Development of pulmonary arteriovenous fistulas in patients with cavopulmonary anastomosis may result in a significant morbidity. Although the use of bubble contrast echocardiography with selective injection into both the branch pulmonary arteries in identifying pulmonary arteriovenous fistulas has been increasing, the actual efficacy of this diagnostic modality has not been properly evaluated. Thus, this study aimed to assess the efficacy of bubble contrast echocardiography in detecting pulmonary arteriovenous fistulas in children with total cavopulmonary connection.

A total of 140 patients were included. All patients underwent cardiac catheterisation. Bubble contrast echocardiographic studies were performed by injecting agitated saline solution into the branch pulmonary arteries. Transthoracic echocardiograms that use an apical view were conducted to assess the appearance of bubble contrast in the systemic ventricles. Then, the contrast echocardiogram results and other cardiac parameters were compared.

No correlation was found between contrast echocardiogram grade and other cardiac parameters, such as pulmonary capillary wedge saturation and pulmonary artery resistance. Moreover, only 13 patients had negative results on both the right and left contrast echocardiograms, and 127 of the 140 patients had positive results on contrast echocardiograms even though they had normal pulmonary capillary wedge saturation. Results showed that bubble contrast echocardiography was a highly sensitive method and was likely to obtain false-positive results.

Bubble contrast echocardiography might be highly false positive in detecting pulmonary arteriovenous fistulas in patients with cavopulmonary anastomosis. We have to consider how we make use of this method. Further standardisation of techniques is required.