Triple rule-out computed tomography for risk stratification of patients with acute chest pain

Abstract

AIMS: Clinical evidence supporting triple rule-out computed tomography (TRO-CT) for rapid screening of cardiovascular disease is limited. We investigated the clinical value of TRO-CT in patients with acute chest pain. METHODS: We retrospectively enrolled 1024 patients who visited the emergency department (ED) with acute chest pain and underwent TRO-CT using a 128-slice CT system. TRO-CT was classified as "positive" if it revealed clinically significant cardiovascular disease including obstructive coronary artery disease, pulmonary thromboembolism, or acute aortic syndrome. The clinical endpoint was occurrence of a major adverse cardiovascular event (MACE) within 30 days, defined by a composite of all cause death, myocardial infarction, revascularization, major cardiovascular surgery, or thrombolytic therapy. Clinical risk scores for acute chest pain including TIMI, GRACE, Diamond-Forrester, and HEART were determined and compared to the TRO-CT findings. RESULTS: TRO-CT revealed clinically significant cardiovascular disease in 239 patients (23.3%). MACE occurred in 119 patients (49.8%) with positive TRO-CT and in 7 patients (0.9%) with negative TRO-CT (p < 0.001). Sensitivity, specificity, positive predictive value, and negative predictive value of TRO-CT was 95%, 88%, 54%, and 99%, respectively. TRO-CT was a better discriminator between patients with vs. without events as compared to clinical risk scores (c-statistics = 0.91 versus 0.64 to 0.71: integrated discrimination improvement = 0.31 to 0.37: p < 0.001 for all comparisons). Patients with a negative TRO-CT showed shorter ED stay times and admission rates compared to patients with positive TRO-CT, irrespective of clinical risk scores (p < 0.001 for all comparisons). CONCLUSION: Triple rule-out CT has high predictive performance for 30-day MACE and permits rapid triage and low admission rates irrespective of clinical risk scores.