QT dispersion in medicine: electrophysiological holy grail or fool's gold?

Abstract

If subjects at high risk of sudden cardiac death were easily identifiable, then targeted therapy might be able to reduce cardiac deaths. Unfortunately, we do not yet possess an applicable screening method for this purpose. Techniques exist for this such as signal‐averaged electrocardiography (ECG), T‐wave alternans and heart rate variability, but they have variable success and tend to require specialized equipment, making them difficult in routine practice. Another possibility is QT interval analysis, which stems from the fact that individuals with long QT syndromes are known to be at high risk of sudden cardiac death. Taking this principle one step further, it is possible that the variation of QT intervals within an ECG in more routine patients may also contain prognostic information. ‘QT interval dispersion’ is at present undergoing vigorous assessment for this purpose. Several years ago, Campbell et al .1 enthusiastically called it the ‘electrophysiological Holy Grail’. The number of studies indexed in the Medline on QT dispersion has risen 34‐fold since its description in 1990. There is therefore a need to synthesize this information and discuss whether this technique could be adopted in clinical practice. A Dutch physician, Willem Einthoven (1860–1927) introduced the ECG ‘PQRST’ designations we use today, and the ‘QT interval’ has been known since 1887 to represent ventricular electrical activities.2 One hundred years later, a group from Newcastle3 proposed that the interlead QT interval differences within a 12‐lead ECG might reflect regional differences in myocardial refractoriness, and that this might predict cardiac dysrhythmias. Animal4 and human5 studies supported this observation. Using epicardial monophasic action potentials in isolated rabbit hearts, Zabel et al .6 correlated this QT interval variation with the degree of homogeneity in ventricular repolarization. This correlation would suggest that QT dispersion is at best a surrogate marker, …