On the morning of Tuesday, September 13, Marietess underwent an operation in which:

• a bilateral and bi-directional caval pulmonary anastomosis was created (this term refers to the joining of the superior vena cavas to the pulmonary arteries);
• the opening in the septal wall between her two atria was enlarged;
• her Blalock-Taussig shunt was removed; and
• her patent ductus arteriosus was closed.

Marietess Tena Capili - pre-operative heart

Diagram 8.2 Marietess Tena Capili - post-operative heart 1 - Suture narrowing of right superior vena cava (discovered at autopsy) 2 - Connection of right superior vena cava to right pulmonary artery 3 - Divided right superior vena cava 4 - Surgical enlargement of opening in septal wall between atria 5 - Suture narrowing of left superior vena cava (discovered at autopsy) 6 - Takedown of left Blalock-Taussig shunt 7 - Connection of left superior vena cava to left pulmonary artery 8 - Divided left superior vena cava

The surgical team is set out in the accompanying chart.

TABLE 8.1: Persons involved in the operation on Marietess Tena Capili, September 13, 1994
OR team member		Persons involved
Surgeon		J. Odim
Surgical assistant		S. Sharpless
Anaesthetists		J. Swartz, R. Debrouwere (resident)
Scrub nurses		C. Weber, S. Scott
Circulating nurses		C. Youngson, B. Zulak
Perfusionists		D. Smith, C. McCudden

TABLE 8.2: Length of phases of the operation on Marietess Tena Capili, September 13, 1994
Phase of the operation		Time taken
Induction		2 hours 10 minutes
Bypass		4 hours 34 minutes
Aortic cross-clamp		22 minutes
Total surgical time		9 hours 1 minute
Total operating-room time		11 hours 56 minutes

There were four significant events during surgery. They were:

• Cannulation issues;
• Problems with distension of the heart;
• Coagulopathy and surgical bleeding; and
• Superior vena cava syndrome.

Cannulation issues

Before Marietess went on bypass, a problem with cannulation gave rise to bleeding and a fall in blood pressure. Youngson testified that it appeared to her that Odim was having difficulty cannulating the superior vena cava. In particular, she thought, he was not using an appropriately sized cannula. She said:

No, to me it looked too big. I pointed it out to him at some point in time-I can remember looking over, I was circulating nurse, I was around at the top of the bed and looking over the top and watching him cannulate.

I could see that this cannula wasn't going to go into this vessel. You know, I didn't say anything at first. Then as he struggled to get it in, I know I sort of quietly said, it's too big, it's too big. And I thought, well, maybe he would be the only one that would hear me say that. And he tried a couple of times, and I think he tore the vessel. So he stopped, and he had to suture the vessel. I think then he went to a smaller cannula and cannulated.

I think at that point in time, I can't remember if this was the first, second or third cannula, venous cannula that went in. I think it was either the first or second, and Marietess was bleeding at that point in time and starting to-her blood pressure was starting to fall and things were starting to, you know, get a little more urgent. (Evidence, pages 8,581-8,582)

Once the cannulas were inserted and the lines connected, blood was supposed to flow to the bypass machine. However, the perfusionist reported that the flow from the child was low. Youngson testified that at this point the operative field was very disorganized. However, she was able to see that one of the cannulas was clamped, preventing a flow of blood away from the body. She told the surgeon and the line was unclamped. Swartz confirmed this event in her testimony. Marietess was transfused and treated for the fall in blood pressure. These events took place at approximately 1145 hours.

When Odim was questioned about this series of events, he suggested that the problem was that the anaesthetist had not prepared for what he termed 'volume overload' in time to prevent a drop in blood pressure. He gave the following account:

I know that that dissection was difficult, because it was in the area of the chest of the previous shunt and there had been a couple of operations on that side. So dissecting out the shunt was quite difficult on the left side and the left superior vena cava was essentially quite near that area.

