STUDY OBJECTIVES: This study tested the hypothesis that serial creatine phosphokinase (CK)-MB sampling in the emergency department can identify acute myocardial infarction (AMI) in patients presenting to the ED with chest pain and nondiagnostic ECGs. DESIGN: Patients more than 30 years old who were evaluated initially in the ED and hospitalized for chest pain were studied. Serial CK-MB levels were analyzed prospectively using a rapid serum immunochemical assay for identification of AMI patients in the ED. Presenting ECGs showing new, greater than 1-mm ST elevation in two or more contiguous leads were considered diagnostic for AMI. All other ECGs were considered nondiagnostic ECGs. CK-MB levels were determined at ED presentation and hourly for three hours (total of four levels). Patients with at least one level of more than 7 ng/mL were considered to have a positive enzyme study. The in-hospital diagnosis of AMI was determined by the development of typical serial ECG changes or separate standard cardiac enzyme changes after admission. SETTING: Eight tertiary-care medical center hospitals. METHODS AND MAIN RESULTS: Of the 616 study patients, 108 (17.5%) were diagnosed in the hospital as AMI; 69 of these AMI patients (63.9%) had nondiagnostic ECGs in the ED. Of the patients with nondiagnostic ECGs, 55 (sensitivity, 79.7%) had a positive ED serial CK-MB enzyme study within three hours after presentation. Combining serial ED CK-MB assay results with diagnostic ECGs yielded an 88.4% sensitivity for AMI detection within three hours of ED presentation. The predictive value of a negative serial ED enzyme study for no AMI was 96.2% (specificity, 93.7%). CONCLUSION: Serial CK-MB determination in the ED can help identify AMI patients with initial nondiagnostic ECGs. Use of serial CK-MB analysis may facilitate optimal in-hospital disposition and help guide therapeutic interventions in patients with suspected AMI despite a nondiagnostic ECG.
STUDY OBJECTIVES: This study tested the hypothesis that serial creatine phosphokinase (CK)-MB sampling in the emergency department can identify acute myocardial infarction (AMI) in patients presenting to the ED with chest pain and nondiagnostic ECGs. DESIGN:Patients more than 30 years old who were evaluated initially in the ED and hospitalized for chest pain were studied. Serial CK-MB levels were analyzed prospectively using a rapid serum immunochemical assay for identification of AMI patients in the ED. Presenting ECGs showing new, greater than 1-mm ST elevation in two or more contiguous leads were considered diagnostic for AMI. All other ECGs were considered nondiagnostic ECGs. CK-MB levels were determined at ED presentation and hourly for three hours (total of four levels). Patients with at least one level of more than 7 ng/mL were considered to have a positive enzyme study. The in-hospital diagnosis of AMI was determined by the development of typical serial ECG changes or separate standard cardiac enzyme changes after admission. SETTING: Eight tertiary-care medical center hospitals. METHODS AND MAIN RESULTS: Of the 616 study patients, 108 (17.5%) were diagnosed in the hospital as AMI; 69 of these AMI patients (63.9%) had nondiagnostic ECGs in the ED. Of the patients with nondiagnostic ECGs, 55 (sensitivity, 79.7%) had a positive ED serial CK-MB enzyme study within three hours after presentation. Combining serial ED CK-MB assay results with diagnostic ECGs yielded an 88.4% sensitivity for AMI detection within three hours of ED presentation. The predictive value of a negative serial ED enzyme study for no AMI was 96.2% (specificity, 93.7%). CONCLUSION: Serial CK-MB determination in the ED can help identify AMI patients with initial nondiagnostic ECGs. Use of serial CK-MB analysis may facilitate optimal in-hospital disposition and help guide therapeutic interventions in patients with suspected AMI despite a nondiagnostic ECG.