MI (loss of inferior forces, unopposed superior forces), but in sinus rhythm R/S ratio in aVR is <1, so isolated R in aVR does not apply
• If absent → proceed to Step 3

Step 3 — QRS morphology inconsistent with BBB or fascicular block:

• If the WCT morphology does NOT correspond to any recognised BBB or fascicular block pattern (per Willems 1985 criteria) → VT
• If morphology IS consistent with BBB or fascicular block → proceed to Step 4
• Sensitivity of this criterion for VT in this dataset: 74.7% (highest single-step sensitivity in the new algorithm)

Step 4 — Vi/Vt ratio ≤1:

• Vi = vertical voltage (mV) traversed during the INITIAL 40 ms of the QRS
• Vt = vertical voltage (mV) traversed during the TERMINAL 40 ms of the QRS
• Measured in any lead with a bi- or multiphasic QRS where QRS onset and end are clearly visible AND initial ventricular activation is fastest; when initial or terminal 40 ms contain both positive and negative deflections, sum absolute values
• Lead selection: precordial leads in 87% of cases (V3, V5, V2 in decreasing order); limb leads in 13%
• Vi/Vt ≤1 (slow initial, fast terminal) → VT
• Vi/Vt >1 (fast initial, slow terminal) → SVT with aberrant conduction
• Mechanism (Vi/Vt physiological basis):
  • SVT with BBB/aberrancy: initial septal activation via His-Purkinje → rapid initial ventricular depolarisation → high Vi; widening occurs in the territory of blocked bundle → slower terminal activation → lower Vt → Vi/Vt >1
  • VT: initial muscle-to-muscle spread from ectopic focus → slow initial depolarisation → low Vi; once impulse reaches His-Purkinje system, faster activation of remaining myocardium → higher Vt → Vi/Vt ≤1
  • Antiarrhythmic drugs (class I, amiodarone) would reduce both Vi and Vt proportionally → Vi/Vt ratio preserved → criterion still applicable under AAD treatment
• Not applicable in 16/453 (3.5%) WCTs — all were VTs; causes: no biphasic QRS in any lead, or QRS onset/end indeterminate in fast VTs; the mere inability to measure Vi/Vt may itself suggest VT
• Step 4 TA: 82.2% (95% CI 75.8–88.7%) vs Brugada Step 4 morphological criteria: 68% (95% CI 60.5–75.6%) [P=0.004]

Performance Results (Table 3–4)

Overall test accuracy (453 WCTs):

Sensitivity of individual Brugada criteria in this cohort (highlighting the real-world gap from original paper):

1. No RS in any precordial: 22.8% (vs 21% Brugada 1991 — consistent)
2. RS >100 ms: 56.5% (vs 66% Brugada 1991 — somewhat lower)
3. AV dissociation: 10.1%
4. Morphological criteria: 39.4% (lowest; most errors)

Subgroup performance (new algorithm vs Brugada):

• Pre-existing BBB: 92.2% vs 85.8% [P=0.027] — new algorithm significantly superior
• Class I AAD or amiodarone: 97.6% vs 92.9% [P=0.065] — borderline superior; 27% of SVTs with pre-existing BBB + class I/amiodarone had RS >100 ms (Brugada Step 2 false positive)
• Idiopathic VT: 86.5% vs 67.6% [P=0.065] — RS interval <100 ms in 52% of idiopathic VTs (below Brugada threshold)
• Largest source of Brugada errors: Step 4 morphological criteria accounted for 41/70 (59%) of all Brugada misclassifications

Rationale for Superiority over Brugada

1. Vi/Vt replaces complex morphological criteria → higher TA in final step
2. Pre-existing BBB invalidates Brugada RS >100 ms criterion (increases false-positive VT diagnoses for SVT); Vi/Vt is not affected by BBB or AAD-induced conduction changes
3. Fascicular VT: RS <100 ms in 52% of idiopathic VTs → Brugada underperforms; Vi/Vt criterion is independent of RS duration
4. aVR initial R criterion is simpler and more reproducible than multi-lead morphological criteria

The aVR Criterion and Pre-Excited Tachycardia

• aVR criterion positive in 0/17 PXTs → activation over accessory pathway proceeds base-to-apex → negative aVR initial deflection
• This observation suggests aVR initial R may help distinguish VT from both SVT-aberrancy AND PXT — an advantage not shared by most other criteria
• Antunes et al. 1994 criteria for VT vs PXT vs SVT were infrequently applicable; aVR initial R provides a simple additional discriminator

Limitations of the Document

• Single-centre (Indiana University EP lab): referral bias; VT prevalence 73% (331/453) is at the high end for unselected populations; all patients undergoing EP study
• Vi/Vt measurement complexity: requires identifying QRS onset and end in a biphasic lead with fastest initial activation; 3.5% non-applicable; QRS onset/end definition difficult in fast VTs — three simultaneous leads required for alignment; practical reproducibility in non-EP-lab settings is uncertain
• Bundle branch reentry VT: typical BBB morphology → classified as SVT at Step 3 by both algorithms (unless AV dissociation)
• Fascicular VT: typical fascicular block pattern → classified as SVT at Step 3 unless AV dissociation is identified
• SVT with atriofascicular accessory pathway (Mahaim): typical LBBB pattern → classified as SVT at Step 3
• Not a community validation: all WCTs from EP study; performance in emergency department or acute general cardiology settings may differ
• PXT subgroup small (n=17): aVR criterion performance in PXT needs further confirmation
• The 2007 12-lead algorithm was later simplified to an aVR-only 4-step version (Vereckei 2008, Heart Rhythm), which the same group showed had comparable accuracy and improved clinical usability

Key Concepts Mentioned

• concepts/Wide-Complex-Tachycardia — primary subject; this paper introduces the Vi/Vt criterion and the initial-R-in-aVR criterion; first head-to-head comparison against Brugada
• concepts/ECG-Conduction-Disturbances — BBB/fascicular block morphology criteria (Step 3) based on Willems 1985 intraventricular conduction disturbance criteria

Key Entities Mentioned

• entities/Brugada-Syndrome — note: the Brugada criteria referenced throughout are from Pedro Brugada 1991 (sources/vt-brugada-circ-1991), distinct from the channelopathy

Wiki Pages Updated

• wiki/sources/vt-vereckei-ehj-2007.md — created (this file)
• wiki/sourceindex.md — added entry
• wiki/wikiindex.md — updated Wide-Complex-Tachycardia entry
• wiki/concepts/Wide-Complex-Tachycardia.md — added source link, updated source_count 5→6