The number of patients with cirrhosis who require surgery is on the rise. Despite advances in antiviral therapeutics, the prevalence of cirrhosis secondary to hepatitis C continues to increase, as does the prevalence of cirrhosis due to chronic alcoholic liver disease. Additionally, nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH) are gaining more attention, especially in association with metabolic syndrome and obesity. At the same time, the amount of medications and treatments aimed at improving survival among patients with cirrhosis has been increasing. Therefore, it can be expected that a growing number of patients with liver disease, both known and as yet undiagnosed and asymptomatic, will undergo surgery.
An estimated 1 in 700 patients admitted for elective surgery has abnormal liver enzyme levels. Some authors have estimated that as many as 10% of patients with advanced liver disease will undergo surgery in the last 2 years of their lives.  This article focuses on the challenges of perioperative care of patients with liver disease.
Identification of the surgical risk is imperative in the care of any patient, especially as patients develop an increasing number of chronic comorbid medical conditions. Patients with liver disease are at particularly high risk for morbidity and mortality in the postoperative period due to both the stress of surgery and the effects of general anesthesia. del Olmo et al compared 135 patients with cirrhosis with 86 patients without cirrhosis, all undergoing nonhepatic general surgery.  At 1 month, mortality rates were 16.3% for patients with cirrhosis compared with 3.5% in the control group. What is further evident in the literature is that decompensated liver disease increases the risk of postoperative complications (eg, acute hepatic failure, infections including sepsis, bleeding, poor wound healing, and renal dysfunction). Assessing risk in these patients is a challenging but important endeavor.
The liver is vital for protein synthesis, coagulation homeostasis, glucose homeostasis, bilirubin excretion, drug metabolism, and toxic removal, among other critical functions. In general, the liver has substantial functional reserve because of its dual blood supply: portal-venous (75%) and hepatic-arterial (25%). Hence, clinical manifestations of liver damage occur only after considerable injury.
Liver disease comprises a large spectrum of hepatic dysfunction. It includes asymptomatic transaminitis, cirrhosis, and end-stage liver disease. The most common causes of advanced liver disease are chronic viral infections (hepatitis C [HCV] and B [HBV]), alcohol abuse, NAFLD/NASH , autoimmune disease, drugs or toxins, metabolic disorders (eg, alpha-1 antitrypsin deficiency, hemochromatosis, and Wilson disease), and biliary tract diseases.
Surgical Risk Assessment
Basis for risk assessment
Secondary to the loss of hepatic reserve capacity and because of other systemic derangements that are the result of liver dysfunction (such as hemodynamic impairments), patients with liver disease have an inappropriate response to surgical stress. These individuals are accordingly at an increased risk of bleeding, infection, impaired wound healing, postoperative hepatic decompensation, including hepatic coma or death. Therefore, the decision to perform surgery in these patients must be heavily weighed.
Prediction of surgical risk is based on the degree of liver dysfunction, the type of surgery, and the preclinical status of the patient. The extent of liver dysfunction and type of surgery play key roles in determining a patient’s specific risk. In addition, liver disease can affect almost every organ and system in the body, including the cardiorespiratory and circulatory systems, the brain, the kidneys, and the immune system.
The extent to which secondary manifestations of liver disease affect these systems may be just as important as the manifestations of primary liver dysfunction in predicting the outcome after surgery. Such comorbid conditions responsible for perioperative morbidity and mortality (eg, coagulopathy, intravascular volume, renal function, electrolytes, cardiovascular status, and nutritional status) should be identified and addressed before surgery. Optimal preparation may decrease death and complications after surgery. Issues to anticipate and address include manifestations of acute liver decompensation including encephalopathy, acute renal failure, coagulopathy, adult respiratory distress syndrome, and sepsis. [3, 4, 5]
Quantitative risk stratification
Two risk stratifications schemes have been used to estimate the perioperative risk of patients with cirrhosis: the Child-Turcotte-Pugh score and the Model for End-Stage Liver Disease (MELD) score.
The Child-Turcotte-Pugh (CTP) score incorporates a combination of 3 biochemical elements (ie, prothrombin time [PT], albumin level, and bilirubin level) and 2 clinical features (ie, presence of ascites and encephalopathy) to assess the primary functions of the liver (see the table below). A patient’s score is translated to 1 of 3 CTP classes: A, B, or C, with A reflecting the least severe disease. Patients who are CTP class B and C have worse outcomes and are candidates for liver transplantation.
