On May 2, 1994, Giddins presented Marietess's case at a CVT conference. In Odim's opinion, Marietess was at the point of outgrowing her shunt, which was creating a volume overload on her right ventricle. He suggested a bilateral, bi-directional, cavo-pulmonary shunt, with removal of the existing B-T shunt and an atrial septectomy. This surgery is a Fontan-type procedure, sometimes referred to as a bilateral Glenn procedure or a Kawashima operation.

In the procedure that Odim proposed, the right superior vena cava was to be surgically connected to the right branch of the pulmonary artery, instead of to the right atrium. The left superior vena cava was to be surgically connected to the left branch of the pulmonary artery, instead of entering the coronary sinus. It is called bilateral because it involves the superior vena cavas on both the left and right sides of the body. It is called bi-directional because the blood from the right superior vena cava flows to both the right and left lungs. It is called a cavopulmonary shunt because it was to connect the vena cavas with the pulmonary arteries.

All the oxygen-depleted blood that normally drained into the right atrium would, in this case, be draining through both of the superior vena cavas into the lungs. The right superior vena cava was attached to the right pulmonary artery and the left superior vena cava was attached to the left pulmonary artery. In addition, a septostomy was performed to remove the remnants of the septal wall between the right and left atria. The result was that blood flow from her body would continue to drain into the right superior vena cava, which would also drain the right upper body. The left side of her upper body would continue to drain into the left superior vena cava. Therefore most of the venous or blue blood, which would normally be directed through the right side of the heart, would flow through the pulmonary arteries to the lungs by gravity, without the benefit of any pumping action from the right ventricle.

According to Dr. Glenn Taylor, the success of the operation depended on two things. First, there had to be low resistance to blood flow in the pulmonary arteries. Second, there had to be maintenance of a satisfactory gradient of pressure between the central venous pressure and the pressure within the common atrial chamber, which was also receiving the venous drainage from the lungs. The central venous pressure was the driving force for perfusion of the lungs, in comparison to systemic arterial pressure being the driving force where there was a Blalock-Taussig shunt (Exhibit 336, page 8.1).

While the pressures in a heart are always important, in this repair the pressures are particularly important. This is because the patient's blood circulation depends on a single pump, the right ventricle, to send the blood through both the body and the lungs. If there is any backup of blood flow, because pressures downstream are elevated, the child's heart will quickly start to fail. A variety of factors, including medications, blood clots, and suturing, can work independently or jointly to raise or lower pressures in the veins, with dramatic and potentially fatal impact.