RESUMO
BACKGROUND: Cardiac catheterization is the gold-standard modality for investigation of cardiovascular morphology before bidirectional cavopulmonary connection, but requires general anaesthesia and is associated with procedural risk. AIMS: To assess the diagnostic accuracy and safety of computed tomography in diagnosing great vessel stenosis/hypoplasia compared with cardiac catheterization and surgical findings. METHODS: Twenty-seven patients (10 after Norwood stage I) underwent computed tomography before surgery between January 2010 and June 2016; 16 of these patients also underwent cardiac catheterization. Proximal and distal pulmonary artery, aortic isthmus and descending aorta measurements, radiation dose and complications were compared via Bland-Altman analyses and correlation coefficients. RESULTS: The accuracy of computed tomography in detecting stenosis/hypoplasia of either pulmonary artery was 96.1% compared with surgical findings. For absolute vessel measurements and Z-scores, there was high correlation between computed tomography and angiography at catheterization (r=0.98 for both) and a low mean bias (0.71mm and 0.48; respectively). The magnitude of intertechnique differences observed for individual patients was low (95% of the values ranged between -0.9 and 2.3mm and between -0.7 and 1.7, respectively). Four patients (25%) experienced minor complications from cardiac catheterization, whereas there were no complications from computed tomography. Patients tended to receive a higher radiation dose with cardiac catheterization than with computed tomography, even after exclusion of interventional catheterization procedures (median 2.5 mSv [interquartile range 1.3 to 3.4 mSv] versus median 1.3 mSv [interquartile range 0.9 to 2.6 mSv], respectively; P=0.13). All computed tomography scans were performed without sedation. CONCLUSIONS: Computed tomography may replace cardiac catheterization in identification of great vessel stenosis/hypoplasia before bidirectional cavopulmonary connection when no intervention before surgery is required. Computed tomography carries lower morbidity, can be performed without sedation and may be associated with less radiation.