The CTP score was first developed to predict mortality after portocaval shunt surgery, but it has since been used to predict perioperative morbidity and mortality rates for patients undergoing hepatic and nonhepatic intra-abdominal surgeries. [1, 6, 7, 8] Patients with CTP class A disease are estimated to have a 10% mortality rate after abdominal surgery. That mortality rate increases to 30-31% for CTP class B and 76-82% for CTP class C. [1, 8] A 2010 study by Telem et al demonstrated lower mortality rates (CTP 2%, CTP B 12%, CTP C 12%) when surgeries were performed at an institution specializing in hepatology and liver transplantation. 
However, the CTP scoring system has been challenged for its ambiguity and interobserver variability because it includes subjective parameters (eg, degree of ascites and encephalopathy). Additionally, all the factors are weighted equally. Patients within a given class are not homogenous but also not distinguished between, a feature for which it has also been criticized. 
The MELD score was originally developed to predict short-term mortality for patients undergoing transjugular intrahepatic portosystemic shunt (TIPS) placement. It has since been adopted as the tool to prioritize patients with cirrhosis for liver transplantation.
The MELD score is based on a patient's serum bilirubin, creatinine, and international normalized ratio (INR) for prothrombin time and is calculated from a validated predictive equation, as follows: (3.8 × ln bilirubin value) + (11.2 X ln INR) + (9.6 ln creatinine value), where bilirubin and creatinine values are in milligrams per deciliter (mg/dL) and ln represents natural logarithm. See the MELD Score calculator.
The MELD score originally included the etiology of liver failure, but this criterion was subsequently dropped from the equation because it was proved prognostically insignificant. The Organ Procurement and Transplantation Network (OPTN) also provides an online calculator.
With regard to its original utilization, a MELD score < 8 predicts good outcome after TIPS and a score >18 predicts poor outcome, with best outcomes seen in patients with scores < 14. Avoidance of TIPS is generally recommended in patients with a MELD score >24, unless the procedure is used as a measure of last resort to control active variceal bleeding. Since its implementation, the MELD score’s use has been expanded to also predict the risk of mortality and morbidity after other procedures. A MELD score of at least 8 predicts an increased risk of postoperative complications, including death in patients undergoing cholecystectomy  and cardiac surgery requiring cardiopulmonary bypass. 
Several authors have also shown that the MELD score predicts morbidity and mortality after hepatic resection for hepatocellular carcinoma. Cucchetti et al showed that MELD scores < 9 were associated with 0% postoperative liver failure; MELD scores 9-10 were associated with 3.6% postoperative liver failure; and MELD scores >10 were associated with 37.5% postoperative liver failure.  Teh et al showed that a MELD score less than or equal to 8 was associated with 0% postresection mortality compared with 29% mortality for MELD scores >8. 
In general, the MELD score fairs well compared to the CTP score. However, some might argue that the MELD score may be a more objective predictor of postoperative mortality than the CTP score, [11, 12, 15, 16] especially as patients fall along a continuum of values instead of into 3 discrete groups.
The MELD score has been validated as an independent prediction tool to calculate postoperative mortality. A retrospective analysis by Northup et al found that the MELD score was the only statistically significant predictor of 30-day mortality. For example, with a MELD of 5 was associated with 5% risk; 10, with a 7% risk; 15, with an 11% risk; 20, with a 17% risk; or 25, with a 26% risk. 
Teh et al performed a retrospective, multivariate analysis that showed among patients with cirrhosis undergoing multiple types of major surgeries, the MELD but not the CTP score predicted increased mortality at 30 and 90 days, 1 year, and over the long term.  Age and American Society of Anesthesiologists (ASA) class also predicted postoperative mortality. The MELD was the strongest predictor of mortality after 7 days and over the long term. For example, the 30-day mortality associated with MELD < 8 was 5.7% but >50% for MELD score >20.  The relative risk of mortality also increased 14% for each 1 point increase in the MELD score.
