Variation in the identification of Q wave initiation and its contribution to QT measurement

Cardiac repolarization abnormalities can be assessed from measurements of the QT duration taken from paper electrocardiogram recordings. Errors associated with determining the end of the T wave are known, but those associated with the start of the Q wave have so far been neglected. This paper quantifies the variation in manual identification of the start of the Q wave, and assesses its contribution to errors in the manual measurement of QT. A randomized study of errors in the timing of Q wave initiation from electrocardiograms plotted on paper was conducted. Four electrocardiogram leads were recorded in eight subjects relaxing in a semi-recumbent position. Manual measurements were made of the time of Q wave initiation in 512 electrocardiograms, presented with different superimposed noise, recording speed and recording gain. The greatest mean difference between four cardiologists amounted to 6.7 ms. A recording gain of 5 mm mV-1, in comparison with 10 mm mV-1, resulted in a difference in Q wave timing of 3.2 ms (P < 0.05). A further increase in gain, or the addition of noise up to 20 microV made no significant difference to Q wave measurements. Provided ECGs of at least 10 mm mV-1 are used, the effect of variation in Q determination on QT measurement is likely to be small.