It would be interesting to know a bit more onhow these two cases were managed and whether this has led to an alteration in the case selection for this procedure in your institution
Oh! Generally speaking there are strategic and technical issues to discuss in those cases. The former are related to indications and surgical planning. If one operate lesion abutting hepatocaval confluence it should be always prepared for caval clamping, and such patient is not good candidate for minimally invasive surgery. Technical issues are related to technique of transection of the liver and the approach to hepatic veins. All those are clearly surgical problems. I would like to tell you just that I am impressed that you save the patients. Finally to answer briefly to your question: I will immediately compress with gauze, put the liver back if it is lifted by retractor, turn patient in Trendelenburg position and open the abdomen.
First, we are used to manage acute and massive bleeding, and this is important.
Then, aggressive fluid resuscitation and blood transfusion were started using 2 rapid infusion systems Large bore iv catheters were inserted. Vasopressor support using noradrenaline was also initiated. Emergency conversion to laparotomy and undocking of the robot (27 seconds) were done, and table was in Trendelenburg position. Surgical haemostasis was very difficult to achieve requiring total clampation of inferior vena cava. Hemodynamic instability with severe hypovolaemia led to cardiac arrest. I asked the surgeons to perform thoracotomy and to start Internal cardiac massage with return of spontaneous circulation after 4 minutes. Total blood loss was 16.500 ml. Duration of surgery was 335 minutes. The patient had a favourable postoperative outcome with complete neurological recovery, allowing extubation on postoperative day 6, with a PACU stay of 8 days and a total LOS of 24 days.
Second case was almost similar, but with less blood loss and no cardiac arrest, but required huge doses of vasopressor. The operation was a cephalic duodenopancreatectomy.
In both cases we installed after intubation a CVC and a radial artery catether, and we usually have a large bore peripheral cannula in stand-by.
Technically speaking, having a competent assistant that can compress with a gauze or with the liver parenchyma the minutes it takes to undock the robot, for the surgeon to scrub up, and convert to laparotomy makes a big difference. The compression can be continued through a port until an open abdominal access and exposure is achieved.
Also turning up a bit the laparoscopic insuflation pressure to 14-15mmHg (that is, higher than the CVP) can also provide short term hemostasis for vena cava or hepatic veins bleeding.
I am very impressed for the good outcomes of that patients due to an immediate laparotomy and bleeding control. However, I am sure that the purpose of the question was how to prevent such a potential life-threatening accident. The first idea involves a careful preoperative imaging study: for example, it could have shown some vascular abnormalities of hepatic veins confluence. Moreover, I am strongly convinced that the crucial point is the use of an intraoperative ultrasonography probe managed by one of the robotic arm. This technology is also able to show the liver (and other major vascular structures) vascularization through a "picture in picture" image in the surgeon's console. This real time imaging device is able to best guide any liver resection or vessel dissections.