The same study by Telem et al demonstrated an improved though still increased mortality rate for MELD ≥15 (29%); if an albumin ≤2.5 mg/dL was added to a MELD score ≥15, mortality increased to 60%. 
The 2007 review by Hanje et al posited the general recommendation that patients with MELD < 10 can undergo elective surgery and patients with a MELD of 10-15 should proceed with caution. Notation was made that patient- and surgery-specific factors also contribute to the risk/benefit decisions. 
More recently, the MELD score has been adapted with additional clinical risk factors, creating the "integrated MELD" score (iMELD): iMELD = MELD + (0.3 X age) - (0.7 X serum sodium [mEq/L]) + 100. In a retrospective study of 190 patients with cirrhosis, Costa et al demonstrated that the iMELD score had better prognostic strength compared with the MELD or CTP scores. In brief, iMELD scores of < 35, 35-45, and >45 were associated with perioperative mortality rates of 4%, 16%, and 50%, respectively. 
Most recently, a 2011 study again showed that perioperative mortality increased with the CTP and MELD scores (10% Child A, 17% Child B, 63% Child C; P < .01; 9% MELD < 10, 19% MELD 10–15, 54% MELD >15; P < .001). However, on multivariate analysis, the CTP and ASA classes, need for intraoperative transfusions, and preoperative sodium < 130 mmol/L were independent prognostic factors; MELD score was not prognostic. Further subgroup analysis of elective surgeries revealed only preoperative creatinine ≥1.1 mg/dL to be predictive of mortality. For emergent surgeries, CTP but not MELD score was again predictive of outcomes. 
Table. Child-Turcotte-Pugh classification of liver disease. (CTP A = 5-6 points, CTP B = 7-9 points, CTP C = 10-15 points) (Open Table in a new window)
|Criterion||1 point each||2 points each||3 points each|
|Ascites||None||Controlled with diuretics||Poorly controlled|
|Encephalopathy||None||Grade I-II||Grade III-IV|
Total bilirubin, µmol/L
(normal = 17.1 µmol/L or 1.0 mg/dL)
34 – 50
(> 3 mg/dL)
|Albumin, g/L||>35 (>3.5 g/dL)||25-35 (2.5-3.5 g/dL)||< 25 (< 2.5 g/dL)|
Other risk stratification systems
The ASA physical status class risk stratification system is based on comorbid conditions that are a threat to life or that limit activity and thus helps in predicting preoperative risks. In general, an ASA class greater than 2 increases the risk 1.5- to 3.2-fold.  The ASA class independently predicted postoperative mortality in patients undergoing hepatic resection for hepatocellular carcinoma.  Teh et al also found the ASA class significantly predicts increased mortality and morbidity among patients with cirrhosis undergoing major surgery, with ASA class V the strongest predictor of postoperative mortality at 7 days.  The mortality related to ASA IV was the equivalent of 5.5 MELD points in terms of risk.
It is also important to not overlook the preoperative cardiopulmonary evaluation. This is required of any patient, regardless of the functional status of their liver. Cardiac risk stratification should include an assessment of functional capacity (metabolic equivalent [MET] or exercise duration). Additional noninvasive testing such as stress testing might be considered if it will change perioperative management. Cardiac surgery performed in patients with cirrhosis is associated with a high surgical mortality rate and will be discussed separately. 
In 1997, the American College of Physicians (ACP) published guidelines in the form of algorithms for assessing and managing perioperative risks based on the results of the tests mentioned above. The Goldman cardiac risk index is used to predict postoperative pulmonary and cardiac complications.  It is a classification system based on points assigned to a patient's clinical history, physical findings, ECGs, general medical status (based on arterial blood gases [ABGs], electrolytes, and liver disease), and type of operation. The guidelines were updated by the American College of Cardiology/American Heart Association in 2007 and now include the Revised Cardiac Risk Index. [24, 25]
Preoperative Assessment and Management
The evaluation of any patient undergoing surgery should include thorough history taking and physical examination. In asymptomatic patients, this is an extremely valuable screening tool. Risk factors (eg, pervious blood transfusions, tattoos, illicit drug use, sexual history, alcohol use, and personal or family history of jaundice) for liver disease should be explored.
