Back to the basic: measuring QRS voltage-interval in left ventricular hypertrophy.

Back to the Basic: Measuring QRS Voltage-Interval in Left Ventricular Hypertrophy To the Editor: The electrocardiographic patterns of left ventricular hypertrophy or left ventricular strain have received considerable attention for investigators in the last half a century. A variety of ECG measurements, involving QRS voltage ( V or mV) and timeinterval, have been used to define left ventricular hypertrophy (LVH) in the Sokolow-Lyon (1949), Cornell (1985), and Cornell product (1995) criteria.1–3 Both QRS interval (0.04 milliseconds) and ST deviation ( V) improve the sensitivity and specificity of the voltage criteria.3 In general, these ECG criteria have been validated by echocardiography, a method that overestimates left ventricular mass relative to MRI. In the August issue of Hypertension, K. Alfakih et al presented the first new gender-specific ECG criteria based on cardiac MRI.4 They found that at a specificity of 95%, the SokolowLyon product enhances sensitivity in females (26.2%), Cornell criterion in males (26.2%), and Cornell product criteria in both males and females (25.0%, 23.8%). These results were based on ECG recorded at 25 mm/s and at 1 mV/cm standardization. However, a gross mistake is evident in the methods and results (Table 1 of Alfakih et al): LVH voltage was defined as a Sokolow-Lyon voltage (SV1 RV5 or RV6) 35 mV (instead of 35 mm), Cornell voltage (RaVL SV3) 28 mV for men and 20 mV for women (versus 28 mm and 20 mm), and Cornell product [(RaVL SV3) QRS duration] 2440 mV ms (versus 2400 mm ms). Numbers in parenthesis correspond to the actual values for such indexes.1–3 The authors even present an incorrect version of the LIFE study reported in Lancet 2002, when quoting QRS interval and Cornell voltages “with an adjustment of 8 mV in females and a partition value of 2440 mV ms”. Because QRS amplitudes were expressed in 1 mm 1mV, instead of 100 V or 0.1 mV/mm, QRS voltages in Alfakih et al’s Table 1 do not represent Sokolow, Cornell, or Cornell product criteria for any sex group, with or without LVH. These measurements should be corrected and expressed in electrical units ( V or mV ms), a reliable parameter for the genderspecific LVH. This unexpected “error” from a Research Heart Center probably arises from the equivocal use of metrical (1 mm) instead of electrical (100 V 1 mm) units in virtually all medical and ECG textbooks, even after 6 decades of Ashman works on ECG units: 1 mm square 100 V 0.04 ms,5 which may explain an identical mistake in Cornell Voltage-duration product (2483 to 3594 mV ms or 3595 mV ms, Table 6 in Reference 6) in a recent article of the LIFE study published in Hypertension.6