Left ventricular mass and hypertrophy assessment by means of the QRS complex voltage-independent measurements

ECG QRS-complex voltage-based criteria are relatively insensitive for detection of increased left ventricular mass (LVM). We developed and evaluate a new ECG index for LV hypertrophy (LVH) detection regardless of the QRS voltage. Methods: Study population consisted of 106 patients (73 in, 33 f, aged 60 10 years) with established coronary artery disease (CAD). All patients had LVM assessed echocardiographically and indexed to BSA (LVMIECHO). LVH was diagnosed if LVMIECHO > 117 g/m(2) in men and > 104 g/m(2) in women. LV geometry was also determined. Analysed ECG variables, obtained from 12 leads recorded simultaneously, were: the QRS complex duration (QRSd, ins), the average 12-lead time to maximal deflection (TMD, ms), the average 12-lead QRS complex voltage (12QRSV, mV), the average product of 12 lead QRS voltage and duration (12QRSVd, mV ms), Sokolow-Lyon voltage and V-d product (SLV, SLVd), Cornell voltage and V-d product (CV, CVd). A newly developed index, LVMECG, was calculated, as LVMECG=[(2 x TMD+QRSd/pi)(3)-(QRSd/pi)(3)] *0.0001 (ms(3)), and indexed to BSA (LVMIECG, ms(3)/m(2)). Results: Means of the QRS voltage-related parameters were similar in patients with LVH and normal LVM. Greater differences existed between both groups when the QRS voltage-duration products were compared. LVMIECG was most powerful in distinguishing between groups (130 +/- 33 LVH vs 91 +/- 21 normal LVM, p < 0.001). LVMIECG correlated with LVMIECHO better (r = 0.77,p < 0.001) than other indices (r coefficients between 0.24 for SLV and 0.49 for CVd). None of the examined indices allowed for distinction between eccentric and concentric LVH. The new index showed better statistical performance (area under ROC = 0.861) compared to the other indices (AUC range 0.545-0.697, p < 0.001 vs LVMIECG). At the specificity level of 92%, the value of LVMIECG > 120 ms(3)/m(2) had the sensitivity of 64% for detection of increased LVM. The sensitivities of the other parameters were significantly lower (sensitivity range 18-42%). Relative intra- and interobserver errors and correlation coefficients for LVMIECG calculation were 0.4% and 1.6% and r=0.94 and 0.98, respectively. Conclusions: In patients with CAD an assessment of LV mass and detection of hypertrophy using the QRS complex time-dependent index is feasible. The new index correlated well with echocardiographically-determined LVM and showed better statistical performance than indices which include QRS-voltage measurements. The results are promising and warrant further studies to evaluate the utility of the new index as a risk predictor. (c) 2005 Elsevier Ireland Ltd. All rights reserved.