In the nearly 30 years since Wellens et al. categorised most of the currently used electrocardiographic (ECG) criteria for distinguishing supraventricular tachycardia (SVT) with aberrant interventricular conduction from ventricular tachycardia (VT),1 there have been relatively infrequent and largely incremental enhancements to the art and science of making this distinction. In most published series, a correct diagnosis can be made using these criteria in up to 90% of cases. Despite this, few clinical situations evoke more anxiety in physicians and nurses than a patient with wide complex tachycardia (WCT). Often, residents and even staff physicians are seen poring over the WCT ECG and muttering under their breath, “I should know this!” Why does this situation continue to present such a clinical challenge? And what are the consequences of making an incorrect diagnosis?
WCT has several potential causes, including VT; SVT with one of the following: aberrant interventricular conduction (SVT-A), atrioventricular (AV) conduction over an accessory pathway (Wolff-Parkinson-White), QRS widening due to drug effect/electrolyte abnormalities or an abnormal bizarre baseline QRS (cardiomyopathy); or ventricular pacing. Although the proportion of cases falling into each category varies slightly depending on patient population, VT comprises about 67% of WCT in most series, with SVT-A accounting for another 25%. Thus, the most important distinction to make is between VT and SVT-A.
ECG differentiation of WCTs can be divided into two major areas: configurational (morphology of QRS) and relational (AV relationship during WCT). Configurational distinctions are based on QRS patterns that resemble aberrant conduction and are thus more consistent with SVT-A, or do not resemble aberration patterns and are thus likely to be VT. Some of the more commonly used configurational criteria that are uncommon in SVT-A and thus strongly suggest VT are prolonged QRS duration (>140ms for right bundle branch block [RBBB] pattern QRS, >160ms for left BBB [LBBB]); leftwards frontal plane axis (especially if between -90 and 180├é┬░); a fully concordant pre-cordial R wave pattern (fully positive or fully negative); and specific patterns in leads V1 and V6 that are or are not compatible with aberrant conduction (see Figure 1). Particular combinations of BBB and axis are distinctive, such as LBBB with rightwards inferior axis (+90–180├é┬░), which is almost never seen in SVT-A, and RBBB with normal axis (0–+90├é┬░), which is almost never seen in VT. Another pattern that is rare in SVT-A is the absence of any RS complex among the pre-cordial leads; Brugada et al. used this as the basis of an algorithm that they found to have 99% sensitivity and 97% specificity for diagnosing VT.2
Capitalising on the fact that in SVT-A the electrical vectors move rapidly at the beginning of the QRS and more slowly at the end (activating the ‘locked’ ventricle), whereas in VT slow conduction is the rule throughout the QRS complex, Vereckei and colleagues recently reported a new criterion that quantifies these differences.3 They calculated the ratio of the voltage amplitude change in the first versus last 40ms of a bi- or multiphasic QRS and found a ratio >1 suggested SVT-A and, if ≤1, VT was diagnosed. They also found that an initial R wave in lead aVR strongly suggested VT. This, as well as most of the published algorithms in the literature dealing with QRS configuration distinctions, reports predictive accuracies from 85 to 95%. Unfortunately, the configurational criteria are most helpful when the patient’s baseline QRS complex is not significantly widened. If the baseline QRS is very abnormal, the configurational criteria lose much of their predictive capacity4 with the exception that if the QRS complexes of the WCT are identical in configuration to those during baseline ECG, chances are good that the WCT is SVT.5 In many cases, however, the baseline ECG is simply not available for comparison. Recently, some of the long-trusted ECG differentiating criteria have undergone re-evaluation in light of changes in the patient population.6 In particular, patients with heart failure are living for longer with improved medical therapies. Many of these patients have widened, very abnormal baseline QRS complexes and some have episodes of SVT conducted with bizarre patterns that suggest VT but are identical to their baseline QRS configuration. Perhaps because of this, one differentiating criterion that has ceased to be helpful is a negative concordant pre-cordial pattern. In a recent series of WCTs, this pattern was found as frequently in SVT (10%) cases as in VT (12%), whereas previously it was seen almost exclusively in VT.
In the relational area, the AV relationship during WCT is conceptually straightforward and has long been a useful tool in diagnosing WCTs, since – with only extremely rare exceptions – SVTs have at least as many P waves as QRS complexes (AV ratio ≥1). In contrast, since VTs do not require atrial participation, the AV ratio is ≤1 (that is, AV dissociation or retrograde 2:1 or Wenckebach pattern is present). The AV relationship is independent of complicated morphological criteria as well as the pattern of the patient’s QRS complex during normal rhythm. While these features of the AV relationship are attractive, it is often difficult to discern clear atrial activity during WCT, or the presence of atrial fibrillation or flutter confounds the issue. In a recent series,6 a diagnostic AV pattern was observed in only 36% of VTs; thus, while it is very specific, it is not a sensitive criterion. Since the AV relationship is such a specific differentiator when positive, some stepwise algorithms for analysing WCTs incorporate it as a first step./>/>/>/>/>/>/>