When we had, that was sort of the last structure that was cannulated, the aorta was cannulated, the right atrium was cannulated, and the right SVC or superior vena cava was cannulated, and that was a large structure. When we cannulated the left SVC, it was a little difficult when we opened, and I had to make a couple of passes with the cannula. During those attempts was when the blood volume was infused.

Now, typically what happens is if there is some coordination between surgeon and anaesthetist, the blood pressure is low prior to opening that vessel, you volume load beforehand so that when you do open the vessel to put the cannula in, your tank is not empty. And I think the volume loading occurred a little late here.

Q: Okay. So let's back up on that. You indicated if there is coordination between the surgeon and the anaesthetist -

A: That's right.

Q:-then you might volume load beforehand, anticipating that problem?

A: Exactly, because when you cannulate and you compress structure and open structure, you know by definition that you are going to have some blood loss as you open the vein. If your blood pressure is low beforehand, what anaesthesiologists will do is actually start giving volume, knowing that the next step is going to be opening that cava. (Evidence, pages 25,457-25,458)

Odim said that three vessels had already been cannulated and that Swartz should have provided more volume beforehand, thus making the procedure smoother. He said that this would have prevented the drop in blood pressure and the need to treat Marietess with drugs. He also indicated that he could not recall the clamped line to which Youngson referred.

Problems with distension of the heart

At 1440 hours, Marietess's heart became distended (or filled with blood due to a lack of drainage of blood from the heart). Swartz testified that the problem occurred because the catheter that was inserted into the heart to drain (or vent) it of blood had been displaced. She testified:

In this case, it appeared what happened was that the vent had come out and the heart had gradually swollen up or filled up. And if it fills up to the point where it is not being-where the arterial blood going in is not enough, is not of high enough pressure with the pressure within the heart, the heart becomes ischemic. (Evidence, page 16,092)

In this situation the heart was not short of blood-it was actually filling up with blood. However, that blood was not leaving the heart, and the pressure from this build-up was keeping the oxygen-rich blood from entering the heart muscle. Swartz testified that Odim replaced the catheter and she treated Marietess with neosynephrine. Surgery then proceeded.

Surgical bleeding and coagulopathy

At 1648 hours, the team began to wean Marietess from bypass. Her blood pressure was low and she had a metabolic acidosis, suggesting that her heart was not beating well enough to meet the metabolic needs of her body.

Marietess was bleeding from the sites where the vena cavas had been connected to the pulmonary arteries, and she also had a coagulopathy. In response, the team transfused over five times her blood volume into her. The coagulopathy was also treated with medication, and Odim addressed the surgical bleeding.

The superior vena cava syndrome

After Marietess's chest was closed, Swartz became concerned that, when she listened to the chest with a stethoscope, the left side of her chest did not sound as good as the right side. Swartz therefore requested that a chest X-ray be performed. While this was being done, Odim left the OR to speak with the parents. As the X-ray was being taken, Marietess's head became increasingly swollen and purple-coloured. Swartz thought that Marietess was developing what is known as superior vena cava (SVC) syndrome. This syndrome results from partial or complete obstruction of blood flow in the superior vena cava. Swartz and Odim provided different accounts of what transpired once Odim returned to the OR.

Swartz testified that when Odim returned, the pressures in Marietess's left external jugular vein were measured. According to Swartz, Odim was concerned that:

the adrenaline was causing vasoconstriction and that we were infusing the adrenaline into the external jugular vein. He had put in a left atrial line, and we moved the adrenaline from the left external jugular to the left atrial line, and then we measured the pressure in her external jugular. (Evidence, page 16,127)

A pressure reading of 35 millimetres of mercury (mm Hg) was measured in Marietess's left external jugular vein (a vein that connects to the left superior vena cava that is returning blood to the lungs). This measurement was about twice what Swartz believed the value should have been. Swartz concluded that the increased pressure showed a blockage in blood flow from the brain. This blockage was causing the brain to swell and was also starving the lungs of blood.

