The treatment strategy for a patient with known liver disease depends on the etiology of the liver disease, the chronicity and histologic severity of the underlying liver disease, and the nature of the surgery. Intuitively, the more emergent the surgery, the less opportunity available to correct reversible factors such as electrolyte abnormalities, coagulopathy, and effects of portal hypertension (eg, ascites, hepatic encephalopathy). The presence of decompensated liver disease increases the risk of postoperative complications. Acute liver disease, such as viral hepatitis or alcoholic hepatitis, is associated with more ongoing hepatocyte damage than chronic liver disease, such as steatohepatitis. In general, patients with hepatitis likewise have higher morbidity rates than those with cholestatic disease. Patients with chronic liver disease, including conditions such as chronic hepatitis C, in whom liver function is preserved, may not have an increased operative risk (Friedman, 1999).

In patients with alcoholic hepatitis, elective surgery is associated with high mortality unless the surgery is deferred until prolonged abstinence and biochemical and clinical progress is noted. Patients with cirrhosis have higher mortality rates postoperatively because of sepsis, bleeding, renal failure, and hepatic failure (decompensation of underlying liver disease). Patients with Child’s class B and Child’s class C cirrhosis have poorer outcomes when compared to those with Child's class A

The Child-Pugh score correlates with perioperative mortality in patients undergoing nonshunt surgery and in patients with cirrhosis who are undergoing abdominal procedures. According to a more recent analysis, serum albumin, leukocytosis, and an elevated prothrombin time are the most sensitive indicators of perioperative mortality independent of the Child-Pugh score (Garrison, 1984).

The scoring system for determining liver transplant necessity is based on the Child-Turcott-Pugh criteria. However, according to the United Network for Organ Sharing (UNOS) committee, this system is not designed to assess the probability of survival for patients with chronic liver disease in a nonsurgical situation. Rather, it is for patients undergoing surgical portocaval shunt or esophageal transection. Also, the system does not take into account the severity of any one clinical factor. For instance, a patient with a bilirubin level of 4 mg/dL and one with a bilirubin level of 20 mg/dL both get a score of 3, although the patient with the bilirubin level of 20 mg/dL has the greater chance of dying. The amount of ascites and degree of hepatic encephalopathy are altered easily by medical intervention, and this is not taken into account by the scoring system.

The Mayo End-Stage Liver Disease (MELD) model, which uses serum creatinine, bilirubin, International Normalized Ratio, and disease etiology, may accurately predict patients who have a median survival of 3 months after an elective transjugular intrahepatic portal-systemic shunt (TIPS) procedure. This model is based on a heterogeneous group of patients from geographically diverse areas of the US. Elective abdominal surgical procedures should be avoided in potential candidates for orthotopic liver transplantation.

Surgery in patients with aminopyrine breath test (ABT) results of less than 2.3% is associated with extremely high mortality. However, patients with cirrhosis who have normal ABTs tolerate elective surgery well (Gill, 1983). According to an analysis in one study, the presence of ascites in patients with cirrhosis is associated with a 37-83% mortality rate compared to an 11-53% mortality rate in those without ascites (Conn, 1991). Likewise, emergency surgery is associated with a 45-86% mortality rate, and preoperative infection is associated with a 64% mortality rate.

A bilirubin level of higher than 3 mg/dL, elevated creatinine, and hypoalbuminemia also are known to be associated with increased mortality (Runyon, 1986). The odds ratio for perioperative mortality in patients with liver disease who undergo cholecystectomy is 8.47. In fact, open cholecystectomy in patients with cirrhosis has been called a formidable operation, although more recent studies have reported lower, but still considerable, mortality in patients with cirrhosis who undergo abdominal surgery. However, laparoscopic cholecystectomy can be performed safely in selected patients who have well-compensated cirrhosis and no signs of portal hypertension (Friedman, 1999).

Cardiac surgery in patients with cirrhosis has been associated with a high operative mortality rate (Friedman, 1999). In patients with obstructive jaundice, other factors, such as a hematocrit of less than 30%, a serum bilirubin level of higher than 11 mg/dL, malignant biliary obstruction, azotemia, and cholangitis, are risk factors for operative mortality (Friedman, 1999).

The need for surgery for many conditions has decreased because of advances in imaging and because of improved nonsurgical techniques for managing problems such as extrahepatic biliary obstruction. TIPS has become the treatment of choice for managing refractory variceal bleeding in most instances, and surgical shunts are inserted only in special circumstances. Liver transplantation and hepatic surgery for a variety of causes have their own risks and often are necessary procedures. For example, in general, the anatomy of the liver precludes lobectomy in a bloodless fashion. Also, primary or secondary hepatic tumors may not produce changes in liver function, but postoperative factors can.

PREOPERATIVE FACTORS

Clinical