The impact of operative bleeding on outcome in transplantation of the liver

Abstract

Excessive operative blood transfusion has been correlated with an increased rate of infectious complications and lower survival rate after transplantation of the liver. Two hundred and five consecutive transplants of the liver, performed between January 1988 and December 1989, were studied retrospectively to determine preoperative risk factors associated with an increased operative blood loss and to evaluate the impact of operative transfusion on the outcome of transplantation. Preoperative clinical and laboratory parameters in patients who required 10 units or more of banked erythrocytes were compared with those in patients who received less than ten units of erythrocytes. In evaluating the outcome, the two groups were compared for infection, rejection and graft and patient survival rates. The median operative blood loss for 205 patients was 5 units of banked erythrocytes (range of zero to 52, mean of 6.9 units). Only 41 patients (20 percent) required 10 units or more of erythrocytes. The significant factors on univariate analysis that were associated with an increased operative blood loss were hospitalized patients (United Network for Organ Sharing Status > or = 3), fulminant hepatic failure, previous portosystemic shunt and complete ABO mismatch. Patients who required more blood had higher incidence of coagulation abnormalities, renal dysfunction and high bilirubin levels. A stepwise logistic regression analysis model using all these parameters identified an elevated serum creatinine, decreased platelets and a prolonged partial thromboplastin time as being the strongest risk factors. Using these variables, operative bleeding of more than 10 units could be predicted accurately only 60 percent of the time (sensitivity 60.0 percent with a specificity of 69.1 percent). Septic episodes occurred more frequently in patients with an excessive operative blood loss (p < 0.05), and these patients also tended to have a higher rate of severe cytomegalovirus infections and a lower incidence of acute rejection. Patients who required more blood also had significantly prolonged stays in the intensive care units postoperatively (18.3 versus 6.3 days, p < 0.002) and lower graft and patient survival rates (p < 0.001 and p < 0.05, respectively). We conclude that intraoperative bleeding has remained a significant problem affecting the immediate outcome after transplantation of the liver. Preoperative parameters cannot predict operative bleeding accurately and the mainstay to prevent bleeding is a meticulous surgical technique during the hepatectomy and correction of coagulation abnormalities throughout the procedure.