Study objectives: To test the short- and long-term performance of a simple algorithm to detect or rule out acute coronary artery disease (CAD) in patients referred with acute chest pain (CP).
Design: Prospective, multicenter study with 30-day and 1-year follow-ups.
Setting: Emergency and coronary care units of two university hospitals and four affiliated teaching hospitals in a suburban region of northern Switzerland.
Patients: One hundred sixty-one consecutive patients referred with acute CP with a high CAD risk profile underwent exercise testing (ET) within 24 h if results of two troponin tests and ECG remained normal within 6 h. The patients were discharged if ET results were negative or underwent angiography if the results were positive.
Measurements and results: The primary end point was survival free of death, myocardial infarction (MI), and acute coronary syndrome, as well as the correct diagnosis of CAD in need of revascularization. Failure of the algorithm included events not predicted and false-positive ET results. Symptom-limited ET was performed in 142 patients, and stress imaging was performed in 17 patients. Discharge within 24 h was possible in 76%. Angiography in 21 patients showed no CAD in 4 patients, whereas revascularization was necessary in 17 patients. During follow-up, three patients had an MI (one fatal) and two patients had unstable angina, for an event rate of 3.1%/yr. Overall sensitivity of the algorithm to detect CAD was 71%, whereby six of seven false-positive ET results were due to exercise-induced CP without ECG changes. More importantly, the negative predictive value was 96.4%, indicating that this simple algorithm failed to identify only 3.6% of patients at risk for future events.
Conclusions: Discharge within 24 h after onset of acute CP in high-risk CAD patients is safe, irrespective of CAD history, if ECG findings remain unchanged and troponin test results are normal within 6 h, and if results of adequate ET are negative.