Methods
A prospective study was performed on 100 consecutive patients less than 18 years old who were undergoing either CT or MRI of the chest for any indication. These advanced imaging studies served as gold standard for diagnosis of AoA anatomy. A priori power analysis was completed assuming null proportion of successful tracheal imaging of 85% and demonstrated power > 0.95 with a sample size of 100 when proportion was 97%. If a more conservative null proportion of 90% was used, power remained significant at > 0.80 with when proportion was 97%.
Four total echocardiographic clips in two separate planes were obtained for each patient in 2D and color Doppler with Phillips IE33 or Phillips EPIQ echocardiography machines (Philips Electronics, Washington, USA). In all patients, a right-to-left sagittal sweep was performed with echocardiography from the suprasternal notch or high right parasternal view and used to categorize 1) Left AoA = right superior vena cava (SVC)-trachea-AoA, Figure 1, Video 1; 2) Right AoA = SVC-AoA-trachea, Video 2; 3) DAA = SVC-AoA-trachea-AoA, Video 3. A parasternal short-axis view at the base of the heart showing the PA bifurcation straddling the trachea was used to rule out PA sling, Video 4. No additional echocardiography images were obtained as a part of this study. Study approval was granted by the University of Iowa institutional review board. Age appropriate consent/assent was obtained on the day of advanced imaging to participate in additional research-only echocardiographic images completed either directly before or after CT/MRI.
Patient acuity, use of sedation or anesthesia, imager experience, and subjective ease of obtaining echo were collected. Independent review for AoA sidedness, DAA, and absence of PA sling was performed by 2 advanced-imaging pediatric cardiologists for all echocardiography images and advanced imaging scans. Those obtaining echo images and those reading images were blinded to any previous diagnoses. Kappa coefficient was calculated to determine inter-reader reliability with independent read, and then in cases of initial disagreement, consensus was attained by joint review. Continuous variables were reported as median and range. Categorical variables were reported as percent. The proportion of successful imaging of the trachea and diagnostic accuracy with tracheal sweep were calculated as mean percent values with 95% confidence interval (CI). A one-tailed z-test was used to assess for statistical significance. These calculations were repeated for successful imaging of tracheal-PA bifurcation relationship for both the entire cohort and then again excluding patients who had previous surgical intervention on their PA (ie: Glenn/Fontan, RV-to-PA conduit, LeCompte).