GRACE and TIMI risk scores but not stress imaging predict long-term cardiovascular follow-up in patients with chest pain after a rule-out protocol

OBJECTIVE To determine the long-term prognostic value of stress imaging and clinical risk scoring for cardiovascular mortality in chest pain patients after ruling out acute coronary syndrome (ACS). METHODS A standard rule-out protocol was performed in emergency room patients with a normal or non-diagnostic admission electrocardiogram (ECG) within 6 h of chest pain onset. ACS patients were identified by troponin T, recurrent angina and serial ECG. Dobutamine stress echocardiography (DSE) was performed after ACS was ruled out. Myocardial perfusion scintigraphy (MPS) was performed within 6 months in an outpatient setting according to the physician's discretion. RESULTS 524 patients were included. GRACE and TIMI risk scores were 75 (57-96) and 1 (0-2) in the rule-out ACS group, and 89 (74-107) and 2 (1-3) in the ACS group, respectively (median, interquartile range). Follow-up (median 9.4 (8.9-10.0) years) was complete in 96%. 350 of 379 rule-out ACS patients had an interpretable DSE and 52 patients underwent an MPS. 21 of the rule-out ACS patients (6%) died of a cardiovascular cause compared with 24 (17%) ACS patients (p < 0.001). For rule-out ACS patients, C-statistics were 0.829 and 0.803 for the GRACE and TIMI scores. In these patients, DSE and MPS outcome did not predict long-term cardiovascular mortality. In multivariate analysis, known chronic heart failure, ACE inhibitor use, and GRACE score were independent predictors of cardiovascular mortality. CONCLUSIONS TIMI and GRACE score but not DSE and MPS are accurate predictors of long-term cardiovascular mortality, even in chest pain patients with a normal or non-diagnostic electrocardiogram undergoing a rule-out protocol.