New ECG algorithms with improved accuracy for prediction of culprit vessel in inferior ST-segment elevation myocardial infarction

BACKGROUND: In addition to diagnosing acute myocardial infarction (MI), the electrocardiogram (ECG) may also predict the culprit coronary artery. We aimed to assess the diagnostic accuracy of ECG algorithms predicting the occluded vessel in inferior ST-segment elevation myocardial infarction (STEMI). METHODS: This retrospective cohort study included 300 consecutive patients with inferior STEMI undergoing acute coronary angiography. A new method based on the summation of ST-segment deviations in multiple leads from the first 12-lead-ECG was used to develop algorithms to discriminate between right coronary artery (RCA) and circumflex artery (CX) occlusion. Additionally, older algorithms were reassessed. RESULTS: The RCA was occluded in 235 patients (78%) and the CX in 65 (22%). ST-segment deviations differed significantly between RCA and CX occlusions in leads I, III, aVR, aVL, aVF and V1. ST-segment deviations in lead I showed the highest discriminatory ability of a single lead (area under the receiver operating curve [AUC]=0.77). The summation of multiple leads further increased the discriminatory ability ("III-II+aVF+aVR+V1:" AUC=0.86; "III-II-I+aVF+V1:" AUC=0.85). The best binary algorithm "III-II-I+aVF+V1>0.1 mV" classified 86% of cases correctly and was better than the best old algorithm (83.3%). The simpler algorithm "III+aVR+V1≥0.1 mV" still predicted 85.0% correctly. All algorithms had higher sensitivities for RCA than for CX detection and performed better in right-dominant anatomy. CONCLUSIONS: A new approach summating multiple ST-segment deviations generated ECG algorithms with higher diagnostic accuracy to predict the occluded vessel in inferior STEMI compared to previous studies. These algorithms may facilitate earlier risk stratification for patients at risk of postinfarct complications.