Purpose: Evaluation of different electrocardiographic criteria for left ventricular hypertrophy (ECG-LVH criteria) using left ventricular mass index (LVMI) determined by magnetic resonance imaging (MRI). In addition, the relation between LVMI regression after aortic valve replacement and corresponding ECG changes regarding LVH was studied.
Methods: A group of 31 patients with severe aortic valve disease was studied to assess the presence of ECG-LVH and to measure LVMI and LV end-diastolic volume index (LVEDVI); 13 patients were restudied at 9.8 +/- 2.7 months after aortic valve replacement.
Results: Three criteria had a sensitivity of 100% (SV1 + RV5 or RV6 > 3.0 mV; SV1 or SV2 + RV5 > or = 3.5 mV; SV1 or SV2 + RV5 or RV6 > 3.5 mV), at the cost of specificity (50%, 44.4% and 44.4%, respectively). The R in I > 1.4 mV had a specificity of 100%, at the cost of both sensitivity and accuracy (13.6% and 54.5%, respectively). The Romhilt-Estes point score system, using > or = 4 points as cut-off value for LVH, provided the overall best accuracy of 79.5%. The number of Romhilt-Estes points also showed the best correlation with LVMI, both before and after aortic valve replacement (r = 0.81 and r = 0.67, respectively).
Conclusions: Substantial differences in sensitivity, specificity, accuracy, and correlation with LVMI were found among several ECG-LVH criteria. No single criterion performed optimally since the highest sensitivity, specificity, accuracy, and correlation with LVMI were found with different criteria. The present study may therefore contribute to a more targeted use of the existing ECG-LVH criteria.