Impact of computed tomography (CT)-derived fractional flow reserve on reader confidence for interpretation of coronary CT angiography

Abstract

PurposeTo assess the impact of computed tomography-derived fractional flow reserve (FFRCT) on reader confidence and reader time for interpretation of coronary computed tomography angiography (CCTA). Material and methodsThis IRB-approved, HIPAA-compliant, consent waivered, quality-improvement study included 50 patients (23 women, age 67 ± 12years, body mass index 28.7 ± 5.3 kg/m2). CCTA was acquired on 2nd and 3rd generation dual-source MDCT with use of beta-blockers and nitroglycerin, and FFRCT was calculated (HeartFlow,Redwood City,CA). Two readers with experience level COCATS2 (Core- Cardiology-Training-Symposium) and two with COCATS3 assessed severity of epicardial coronary artery disease (CAD) using CCTA alone and CCTA with FFRCT. Reader confidence for CAD and hemodynamically significant stenosis (HS) was rated on a 4-point Likert-scale (1=high, 2=good, 3=limited, 4=none) for the four major epicardial coronary artery (CA) vessels. Time to interpret was recorded. ResultsThe severity of CAD in the cohort population was CAD-RADS1 (Coronary-Artery-Disease Reporting-and-Data-System) n = 15, CAD-RADS2 n = 8, CAD-RADS3 n = 11, and CAD-RADS4A n = 16. Sixty-three CA in 30 patients had minimal FFRCT values ≤0.8. Reader confidence when using FFRCT increased for CAD and HS (p = 0.0001) with a reduction of “non-confident” patient reads (rank 3&4) by 27% and 75%, respectively. The change in confidence was not associated with CAD-RADS (p = 0.1) but correlated with reader experience (p < 0.001). The median time-to-read a CCTA study decreased by 5 min when FFRCT was available (p < 0.001). ConclusionInterpretation of CCTA in conjunction with FFRCT improved reader for assessment of severity of CAD and HS with reduction of “non-confident” reads and decreased the median time-to-interpretation of a CCTA.