[Validity of some electrocardiographic criteria in left ventricular hypertrophy]

The aim of the study was the assessment of the sensitivity, specificity and accuracy of some ECG criteria of left ventricular hypertrophy (LVH). Left ventricular mass (LVM) measured on the M-mode echogram of the left ventricle was the reference standard. Ninety-four (94) unselected, consecutive clinical patients (34 women, 60 men, average age 47 years) underwent in the same day, ECG and echocardiogram. Exclusion criteria were the presence of ischemic heart disease, atrial fibrillation and Wolff-Parkinson-White syndrome. The ECG tracings were interpreted independently by 2 investigators following 5 independent criteria of LVH: 1) Sokolow-Lyon voltage criterion (SL); 2) a modified Romhilt-Estes point score (REM); 3) left atrial abnormality (LAA); 4) left ventricular strain; 5) a new voltage criterion RaVL + SV3 corrected by sex and age. The left ventricular M-mode echograms were recorded using a left parasternal approach and were interpreted independently by 2 investigators. LVM was measured using the "Penn convention" and taking the R wave peak as end-diastole. The prevalence of LVH (= LVM 215 g) in the study population was 47%. The following results were achieved (sensitivity, specificity, accuracy): SL: 68.2%, 84%, 76.6%; REM: 63.6%, 90%, 77.7%: LAA: 36.4%, 84%, 61.7%; strain 52.3%, 72%, 62.8%; RaVL + SV3: 54.5%, 82%, 69.1%. Our data suggest: 1) the high sensitivity, specificity and accuracy of Romhilt-Estes point score are confirmed; 2) the sensitivity of Sokolow-Lyon voltage criterion is reevaluated; 3) the most sensitive morphological criterion seems to be the left ventricular strain; 4) a new voltage criterion could be useful.