Patients who undergo Fontan procedures are usually kept in a head-up position to assist blood flow from the head to the heart. However, because she had just had an X-ray, Marietess was lying on her back when Odim returned. Odim testified:

And one of the, either technicians or nurses at the head of the table said, oh, by the way, we had been running epinephrine through that line that was transducing this pressure, but we have taken it out and we have put it in another line. And at that point I was concerned about a pharmacological, pharmacologically induced obstruction on this side related to epinephrine and its vasoconstriction or vaso tightening effects in the system. (Evidence, page 25,482)

In his testimony, Odim explained that he believed the signs of the SVC syndrome were being caused by a response to the epinephrine (or adrenalin). He also indicated that he thought other team members shared the same belief, since he believed they had switched the epinephrine to another site, once the swelling had become apparent.

Odim testified that it was not common to infuse a vasoconstrictor, such as epinephrine, into the pulmonary system, since pulmonary vessels are prone to constricting. He said that it was more appropriate to infuse vasodilators into such a system. Odim testified that he had no problem with Marietess being treated with epinephrine, but he did not want the drug infused into the left external jugular vein (which flowed to the lungs).

He testified that he suspected that the problem was caused by the epinephrine because pressures had been stable for the two hours that he took to close the chest. For all these reasons, he concluded that recent administration of epinephrine had led to a constriction and the SVC syndrome.

This differs from Swartz's account in that Odim indicated that the switch in the site of infusion of the adrenalin was made before he returned to the OR. It is a significant point because, while they were in the OR, Odim and Swartz had a serious disagreement about both the cause of the SVC syndrome and how it should be addressed.

While Odim suspected that the adrenalin had caused the pressure elevation, Swartz believed there was a physical obstruction. Odim claims that he reached his conclusion about the role that the drug played in bringing on the SVC syndrome, in part because team members told him that after the pressures increased they had switched the adrenalin site. However, according to Swartz, the adrenalin was moved from the left jugular line to the left atrial line at Odim's direction after he returned to the OR and after the SVC had been detected. There was also no indication, she said, that the other team members believed the problem was linked to the adrenalin.

At the time, Swartz suggested to Odim that either an echocardiogram or a 'line-a-gram' be performed to detect any blockage. In a line-a-gram, dye is injected into the bloodstream or into the line and an X-ray is taken. This X-ray provides an indication of where blood flows and where any blockages might lie. According to Swartz:

I had seen this before, actually. It was a similar problem, and we had done a line-a-gram, we had identified an obstruction. The patient had an obstruction because of clot, actually, and the clot was removed and the pressures came back down. And since that was a simple problem, it seemed to me, maybe this was an equally simple problem. (Evidence, page 16,136)

Swartz believed that it would be preferable to identify the problem in the OR, where, if necessary, the source of the swelling could be addressed, rather than attempting to address it in the PICU.

Odim testified that he believed the problem was caused by the adrenalin and could be best treated with drugs in the PICU. He said he did not wish to carry out the types of tests that Swartz was recommending because he felt such tests would not have indicated the cause of the blockage, merely its location. Odim also testified that the dyes used in a line-a-gram could also lead to problems.

Swartz and Odim consulted with Giddins, who also felt that the problem was a narrowing or spasm in the vein caused by medication flowing through the left external jugular line. Giddins agreed with Odim that the best course of treatment would be to take Marietess to the PICU and begin hyperventilation and vasodilator therapy. Swartz testified that she was very unhappy with the decision to leave the OR. She also testified that Giddins told her that Marietess's appearance was not uncommon.

In his testimony Giddins could not recall a serious disagreement between Swartz and Odim and himself over whether or not Marietess should be taken out of the OR. Giddins was asked if consideration was given to stabilizing her in the operating room.

No. In the opinion of all, at the end of the operation, there were set strategies that were felt to be best carried out in the intensive care unit and, in fact, would not be as ideally done in the operating room. (Evidence, pages 4,019-4,020)

Swartz, however, insisted that she made it very clear that, in her opinion, it was not the best strategy to move Marietess to the PICU.