A complete medication review including other-the-counter (OTC) and herbal agents should be performed. Symptoms or physical signs suggestive of liver dysfunction/disease (eg, hepatosplenomegaly, spider angiomata, jaundice, gynecomastia, palmar erythema, scleral icterus, asterixis, encephalopathy) should prompt further examination with liver function tests, coagulation studies, complete blood cell (CBC) counts and metabolic panels. However, routine preoperative testing of liver function is not recommended because of the low prevalence of liver abnormalities in clinically asymptomatic patients. [26, 27, 28]
Asymptomatic patients with significantly abnormal liver function should have their elective surgery postponed and their liver disease investigated; their perioperative risk should be reassessed after their liver dysfunction is characterized. 
Acuity of liver disease
Though most studies have focused on patients with end-stage liver disease or cirrhosis, patients with fulminant hepatic failure have been associated with an increased risk of surgical morbidity and mortality . This also applies to patients with acute alcoholic hepatitis and acute viral hepatitis. Patients with these conditions tend to have morbidity rates higher than those with chronic cholestatic disease. Therefore, it is prudent to postpone surgery, especially elective surgery, until transaminitis is resolved.  Patients with chronic liver disease but with preserved hepatic function may not have an increased operative risk,  but these individuals need to be closely evaluated nonetheless.
Severity and specific derangements of known chronic liver disease
In patients with known liver disease, especially with cirrhosis, optimal preparation for surgery, that appropriately addresses the primary features and secondary manifestations of liver disease may decrease the risk of complications or death after surgery. This includes laboratory tests to assess blood counts, coagulopathy, electrolyte abnormalities, and markers of hepatic synthetic function.
Coagulopathy and thrombocytopenia
Coagulopathy is one of the primary features of advanced liver disease. In addition to hepatic synthetic dysfunction (all of the coagulation factors with the exception of von Willebrand factor are produced in the liver), malnutrition and vitamin K malabsorption due to cholestasis contribute to this abnormality. Additionally, portal hypertension leads to hypersplenism with resultant platelet trapping and peripheral thrombocytopenia. Vitamin K supplementation and administration of fresh-frozen plasma (FFP) are recommended to correct coagulopathy before surgery. Cryoprecipitate might also be required to reduce the prothrombin time. A prolongedbleeding time can also be corrected with diamino-8-D-arginine vasopressin (DDAVP). Finally, platelet transfusion (see platelets) may be necessary based on the patient’s platelet level and the desired level as dictated by the type of surgery.
A recent study explored the effect of preoperative eltrombopag, an oral thrombopoietin-receptor agonist, on the need for platelet transfusions for patients with chronic liver disease and thrombocytopenia. The authors saw a statistically significant difference in the need for platelet transfusions (28% in the treatment group vs 81%) without a significant difference in bleeding episodes. Portal vein thrombosis occurred much more frequently in the treatment group, leading to early termination of the study. 
Ascites is important to assess and manage before surgery because it can lead to wound dehiscence, abdominal wall herniation, and respiratory compromise secondary to reduced lung expansion. In a study by Conn, ascites in patients with cirrhosis was associated with a 37-83% mortality rate compared with 11-53% in those without ascites.  In general, ascites should be treated aggressively with diuretics and/or large-volume paracentesis before surgery. A low sodium diet is another important component of ascites management. Patients on diuretics need to have their creatinine and electrolytes monitored.
Ascites fluid can also be removed intraoperatively at laparotomy.  It is important to take note of the volume of fluid removed and the patient’s baseline renal function and to consider albumin repletement to maintain intravascular volume and prevent paracentesis-induced circulatory dysfunction. Ascitic fluid should also be analyzed to rule out spontaneous bacterial peritonitis.
Many patients with cirrhosis may have portosystemic encephalopathy at baseline, which increases their risk of postoperative encephalopathy. A retrospective study of 40 patients with chronic liver failure undergoing nonhepatic surgery demonstrated that encephalopathy was associated with an 88% risk of mortality, which was even higher than the 50% risk associated with emergency surgery. 
Multiple factors in the preoperative and postoperative periods may precipitate encephalopathy, such as infection and/or sepsis, diuretics, hypokalemia, metabolic alkalosis, constipation, use of central nervous system (CNS) depressants such narcotics and benzodiazepines, hypoxia, azotemia, and gastrointestinal bleeding. Addressing the underlying precipitant through correction of electrolyte abnormalities, treatment of infection, management of gastrointestinal bleeding, and restriction of sedatives may help prevent or decrease encephalopathy. Hepatic encephalopathy is also often treated by administering lactulose or poorly absorbed antibiotics such as rifaximin.
