Assessment of inter- and intrareader agreement among different stenosis quantification methods in coronary CT angiography

PURPOSE: To assess the inter- and intrareader agreement among different subjective and objective measurement methods to grade coronary artery stenosis in coronary CT angiography (CCTA). METHODS: score of 3 or 4 were included. Five different radiologists (three subspecialized in cardiac imaging) independently graded the most severe coronary artery stenosis using different measurement methods: visual stenosis grading and objective stenosis grading measuring stenosis diameter and area. Interreader agreement of percentage stenosis were calculated for each measurement method and stratified by the coronary calcium score (CACS) using Krippendorf α. Intrareader assessment was performed for one subspecialized radiologist. RESULTS: A total of 61 patients were included (mean age, 65 ± 10 years; 18 women). Interreader agreements for percentage stenoses as determined visually and as measured by diameter and area were moderate independent of measurement method (0.42, 0.46, 0.41 for visual, diameter and area percentage stenoses, respectively) and did not depend on subspecialization. Intrareader agreement was moderate to substantial with values of 0.67, 0.57 and 0.53 for visual, diameter and area percentage stenoses. There was a trend towards lower inter- and intrareader agreement for stenoses with high CACS (e.g. for visual grading: 0.44 vs. 0.2 for interreader agreement between non-calcified stenoses (CACS = 0) and calcified stenoses (CACS ≥ 100)). CONCLUSION: Inter- and intrareader agreement for stenosis quantification in coronary CT angiography was overall moderate and did not differ significantly among visual and objective measurement methods. A trend towards lower agreement was observed for stenoses with higher CACS.