Abstract 2657: Diagnostic Accuracy of Electrocardiographic Criteria for Left Ventricular Hypertrophy: Comparison With Computed Tomography

Introduction: Several electrocardiographic (ECG) criteria for left ventricular hypertrophy (LVH) exist. However, none of these criteria has been validated against cardiac CT, which provides accurate and reproducible assessment of left ventricular mass (LVM). The aim of this study is to evaluate the diagnostic accuracy of different LVH ECG criteria using cardiac CT as the gold standard. Methods: We included 353 consecutive patients (mean age 55±12 years, 48% females) who underwent coronary CT angiography for various clinical reasons. LVH on CT was defined as LVM index (LVMI) >103 gm/m2 for men and >89gm/m2 for women. ECGs were interpreted by physicians blinded to the LVMI. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and receiver operating curves were calculated for the following LVH ECG criteria using LVMI by CT as the gold standard: Sokolow-Lyon voltage, RaVL, Cornell gender-specific voltage, and Cornell product. Results: A total of 28 patients (7.9%) had LVH by CT and 57 patients (16.1%) had LVH by any ECG criteria. Patients with LVH were older with higher prevalence of hypertension. All ECG criteria had similarly low sensitivity and PPV for the diagnosis of LVH, but high specificity and NPV. The Cornell and aVL criteria are the most sensitive criteria while the Cornell product criterion is the most specific. The false positive and false negative rates range from 6.3% to 7.1% and from 77.8%– 82.4% respectively. Conclusion: Using cardiac CT as the gold standard, ECG criteria for diagnosing LVH are very specific, but all have poor sensitivity. The absence of LVH by ECG provides little reassurance of a normal left ventricular mass by cardiac CT.