Patients with chronic liver disease are at risk for renal dysfunction at baseline due to the propensity for hemodynamic derangements that increase the risk of renal hypoperfusion. This risk is increased by diuretics, nephrotoxic agents including nonsteroidal anti-inflammatory drugs (NSAIDs), large-volume paracentesis performed without albumin supplementation, infections, and gastrointestinal bleeding. Hepatorenal syndrome is another concerning occurrence in this patient population.
The risk of renal dysfunction in the postoperative period is increased because of hemodynamic changes and fluid shifts or losses, particularly if ascites fluid is removed at laparotomy. Renal function should be closely monitored pre- and postoperatively, with appropriate measures taken to address or eliminate potential insults. Attention should also be given to the fact that serum creatinine levels often overestimate the glomerular filtration rate (GFR) in patients with cirrhosis; a seemingly normal creatinine level may indeed represent impaired renal function. Vasoactive compounds such as midodrine and terlipressin appear to be at least as effective as intravenous albumin in preventing circulatory dysfunction with resultant renal impairment in patients with cirrhosis who have lost third-spaced volume. [35, 36]
Pulmonary complications of end-stage liver disease include hepatopulmonary syndrome, portopulmonary hypertension, and hepatic hydrothorax. Hepatopulmonary syndrome is associated with vascular shunt, and the risk of hypoxia and ventilation-perfusion mismatch should be addressed before surgery. Portopulmonary hypertension can eventually lead to right heart failure and hypoxia. Hepatic hydrothorax, usually unilateral and in the right hemithorax, can occur and impair ventilation. However, the associated hypoxemia is usually not severe.  Drainage is usually not recommended because the effusion often rapidly reaccumulates. Finally, the risk of chronic obstructive pulmonary disease (COPD) should be assessed in any patient who has previously smoked tobacco or who has alpha-1 antitrypsin deficiency.
Severe malnutrition is associated with an increased need for packed red blood cells, FFP, and cryoprecipitate during liver transplantation. It is also associated with a prolonged postoperative stay. Stephenson et al suggest that preoperative improvement in the patient's nutritional status may improve outcomes. 
In patients with end-stage liver disease, steps to improve nutritional health should be started, preferably in the preoperative period, because they are expected to have increased energy expenditure after surgery.  Supplements can be used. Patients with alcoholic liver disease and Wernicke encephalopathy benefit from preoperative vitamin B-1 supplementation. Advanced liver disease can also predispose to hypoglycemia because of impaired gluconeogenesis and decreased glycogen stores.
Patients with autoimmune hepatitis on daily steroids may be appropriate candidates for stress-dosed steroids with surgery, depending on their average daily steroid dose. D-penicillamine can impair wound healing; patients taking it for Wilson disease should decrease their dose for 1-2 weeks preoperatively and postoperatively. Wilson disease might predispose to an increased risk of neurologic changes postoperatively. Patients with hemochromatosis should be assessed for a history of extrahepatic hemochromatosis-related complications, such as diabetes or cardiomyopathy.
In addition, it is worth noting that patients with a history of alcohol abuse are at increased risk of other complications, including poor wound healing, bleeding, delirium, and infections. Patients who have continued to actively drink are at risk for withdrawal. Though studies indicate that preoperative alcohol cessation improves outcomes, patients with cirrhosis were excluded from the intervention. [40, 41] However, it seems reasonable to recommend a period of abstinence for patients with known chronic liver disease/cirrhosis or to delay elective surgery until such can be achieved.
Impaired hepatic synthetic function and derangement of other hepatic functions are especially pertinent to note when choosing anesthetic and other agents used in the perioperative period. These changes include decreased synthesis of plasma-binding proteins. Hypoalbuminemia impairs drug binding and metabolism and elevates serum drug levels. Impaired drug metabolism, detoxification, and excretion by the liver can prolong drug half-lives. Thus, the absorption, distribution, metabolism, and excretion of anesthetics, muscle relaxants, analgesics, and sedatives may be affected.
