Failure in short-term prediction of ventricular tachycardia and ventricular fibrillation from continuous electrocardiogram in intensive care unit patients

Background: Patients in the intensive care unit (ICU) setting are prone to malignant ventricular arrhythmias. We sought to test whether electrocardiographic (ECG) markers of autonomic tone, ventricular irritability, and repolarization lability could be used in short-term prediction of ventricular arrhythmias in this patient population. Methods: We studied 38 patients with sustained (>30 seconds) monomorphic ventricular tachycardia, polymorphic ventricular tachycardia, or ventricular fibrillation while monitored in the ICU and 30 patients without arrhythmia in the ICU who served as controls. All patients had at least 12 hours of continuously recorded multilead ECG before arrhythmic event. Mean heart rate and measures of heart rate variability, QT variability, and ventricular ectopy were quantified in 1-hour epochs for the 12 hours before the arrhythmic event and in 5-minute epochs for the last hour preevent (and using a random termination time point in controls). Results: A modest downward trend in QT variability and a rise in heart rate were observed hours before polymorphic ventricular tachycardia and ventricular fibrillation events, although no significant changes heralded monomorphic ventricular tachycardia and no changes in any parameter predicted imminent ventricular arrhythmia of any type. There were no significant differences in ECG parameters between arrhythmia patients and controls. Conclusions: In ICU patients, sustained ventricular arrhythmias are not preceded by change in ECG measures of autonomic tone, repolarization variability, and ventricular ectopy. Short-term arrhythmia prediction may be difficult or impossible in this patient population based on ECG measures alone.