The calculation of the heart axis in the frontal plane can be performed wit
h the combination of any two leads. The use of combination of bipolar (I, I
I, III) and unipolar leads (aVR, aVL and aVF) can produce wrong results. Ca
lculation of the electrical axis from leads I and aVF without correction (s
ometimes used in ECG recorders): EA= +/-Arctan (alpha VF/I) results in lowe
r values tin our study: 34 degrees+/-4 degrees, n = 48) as compared to the
values obtained with formula that uses leads I and II: EA= +/- Arctan ((2*I
I-I)/(Sqr(3)*I)) (axis = 37 degrees+/-3 degrees, n=48; p<0.005, paired t-te
st with Bonferroni correction) or with corrected formula which uses leads I
and aVF: EA = +/-Arctan ((2*alpha VF)/(Sqr(3)*I)) (axis = 37 degrees+/-4 d
egrees, n = 48; p<0.005, paired t-test with Bonferroni correction). The cor
rection factor 2/Sqr(3) is required because the unipolar and bipolar leads
have different strengths. Although the difference rarely reach clinical sig
nificance, our results suggests that the ECG recorders should be proofed on
formulas used for the calculation of the electrical axis.