In-hospital shunt occlusion in infants undergoing a modified blalock-taussig shunt

The clinical variables leading to postoperative thrombotic occlusion of a modified Blalock-Taussig shunt (mBTS) remain elusive. In this investigation, we assess several perioperative variables to determine associations with postoperative in-hospital shunt occlusion.

We retrospectively reviewed the medical records of infants receiving a mBTS as a first operation between March 1, 2005, and December 31, 2011. Numerous perioperative variables were collected, focusing on those that would increase resistance to blood flow through the shunt or alter coagulation.

In all, 207 neonates fit our criteria. In-hospital shunt occlusion occurred in 14 patients (6.8%); 3 patients (21.4%) subsequently died during their hospitalization. Pulmonary atresia/ventricular septal defect with or without major aortopulmonary collateral arteries was the most common diagnosis associated with shunt occlusion (57.1%). Of the collected perioperative variables, pulmonary artery size was significantly associated with shunt occlusion (p = 0.03). Preoperative coagulation values were significantly reduced in those patients who experienced shunt occlusion. Additionally, the immediate postoperative activated partial thromboplastin time was significantly reduced in the occlusion group although values remained well above normal as all infants were treated postoperatively with a heparin infusion.

We found that a patient's anatomy (pulmonary atresia/ventricular septal defect with or without major aortopulmonary collateral arteries) and the size of the pulmonary artery being shunted had a significant impact in predicting postoperative in-hospital shunt occlusion. These results emphasize that technical skills and a low resistance to blood flow are necessary for successful shunt function. Although some perioperative coagulation values were significantly reduced in infants who were destined to experience shunt occlusion, they would be difficult to detect clinically.