The relative lack of sensitivity and specificity of electrocardiography (ECG) and the serum creatine kinase-MB (CK-MB) level in acute coronary syndromes results in a large number of unnecessary admissions to coronary care units. Conversely, developing myocardial infarction is undetected in some patients, and these patients may be discharged from the emergency department. Hamm and colleagues studied the diagnostic value of the cardiac-specific contractile proteins troponin T and troponin I for the early triage of patients who visited an emergency department for evaluation of acute chest pain.

Patients included in the study presented to the emergency department because of acute anterior, precordial or left-sided chest pain of less than 12 hours' duration. The pain could not be explained by obvious trauma or by abnormalities on a chest radiograph. Patients were excluded if ST-segment elevation was apparent on ECG or if they had had an acute myocardial infarction within the previous two weeks. Blood was drawn within 15 minutes of arrival at the hospital for measurement of troponin T and troponin I. The tests were repeated four hours later and, if the patient presented with pain of less than two hours' duration, the troponin assays were repeated a third time at six hours after the onset of chest pain. The tests were performed by a trained assistant in the emergency department, and the results were immediately available to the treating physician.

A total of 773 patients were enrolled in the study, and 47 (6.1 percent) were diagnosed as having a myocardial infarction on the basis of routine measurements of CK-MB within 24 hours of presentation. Of these 47 patients, 24 (51 percent) had positive results on the troponin T test at the time of arrival in the emergency department and 44 (94 percent) had positive results four hours later. The troponin I results were positive in 31 (66 percent) of these patients on arrival, and all of the 47 patients had positive test results four hours later. Measurements of CK-MB were elevated in 25 (53 percent) of these patients on arrival in the emergency department and in 43 (91 percent) of these patients four hours later.

Unstable angina was the admitting diagnosis in 315 patients. Seventy patients (22.2 percent) had at least one positive troponin T measurement, and 114 (36.2 percent) had at least one positive troponin I value. CK-MB elevation at any time was found in only 16 (5.1 percent) of these patients.

Troponin assays were negative in 286 patients who were not admitted. Seven of these patients were subsequently found to have a positive result for troponin I on repeat testing at a separate laboratory. Two of these patients had adverse cardiac events soon after discharge, including a nonfatal myocardial infarction five days later and a presumed cardiac death 23 days later.

The authors conclude that troponin T and troponin I levels are highly sensitive for the diagnosis of myocardial infarction (94 percent for troponin T and 100 percent for troponin I). However, the diagnostic specificity for myocardial infarction was low: 22 percent of the patients with unstable angina had positive results for troponin T, and 36 percent had positive results for troponin I. For patients with two negative troponin values at least four hours apart, the likelihood of an acute coronary event was very low. Troponin T and troponin I tests result in more accurate diagnoses than previous, more time-comsuming methods