ABSTRACT
OBJECTIVE: Clinical assessment often cannot risk stratify patients hospitalized with chest pain and non-diagnostic electrocardiography (ECG) or myocardial enzymes. An inappropriate admission of patients with non-cardiac chest pain is an enormous cost factor. METHODS: 2315 patients who presented in the chest pain unit (CPU) with symptoms suggestive of acute coronary syndrome (ACS) were screened. All patients with relevant changes in ECG or myocardial enzymes were excluded. 268 consecutive patients (mean 58 ± 7 years, 88 men) were prospectively included and underwent echocardiography for left ventricular ejection fraction (LVEF), wall motion score index (WMSI) and strain parameter and a coronary angiography (CA) within 2 ± 1 days after admission. RESULTS: Anatomically obstructive coronary artery disease (CAD) (≥70 % diameter stenosis) was present in 110 patients (41 %). The incremental value of LVEF, WMSI, and strain parameters to relevant clinical variables was determined in nested Cox models. Baseline clinical data associated with relevant CAD were age [hazard ratio (HR) 1.31, p = 0.03], arterial hypertension (HR 1.39, p = 0.03) and diabetes (HR 1.46, p = 0.001). The addition of endocardial global circumferential strain (GCS) (HR 1.57, p < 0.001) caused the greatest increment in model power (χ (2) = 43.4, p < 0.001). Optimal cut-off value was calculated as -21.7 % for GCS (sensitivity 87 %, specificity 76 %) to differentiate between these patients. CONCLUSIONS: In patients with suspected ACS but without ECG changes or myocardial enzyme abnormalities, myocardial deformation imaging can identify patients at risk. This approach may be applied to improve decision guidance at the CPU for fast discharge of patients with non-cardiac chest pain or prompt cardiological allocation of patients with CAD. CLINICAL TRIAL REGISTRATION: NCT 02357641.