Published online by Cambridge University Press: 28 June 2012
To determine the reliability of ST-segment interpretation by paramedics from lead-II rhythm strips obtained in the prehospital setting.
Prospective, blinded study of 127 patients transported by an urban/rural emergency medical services system with complaints consistent with ischemic heart disease.
Emergency department physicians asked emergency medical technician-paramedics (EMT-P) via radio to evaluate ST-segments for elevation or depression and grade it as “mild,” “moderate,” or “severe.” Then, this rhythm strip was interpreted blindly by emergency physicians who also interpreted the lead-II obtained from a 12-lead electrocardiogram (ECG) obtained in the emergency department (ED). The field interpretation was compared with the subsequent readings and the final in-patient diagnosis using positive predictive value (PPV), negative predictive value (NPV), and the Kappa statistic. Markedly discrepant interpretations were analyzed separately.
Using physician interpretation as the reference standard, paramedic interpretation of the lead-II ST-segments obtained in the prehospital setting was correct (within ±1 gradation) in 113 out of 127 total cases (89%). Of 105 patients for whom final hospital diagnosis was available, the ST-segment on the rhythm strip obtained in the prehospital setting, had a positive predictive value of 74% and a negative predictive value of 85% for myocardial ischemia or myocardial infarction (MI) (p <0.001, Kappa = 0.59). Discordant interpretations between the paramedics and emergency physicians often were related to a basic misunderstanding of rhythm strip morphology.
Field interpretation of ST-segments by paramedics is fairly accurate as judged both by emergency physicians and correlation with final patient outcome, but its clinical utility is unproved. A small but clinically significant number of outliers, consisting of markedly discrepant false positives, reflects paramedic uncertainty in identifying the deviations of the ST-segment.