• Robin A Ducas, Anthony W Wassef, Davinder S Jassal, Erin Weldon, Christian Schmidt, Rob Grierson, and James W Tam.
• University of Manitoba, Winnipeg, Manitoba, Canada. umducasr@cc.umanitoba.ca
• Can J Cardiol. 2012 Jul 1;28(4):432-7.

BackgroundThere is growing use of prehospital electrocardiograms (ECGs) in establishing early diagnosis of ST segment myocardial infarction (STEMI) to facilitate early reperfusion. This study aimed to determine the predictive value of prehospital ECGs interpreted by nonphysician emergency medical services (EMS) in chest pain presentations.MethodsIn our city of 658,700 people, EMS/paramedics received 21 hours of instruction on STEMI management, ECG acquisition, and interpretation. Suspected STEMI ECGs were wirelessly transmitted to and discussed with a physician for possible therapy. ECGs deemed negative for STEMI by EMS were not transmitted; patients were transported to the closest hospital without prehospital physician involvement.ResultsFrom July 21, 2008 to July 21, 2010, there were 5426 chest pain calls to EMS, 380 were suspected STEMI cases. The remaining ECGs were deemed negative for STEMI by EMS. To audit the nontransmitted ECGs we analyzed 323 consecutive patients over 2 selected months (January and June 2010) for comparison. Of nontransmitted cases there was 1 missed and 2 STEMIs that developed subsequently. Based on 380 transmitted and 323 nontransmitted cases, the sensitivity and specificity of EMS detecting STEMI were 99.6% and 67.6%, respectively. The positive and negative predictive values for STEMI were 59.5% and 99.7%, respectively.ConclusionsOur findings demonstrate nonphysician EMS interpretation of STEMI on prehospital ECG has excellent sensitivity and high negative predictive value. This finding supports the use of prehospital ECGs interpreted by EMS to help identify and facilitate treatment of STEMI. These results may have broad implications on staffing models for first responder/EMS units.Copyright © 2012 Canadian Cardiovascular Society. Published by Elsevier Inc. All rights reserved.

32 keywords...

hide…