Objectives: The aim of this study was to analyze if contrast-enhanced echocardiography (CEE) is as reliable as lung perfusion scintigraphy (LPS) to detect intrapulmonary shunting (IPS) in children with portal hypertension (PHTN) or congenital/surgical portosystemic shunts (PSS) and to define the number of cardiac cycles required to exclude intrapulmonary shunting.

Methods: Inclusion criteria for this cross-sectional study were: (1) presence of PHTN or PSS diagnosed on abdominal ultrasound, (2) technically valid saline contrast echocardiography, (3) lung perfusion scintigraphy within 6 months of CEE. The number of cardiac cycles between right atrial opacification and the arrival of contrast in the left atrium were counted. We analyzed our CEE data at three and five cardiac cycles and compared them with LPS results.

Results: The study population was composed of 78 children (38 girls, 49%) ages 2.1-18.8 years (mean 9.8). Sixty-nine patients had PHTN (88%), and nine had a PSS (11%). Eleven subjects (14%) presented evidence of IPS on LPS. Peripheral oxygen saturation was lower in the subjects with IPS detected on LPS (95.3 ± 1.7% vs 99.0 ± 1.4%; P < 0.01). Comparison of LPS with CEE before three and five cardiac cycles showed that CEE is highly specific (95.7%) as early as three cardiac cycles with markedly better sensitivity (72.7%) when using five cardiac cycles. Furthermore, a negative study using five cardiac cycles ruled out IPS with a 95% negative predictive value. The cardiac cycle at which the bubbles appeared in the left atrium was inversely correlated to the shunt index measured using LPS (r = -0.563; P = 0.001).

Conclusion: CEE is sufficient for the screening of IPS in children with PHTN or congenital/surgical PSS, obviating the need for LPS.