Patients with liver disease are more likely than patients without liver disease to have hepatic decompensation with the use of anesthesia.  General anesthesia reduces total hepatic blood flow, especially the contribution of the hepatic artery. Patients with liver disease tend to have several baseline cardiovascular abnormalities, including decreased systemic vascular resistance and increased cardiac index, which may further affect hepatic blood flow. In addition, catecholamine and other neurohormonal responses are impaired in patients with liver disease; therefore, intraoperative hypovolemia or hemorrhage may not trigger adequate compensatory mechanisms. Anesthetics causing sympathetic blockade further blunt this response. The result of this reduction in hepatic perfusion is a drastic loss of their remaining marginal hepatic function.
Of all the inhaled anesthetics, halothane and enflurane appear to reduce hepatic artery blood flow the most because of systemic vasodilation and a mild negative inotropic effect. [31, 42, 43, 44] Halothane is also associated with the greatest risk of hepatotoxicity, with the incidence of fulminant hepatitis approximating 1 case in 6,000-35,000 patients after exposure.  Isoflurane has fewer effects on hepatic blood flow and less hepatic metabolism ; it is the preferred anesthetic agent in patients with liver disease. Newer haloalkanes, such as sevoflurane and desflurane, also undergo less hepatic metabolism than halothane or enflurane.
The drug effects of neuromuscular blocking agents may be prolonged in patients with liver disease because of impaired biliary excretion. Atracurium has been recommended as the agent of choice because it relies on neither the liver nor kidney for excretion.  Likewise, drugs such as morphine, meperidine, benzodiazepines, and barbiturates should be used with caution because of their dependence on the liver for metabolism. In general, the doses of these agents should be decreased by 50%.  Fentanyl is the preferred narcotic. 
The type of surgery is potentially an important determinant of postoperative hepatic dysfunction. Because of traction on abdominal viscera, intra-abdominal operations are more likely than extra-abdominal surgeries to cause reflex systemic hypotension and to subsequently reduce hepatic blood flow. Hypercarbia-induced splanchnic vasoconstriction is also a threat to hepatic perfusion. Surgeries that result in a large amount of blood loss increase the risk for ischemic hepatic injury, as can intraoperative hypotension. Sufficient surgical hemostasis and autologous platelet-rich plasma have been demonstrated to be useful for prevention of massive hemorrhage. [3, 8]
Examples of specific surgeries and considerations
Cholecystitis and cholelithiasis are common in patients with liver disease. 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 recent studies have demonstrated lowered but still considerable mortality rates in patients with cirrhosis who undergo abdominal surgery. Perkins et al confirmed that a MELD score greater to or equal to 8 predicts an increased risk of postoperative complications in this type of surgery  (see the MELD Score calculator).
However, laparoscopic cholecystectomy can be safely performed in selected patients who have well-compensated cirrhosis and no signs of portal hypertension.  A case-controlled retrospective review of laparoscopic cholecystectomy in 48 patients with Child-Pugh class A (80% of patients) and Child-Pugh class B cirrhosis demonstrated no increase in morbidity and mortality rates or worsening of outcome compared with control subjects.  Another small series had similar results  ;the authors concluded that laparoscopic cholecystectomy is relatively safe in patients with Child-Pugh class A or B cirrhosis. In addition, Ji et al showed that laparoscopic cholecystectomy was associated with lower rates of postoperative complications than open cholecystectomy in patients with cirrhosis matched for disease severity. 
A large study of 747 patients from 1990 to 1997 who underwent liver resection demonstrated that mortality was significantly higher in patients with cirrhosis (8.7%) or obstructive jaundice (21%) than in patients with a normal liver (1%; P < 0.001).  Other groups have also demonstrated that the MELD score predicts risk of postresection morbidity and mortality. [14, 13, 53, 54]
Cardiac surgery in patients with cirrhosis is associated with a high operative mortality rate.  A lone study found the following risk factors for operative mortality: obstructive jaundice, hematocrit value < 30%, serum bilirubin level >11 mg/dL, malignant biliary obstruction, azotemia, and cholangitis.  In a small study, patients with cirrhosis and a CTP class A were found to have 0% mortality; B, 50% mortality; and C, 100% mortality after cardiac surgery,  with another group finding that a CTP score >7 was more sensitive and as specific as the MELD score in predicting poor outcome.  Another small study of 27 patients demonstrated that cardiac surgery could be safely performed in patients with CTP class A and selected patients with CTP class B. Also shown was that the use of cardiopulmonary bypass increased with of mortality.  A 2012 study demonstrated consistent results, showing that a CTP < 8 was associated with mortality comparable to matched controls,aswell as a lower risk of renal failure and dialysis than patients with CTP ≥8. 
