Limitations of electrocardiographic scoring systems for estimation of left ventricular function

Four electrocardiographic scoring systems for the assessment of left ventricular function or presence of myocardial infarction were evaluated in 231 patients with coronary artery disease. Electrocardiographic scores were compared with radionuclide ejection fraction and thallium perfusion studies. The correlation between Wagner's modified QRS score and ejection fraction was only fair (r = -0.60). Askenazi's sum of R wave voltage score correlated poorly with ejection fraction (r = 0.44), as did Gottwik's sum of voltage score from the Frank lead electrocardiogram (r = 0.44). Rautaharju's Cardiac Infarction Injury Score did not reliably predict presence of infarction in the patient group, nor did it correlate well with ejection fraction (r = -0.49). None of the correlations were significantly improved when only patients with a history of a myocardial infarction, a thallium defect compatible with a scar or a diagnostic Q wave were considered. Although Wagner's QRS score correlated best with ejection fraction, all scoring systems had limited clinical usefulness for estimating ejection fraction.