Comparison of four Doppler echocardiographic methods for calculating pulmonary-to-systemic shunt flow ratios in patients with ventricular septal defect

Several methods currently exist for quantifying pulmonary (Qp) and systemic (Qs) shunt flow using Doppler echocardiography, although none is widely utilized. In this study, 39 patients (age 2 months to 12 years, mean 2.1 years) underwent an echocardiographic examination within 1 month of Qp:Qs shunt flow determination by oximetry at catheterization. Qp:Qs was determined by 4 methods that utilized: (1) velocity time integrals and luminal areas to estimate volume flow of the pulmonary artery and aorta; (2) the square of the ratios of pulmonary artery to aorta, multiplied by the ratio of pulmonary to aortic peak flow velocities; (3) inclusion of mitral and tricuspid valve volume flow to pulmonary and aortic volume flow; and (4) ventricular septal defect (VSD) diameter and velocity time integral to calculate left-to-right shunt, which, when added to aortic volume flow (Qs), can be used to estimate Qp. Each of the first 3 methods was statistically correlated to the oximelry Qp:Qs, with r values ranging from 0.54 to 0.66 (p < 0.001). However, the fourth method, based on direct computation of flow across the VSD, had the best correlation to catheterization data (r = 0.82), and further improved when 7 patients with a large VSD (>9 mm/m2), all of whom had bidirectional shunting, were removed (r = 0.90). Thus, we concluded that this latter method demonstrated the best correlation to catheterization-derived shunt flow data, and because this method is somewhat less labor-intensive than the others, should provide clinically useful data well suited for serial evaluation in infants and children with VSD.