An algorithm for identification of ST-elevation myocardial infarction patients by emergency medicine services

Abstract Objective ST-elevation myocardial infarction (STEMI) identification by emergency medicine services (EMS) leading to pre-hospital catheterization laboratory (CL) activation shortens ischemic time and improves outcomes. We examined the incremental value of addition of a screening clinical tool (CT), containing clinical information and a Zoll electrocardiogram (ECG)-resident STEMI identification program (ZI) to ZI alone. Methods All EMS-performed and ZI-analyzed ECGs transmitted to a percutaneous coronary intervention hospital from October 2009 to January 2011 were reviewed for diagnostic accuracy. ZI performance was also compared to ECG interpretations by 2 experienced readers The CT was then retrospectively applied to determine the incremental benefit above the ZI alone. Results ST-elevation myocardial infarction was confirmed in 23 (7.5%) of 305 patients. ZI was positive in 37 (12.1%): sensitivity: 95.6% and specificity: 94.6%, positive predictive value (PPV), 59.5%, negative predictive value (NPV), 99.6%, and accuracy of 93.8%. Moderate agreement was observed among the readers and ZI. CT criteria for CL activation were met in 24 (7.8%): 20 (83.3%) were confirmed STEMIs: sensitivity: 86.9%, specificity: 98.5%, a PPV: 83.3%, and NPV: 98.6%, accuracy of 97.7%. CT + ZI increased PPV ( P < 0.05) and specificity ( P < 0.003) by reducing false positive STEMI identifications from 15 (4.9%) to 4 (1.3%). Conclusions In an urban cohort of all EMS transmitted ECGs, ZI has high sensitivity and specificity for STEMI identification. Whereas the PPV was low, reflecting both low STEMI prevalence and presence of STEMI-mimics, the NPV was very high. These findings suggest that a simplified CT combined with computer STEMI interpretation can identify patients for pre-hospital CL activation. Confirmation of these results could improve the design of STEMI care systems.