DIFFERENCE IN QT INTERVAL MEASUREMENT ON AMBULATORY ECG COMPARED WITHSTANDARD ECG

Measurement of the QT interval on standard ECG has diagnostic importan ce in the congenital long QT syndrome, in pharmacological therapy of a rrhythmias, as well as in ischemic heart disease. It has been suggeste d that QT prolongation on ambulatory ECG (Holter) may have similar imp ortance. To assess agreement between methods, QT interval measurement on standard ECG was compared to that on Holter. Simultaneously obtaine d ECG and Holter tracings (25 mm/s) of the same complexes in leads V-1 and V-5 were studied in 14 patients (age range 4-36 years). ECG pairs (n = 100, 49 V-1 and 51 V-5) were compared over a range of QT interva l from 300-620 ms, as determined with the use of calipers by two obser vers blinded to pairing relationship. Correlation between methods was high for both observers (observer 1: r[V-1] = 0.872, r[V-5] = 0.973; o bserver 2: r[V-1] = 0.972, r[V-5] = 0.988), and interobserver variabil ity was small (> 85% of measurements within 20 ms). As compared to ECG , Holter underestimated QT interval in V-1, mean difference (QT [Holte r]-QT [ECG]) observer 1 (-23 ms, P < 0.001), observer 2 (-7 ms, P < 0. 05), and overestimated QT in V5, mean difference observer 1 (+13 ms, P < 0.001), observer 2 (+13 ms, P < 0.001). However, individual variati on between methods was wide, as expressed by the difference between in dividual measurements (95% confidence interval [V-1]: observer 1 [-99 to +53 ms] observer 2 [-47 to +33 ms]; [V-5]: observer 1 [-33 to +59 m s] observer 2 [-17 to +43 ms]). Furthermore, when using the QTA (inter val from onset of Q wave to apex of T wave) similar variability was ob served. In the assessment of QT interval, potential sources of error o f this magnitude could limit the clinical utility of ambulatory monito ring in detecting prolongation of the QT interval for diagnostic purpo ses.