Emergency Physician Interpretation of Prehospital, Paramedic-Acquired Electrocardiograms

Hypothesis: Emergency physician interpretation of prehospital, paramedic-acquired, electrocardiograms (ECG) is accurate judged by comparison with that of a reference cardiologist. Methods: Twelve-lead ECGs were obtained by paramedics in the field from 150 patients with acute chest pain. The ECGs were transmitted by cellular telephone to a central location. Each ECG was assessed for evidence of acute myocardial infarction (AMI) by: 1) a third-year, emergency medicine resident (EMP-R); 2) a residency-trained, board-certified, emergency physician (EMP-RT); 3) an emergency physician board certified under the practice option (EMP-PT); and 4) a board-certified cardiologist. Agreement between each emergency physician and the cardiologist was assessed by the kappa statistic. Hospital records were reviewed for final diagnosis of each patient. Results: Sixteen of 150 (10.7%) patients received a hospital discharge diagnosis of AMI. Sensitivity of physician interpretation ranged from 0.31 to 0.56. All physicians achieved specificity of 0.99. Falsepositive rates for the physicians ranged from 0.18—0.29. The mean positive predictive value for the four physicians was 0.77+0.05; the mean negative predictive value was 0.94±0.01. The total agreements between the EMP-R, EMP-RT, and EMP-PT and the cardiologists were 0.97, 0.96, and 0.97, respectively. Kappa values for agreement between the emergency physicians and the cardiologist ranged from 0.65-0.79. Conclusions: Residency-trained or board-certified emergency physician interpretations of prehospital, paramedic-acquired 12-lead ECGs show a high degree of agreement with reference cardiologist interpretations.