Like other major open abdominal procedures, open abdominal aortic aneurysm (AAA) repair may be associated with an increased risk of postoperative mortality, even with mild chronic liver disease.  However, a recent study by Marrocoo-Trischitta et al noted that the procedure can be safely performed in patients with compensated cirrhosis, though they noted a nearly double length-of-stay in patients with cirrhosis, despite predominantly mild disease with CTP score A and average MELD 8. They also cautioned that patients with MELD ≥10 may not benefit from the vascular procedure given their overall limited life expectancy.  Bush et al have also demonstrated endovascular AAA repair to be safe in patients with multiple comorbid medical conditions, including hepatic disease. 
Various types and indications for orthopedic procedures also affect perioperative risk for patients and cirrhosis. Cohen et al performed a retrospective review of outcomes of primary total hip arthroplasties and total knee arthroplasties for cirrhotic patients versus matched controls. Significantly worse outcomes were seen in patients with cirrhosis (20.7% vs 3.23%). Higher complication rates were seen in cirrhotic patients undergoing emergent total hip arthroplasties for hip fracture repair (80% had a major complication, with 60% mortality rate). More advanced liver disease trended towards worse clinical outcomes, but primary total hip arthroplasty or total knee arthroplasty could be safely performed in patients with CTP class A or B. 
In some parts of the world, parasitic diseases, such as hydatid disease or echinococcosis, may cause liver lesions that need to be surgically removed. In such cases, the surgical technique is important, and sepsis can cause perioperative morbidity. 
Patients undergoing emergency surgery are at substantial risk for liver dysfunction. Intuition suggests, the more urgent the surgery, the less opportunity that is available to correct reversible factors, such as electrolyte abnormalities, coagulopathy, and clinical manifestations of portal hypertension (eg, ascites, hepatic encephalopathy).
Emergency surgery is an important predictor of adverse outcome. In a series of 100 patients with cirrhosis who underwent abdominal surgery for a variety of reasons, 80% of nonsurvivors and 40% of survivors who had serious complications had undergone emergency surgery.  A series of 92 patients with cirrhosis who underwent abdominal surgeries had a 50% mortality rate in association with emergency procedures (22% for CTP class A, 38% for CTP class B, 100% for CTP class C) versus 18% for elective surgery (P = 0.001).  This study showed that the most accurate predictor of outcome is the patient's preoperative CTP class.
A 2004 study demonstrated that patients with cirrhosis had a higher perioperative morbidity and mortality rate with emergency surgery than with elective surgery. Mortality rates significantly differed between the groups (emergency group, 1 mo = 19% mortality rate, 3 mo = 44%; elective group, 1 mo = 17% mortality rate, 3 mo = 21%; P < 0.05).  A 2007 study found that 100% of patients with cirrhosis undergoing emergency surgery died, with a median survival 2 days  ; all these patients had higher MELD scores and were ASA class V.  The 2011 study by Neeff et al again demonstrated significantly increased mortality with emergent surgeries (CTP A 0%, CTP B 30%, CTP C 72%, P = .0001; MELD < 10 20%, MELD 10-15 33%, MELD > 15 60%, P = .03). 
More specifically, cirrhosis diagnosis made at the time of laparotomy for trauma is associated with an increased risk of mortality (45% vs 24%).  Another study confirmed the elevated mortality when cirrhosis and trauma combined: 12% versus 6% overall mortality, 40% versus 15% mortality after emergent abdominal exploration. [64, 65] Postoperative ICU admission and care is recommended for such patients, even for mild injuries.
Alternatives to surgery
Relatively noninvasive techniques or advances in medical management have replaced surgical intervention for many conditions (eg, extrahepatic biliary obstruction, refractory variceal hemorrhage, coronary artery disease). TIPS has become the treatment of choice for managing cases of refractory variceal bleeding, and surgical shunts are created only in special circumstances.
