Reliability of individual frontal plane axis determination

In a sample of 649 healthy men, the QRS and T amplitude frontal plane QRS and T axes and mean vectors were determined from leads I and III (standard reference), II and III, I and II, I and aV F , aV L and aV F , and from a vectorial combination of leads I, II, and III. The group means were similar, but there were large intraindividual differences of the QRS axis (up to ± 35°) and, to a somewhat lesser degree, of the T axis between the different lead combinations. Hypothetically, the axes from different lead combinations should be identical. The intraindividual discrepancies of the mean frontal plane vectors were also large (about 20%). In a smaller sample of 50 healthy men the axis was also determined from Frank X, Y leads, and the QRS and T axes were determined from both amplitudes (as in the larger sample) and areas. The results were similar to those obtained in the larger sample: no significant differences in the group means, but large intraindividual discrepancies between the various lead combinations both for amplitude and area axes. It is concluded that the normal standards obtained for the axis determined from leads I and III are also valid for the other lead combinations, but that for the individual patient there is no assurance that the axis determined from any one lead combination will be the true axis, even with a liberal range of error. The major part of the variation in axis as measured from various lead combinations is attributable in this study to time disparity between components measured in the individual leads, but skin impedance contributes substantially in many conventional measurements. For a patient in whom the axis is considered to be of diagnostic importance, it is probably worth averaging axes determined by several different lead combinations.