Usefulness of clinical characteristics in predicting the outcome of electrophysiologic studies in unexplained syncope.

ABSTRACT

Guidelines for the use of electrophysiologic studies in syncope have not yet been formulated. To confirm the sensitivity and specificity of a previously derived model to predict the results of electrophysiologic testing in syncope, the importance of 6 clinical predictors was assessed in a new data set of 141 consecutive patients with unexplained syncope who were referred for electrophysiologic studies. The 6 predictors were organic heart disease; premature ventricular beats, sinus bradycardia, first-degree heart block and bundle branch block by electrocardiogram; and nonsustained ventricular tachycardia by Holter monitor. Organic heart disease and nonsustained ventricular tachycardia by Holter monitoring were highly sensitive for serious ventricular tachyarrhythmias at electrophysiologic study (sensitivity 100%), whereas sinus bradycardia, first-degree heart block or bundle branch block by electrocardiogram were sensitive for bradyarrhythmic outcomes (sensitivity 79%). Because these variables are so sensitive for serious outcomes of electrophysiologic testing in syncope, invasive studies in patients without these clinical predictors are likely to be of very low diagnostic yield.