Percutaneous stenting or endoscopic retrograde cholangiopancreatography (ERCP) is now commonly used for biliary strictures and choledocholithiasis. Coronary angioplasty and percutaneous coronary interventions have decreased the need for coronary artery bypass grafting (CABG). The use of proton-pump inhibitors (PPIs) along with antibiotic treatment of Helicobacter pylori has usurped the need for surgical treatment of peptic ulcer disease (PUD) with antrectomy and/or vagotomy.
In patients with cirrhosis, liver failure is the most common cause of postoperative death.  Hepatocellular injury is most commonly due to the effects of anesthesia, intraoperative hypotension, sepsis, or viral hepatitis. A low threshold is generally maintained for postoperative transfer to the intensive care unit (ICU).
Patients must be observed closely for signs of acute hepatic decompensation, such as worsening jaundice, encephalopathy, and ascites. Sedatives and pain medications should be carefully titrated to prevent an exacerbation of hepatic encephalopathy; the increased half-life of hepatically metabolized drugs will make patients with liver disease more sensitive to standard doses. Benzodiazepines can be particularly problematic in patients predisposed to hepatic encephalopathy. Poor stooling, for example due to postoperative ileus or narcotic- or immobility-related constipation, despite lactulose dosing, can also contribute to postoperative encephalopathy.
Renal function should also be monitored because of the risk of hepatorenal syndrome and fluid shifts that occur due to surgery. These patients should also be monitored for surgical site complications such as infections, bleeding, and dehiscence. Additionally, it is now recognized that an elevated international normalized ratio (INR) in the setting of chronic liver disease does not appear to protect patients from hospital-acquired deep venous thromboses or pulmonary emboli. 
Serious sequelae of decompensated cirrhosis include severe sepsis and secondary disseminated intravascular coagulation (DIC). These potential complications emphasize the need for maintaining a low threshold for ICU-level monitoring.
Surgery in a patient with liver disease, especially end-stage liver disease with cirrhosis and portal hypertension, poses a formidable challenge for all physicians involved. Targeted interventions before surgery may help to prevent complications and improve outcomes.
The cornerstones of perioperative management are medical treatment of the complications of liver disease, including coagulopathy, ascites, encephalopathy, and malnutrition. Attention must also be paid to risk factors for infection and renal dysfunction after surgery. Sepsis, coagulopathy, and emergency surgery are most strongly correlated with postoperative mortality.
Evolving knowledge of the effects of anesthesia, improving surgical techniques, and use of improved diagnostic tests will help reduce perioperative complications.  Established risk stratification systems such as the CTP score, the MELD score, and the ASA physical status class should also be used when evaluating a patient with liver disease for potential surgery. Surgery-specific factors should also be strongly weighed. Therefore, a multidisciplinary approach to postoperative care is imperative and should include input from anesthesiologists, surgeons, internists, and hepatologists.
General considerations are as follows (see image above):
Surgery is contraindicated in patients with CTP class C, high MELD score, ASA class V, acute hepatitis, severe coagulopathy, or severe extrahepatic manifestations of liver disease (eg, acute renal failure, hypoxia, cardiomyopathy).
Avoid surgery if possible in patients with a MELD score of greater than or equal to 8 or CTP class B unless they have undergone a thorough preoperative evaluation and preparation.
Use caution with sedatives and neuromuscular blocking agents.
Optimize medical therapy for patients with cirrhosis.
- Correct coagulopathy with vitamin K and FFP to achieve prothrombin time within 3 seconds of normal.
- The goal platelet count is >50-100 × 103/L but may vary depending on the specific surgery.
- Minimize ascites to decrease risk of abdominal-wall herniation, wound dehiscence, and problems with ventilation.
- Address nutritional status.
Perform close postoperative monitoring
- Admission to the ICU may be appropriate after prolonged surgeries, intraoperative hypotension, excessive blood loss, or cardiac and/or pulmonary surgery.
- Monitor for signs of acute liver failure, including worsening jaundice, encephalopathy, and ascites.
- Monitor renal function.
- Monitor and correct electrolyte abnormalities, especially hypokalemia and metabolic alkalosis.