Session S31.2

Validation of New Enhanced ACC/ESC STEMI Criteria on a Population of Patients with Suspected Acute Coronary Syndrome

JJ Wang*, GS Wagner, TN Martin, JW Warren,
M Mirmoghisi, BM Horácek

Philips Healthcare
Andover, MA, USA

Currently used ECG criteria for identifying patients with ST-elevation myocardial infarction (STEMI) perform with high specificity (SP), but low sensitivity (SE). In our previous study, based on data acquired during balloon-inflation angioplasty, we found that the SE of ACC/ESC STEMI criteria can be improved without any loss of SP by considering ST elevation in 3 added pairs of contiguous leads (namely, aVL, –III; III, –aVL; and –V2, –V3). The aim of the present study was to validate our previous findings on an independent dataset consisting of patients admitted to the Western Infirmary Glasgow with symptoms suggesting an acute coronary syndrome. We analyzed a set of 116 12-lead ECGs of patients with (n = 58) and without (n = 58) acute myocardial infarction confirmed by MRI. ST deviation was measured at J point in 12 standard leads by a cardiologist (GSW) from pink paper tracings, with resolution of 50µV; these tracings were subsequently scanned (472 pixels per 1 mV) and high-resolution ST measurements were independently made from computer monitor by 2 readers (JWW, MM) and an average of these 2 values was used. By using manual measurements, we found that the detection of ischemic state by means of the existing STEMI criteria achieved SE/SP of 50/97% and of 66/97% with the enhanced STEMI criteria. Computer-assisted measurements yielded SE/SP of 43/98% for the existing STEMI criteria and 55/98% for the enhanced STEMI criteria. To understand why the SE was lower in the latter measurements compared to manual ones, we gradually lowered thresholds for ST elevation in 5-µV steps and observed how it affected SE/SP of computer-assisted measurements. We found that at thresholds lowered by 30µV, SE/SP values equaled those achieved with manual measurements. Thus, we attributed the discrepancy to “rounding” in cardiologist’s readings. Based on these results we conclude that, in our test set, existing ACC/ESC STEMI criteria enhanced by just 3 additional features yielded improved sensitivity of ischemia detection without any detrimental effect on specificity. This finding, which needs to be corroborated on a larger study population, indicates that the current ACC/ESC criteria can still be further improved.

(Abstract Control Number: 22)