RT Journal Article SR Electronic T1 028 A comparison of the MACS-ECG algorithm versus contemporary computer algorithms JF Emergency Medicine Journal JO Emerg Med J FD BMJ Publishing Group Ltd and the British Association for Accident & Emergency Medicine SP 791 OP 791 DO 10.1136/emermed-2019-RCEM.28 VO 36 IS 12 A1 Niall Morris A1 Richard Body YR 2019 UL http://emj.bmj.com/content/36/12/791.2.abstract AB Background The Manchester Acute Coronary Syndromes ECG (MACS-ECG) model was derived and validated with the aim of providing an objective measure of ECG ischemia in the setting of suspected non-ST-elevation myocardial infarction. We wanted to produce an objective measure to improve existing clinical decision aids such as TMACS or the HEART score. To establish whether the MACS-ECG model warrants further study we compared the diagnostic performance against existing computer algorithms.The MACS-ECG model was derived in a single center cohort and combined a nuanced interpretation of repolarization abnormalities with novel signs of ischemia. These variables were selected using backward logistic regression using SPSS with the primary outcome being NSTEMI.This model was validated in a secondary analysis of the Bedside Evaluation of Sensitive Troponin study (BEST). We recruited patients from 17 Emergency Departments. We included adults with a suspected ACS. We excluded patients with STEMI or obvious non-cardiac cause.The primary outcome was NSTEMI, using the Fourth universal definition of myocardial infarction.In the validation study we also coded the ECG machines interpretation as acute ischemia, infarction or no statement.When the existing ECG algorithms produce an ‘acute ischemia’ statement, the sensitivity (Sn) was 13.3% (8.2–20.0%), specificity (Sp) 93.9% (92.1–95.5%), positive predictive value (PPV) 27.9% (19.1–39.0%) and negative predictive value (NPV) 86.0% (85.2 86.8%).An ‘infarction’ statement produced Sn 21.0% (14.6–28.6), Sp 88.3% (85.8–90.4%), PPV 24.0% (17.9–31.4%) and NPV 86.3% (85.3–87.3%).Combining infarction and ischemia produces the following diagnostic characteristics: Sn 34.3% (26.5–42.7%), Sp 82.2% (79.4–84.8%), PPV 25.4% (20.6–30.9%) and NPV 87.6% (86.2–88.9%).The MACS-ECG model was more specific and sensitive than existing computer algorithims. When it produced an ‘acute ischemia’ statement, the Sn was 22.4% (15.8–30.1%), Sp 95.2% (93.5–96.6%), PPV 45.1% (34.8–55.8%) and NPV 87.4% (86.4–88.4%).