An Assessment of the Prognostic Value of Exercise Treadmill Testing Utilizing the Duke Score in Risk Stratification of Emergency Department Chest Pain Unit Patients

OBJECTIVES: Exercise treadmill testing (ETT) has been a standard for evaluating outpatients at risk for cardiovascular (CV) events. Few studies demonstrate the prognostic utility of ETT in an emergency department chest pain unit (CPU) or have used the Duke score (DS) [exercise duration - 5x ST deviation - 4x treadmill angina index] to grade ETT performance. The authors assess the utility of the DS in the CPU to predict CV events. METHODS: This retrospective study enrolled consecutive CPU patients from November 2000 to October 2001 with chest pain. Patients were required to reside in a ten county area surrounding the tertiary referral center and grant research authorization. They were retrospectively followed for one year to identify CV events defined as death, myocardial infarction, unstable angina, acute congestive heart failure, stroke, arrhythmia, and cardiac revascularization. Those undergoing ETT were stratified into low (DS >/= 5) and moderate/high (DS < 5) risk groups. A logistic regression model was fit to evaluate differences in risk of having a CV event within one year between the DS >/= 5, DS < 5, and no ETT groups. RESULTS: During the study period, 1048 patients entered the CPU with 800 meeting inclusion criteria. Of these, 599 patients received an ETT and 201 had contraindications or failed the CPU protocol prior to stress testing. The CV event rates were 1.5% (7/454), 17.9% (26/145), and 19.9% (40/201), respectively, among those with DS >/= 5, DS < 5, and those who did not undergo an ETT. Patients who did not undergo ETT were 10.5 times (95% CI, 4.4-24.7) more likely to experience a CV event compared to patients who had a DS >/= 5. Patients with DS < 5 were 11.2 times (95% CI, 4.7-26.8) more likely to have a CV event. CONCLUSIONS: Patients with DS >/= 5 developed minimal CV events when compared to those with DS < 5. The DS appears to be effective in risk stratification of chest pain patients entered into the CPU.