Return To Abstract Listing

Cardiac Imaging

E2807. Accuracy of Coronary CT Angiography in Stenosis Assessment Compared to the Reference Standard

Wahl D,  Goykhman I,  Hochhold J,  Son H,  Corral A. Einstein Medical Center, Philadelphia, PA

Address correspondence to D. Wahl (

Objective: Chest pain accounts for as many as 5 million emergency department (ED) visits each year. Low- to intermediate-risk patients with atypical symptoms presenting to the ED are increasingly being referred for CT angiography (CTA) of the coronary arteries, with the intention of reliably ruling out coronary artery disease. Previous studies have demonstrated strong sensitivity and negative predictive value of coronary CTA. The aim of this study is to compare the extent of coronary stenosis found on coronary CTA with the extent of stenosis found on cardiac catheterization in a cohort of patients who present through the emergency department. In analyzing 3 years ED coronary CTA studies, our goal is to evaluate how accurately CTA can predict the degree of coronary artery stenosis when compared to catheterization.

Materials and Methods: Two hundred coronary artery CTA studies were performed through the emergency room or emergency room observation unit between October 3, 2013 and May 26, 2016 at two of our three medical centers. A retrospective search was performed to establish which patients subsequently underwent catheterization within 1 month. Catheterization reports were obtained and reported stenoses were compared with CTA interpretation. Estimates of stenosis from the original CTA reports were used. Five cardiologists interpreted the original catheterization reports, and two radiologists interpreted the original CTA reports. The estimated percentage stenosis was documented for each study. Good correlation was considered any CTA stenosis estimate which was reported as within 10 percentage points of the estimate on catheterization. For any stenosis greater than 70% on catheterization, an estimate of greater than 70% on CTA was considered good correlation regardless of the absolute difference.

Results: In total, 43 of 200 patients underwent subsequent catheterization. One patient was excluded due to motion degradation of the CTA examination. Of the 42 evaluated patients who underwent subsequent coronary catheterization, there were 43 specific stenoses described on cardiac catheterization. The degree of stenosis ranged from 20% to greater than 70%. The extent of stenosis in 26 of the 43 lesions (60.5%) was within 10% of the estimated degree of narrowing on coronary CTA. No cases of false negative CTA results were observed, corresponding to a negative predictive value of 100%. No cases with stenosis greater than 30% were reported as normal on CTA. For stenoses greater than 30%, the sensitivity was measured to be 100%.

Conclusion: Coronary CTA has been established as a sensitive and reliable examination to use in the emergency department in low- to intermediate-risk patients. Assuming catheterization remains the reference standard for stenosis assessment, coronary CTA is accurate in detecting stenoses greater than 30%.