Pulmonary embolism (PE) is a major health problem associated with a significant morbidity and mortality. Immediate recognition of submassive and massive cases is extremely important in order to commencement of early and appropriate therapy that could be life saving. The aim of this study was to assess the ability of two scoring systems, electrocardiography (ECG) and simplified Welts (sWells) clinical scorings in predicting anatomic severity of PE. Hence, ECG and sWells scorings were combined in order to test the hypothesis if this new scoring does enhance the prediction of severity. Fifty six patients with proven PE with high (50 patients) and moderate (six patients)-probability of ventilation/perfusion (V/Q) scan were retrospectively studied. Baseline ECGs were analysed by two independent observers in order to constitute ECG scorings. Baseline sWells scores were also calculated. Anatomic severity of PE was calculated by scintigraphically and categorized into mild (< 50% perfusion defect) (group 1), and severe (>= 50% perfusion defect) (group 2) diseases. The mean of ECG scores, sWells scores and the combined scores were 5.23 +/- 3.42 and 5.85 +/- 3.82; 6.60 +/- 1.88 and 7.03 +/- 2.40; and 10.73 +/- 3.60 and 11.60 +/- 4.32 in groups land 2, respectively (p > 0.05). An ECG score of 6.5 predicted severe disease (perfusion defect >= 50%) with a sensitivity of 41.7% and a specificity of 82%). sWells and combined scores did not provide better sensitivity or specificity values based on ROC curve analysis. Our results indicated that ECG scoring could be valuable test in predicting anatomic severity of PE, adding sWells scoring to ECG scoring did not provide any beneficial effect. (c) 2006 Elsevier Ltd. All rights reserved.