The ECG may vary during the day (intra-day), and between days (interday), for the same subject. Variability in ECG characteristic measurements between different investigators is well documented and is often large. During days 1-6 of each placebo Period of a two-way crossover Phase I study, digital ECGs were recorded at about 8 and 12 AM in 16 healthy volunteers (8 men, 8 women). Two observers independently analyzed leads V-2 and V-6 using EClysis software. The durations and amplitudes of major ECG waves and the intervals between major electrocardiographic events were analyzed in a mixed model ANOVA, in which subject, observer, time, and day were treated as random factors. The influence of various corrections for heart rate on the variability of QT intervals was investigated. The difference among subjects explained between 44-81% of the total variability in ECG intervals and amplitudes. Overall, interand intraday variability was not statistically significant for any variable. The individualized exponential correction of the QT interval for heart rate eliminated the QT interval dependence on the RR interval in all subjects. Changes in T wave morphology and shortening of the QT intervalfrom morning to noon were observed in ten subjects. The interobserver variability was close to zero (SD < 0.005 ms) for all variables except the PQ interval (SD 1.4 ms). The various sources of variability in determinations of ECG wave characteristics should be considered in the design of clinical studies. The use of EClysis software for ECG measurements in this study made the results highly observer independent. (PACE 2003; 26[Pt.II]:361-366).
