April 15, 1999 Table of Contents

Special Considerations in Interpreting Liver Function Tests

DAVID E. JOHNSTON, M.D., University of New Mexico School of Medicine, Albuquerque, New Mexico
Am Fam Physician. 1999 Apr 15;59(8):2223-2230.

A number of pitfalls can be encountered in the interpretation of common blood liver function tests. These tests can be normal in patients with chronic hepatitis or cirrhosis. The normal range for aminotransferase levels is slightly higher in males, nonwhites and obese persons. Severe alcoholic hepatitis is sometimes confused with cholecystitis or cholangitis. Conversely, patients who present soon after passing common bile duct stones can be misdiagnosed with acute hepatitis because aminotransferase levels often rise immediately, but alkaline phosphatase and -glutamyltransferase levels do not become elevated for several days. Asymptomatic patients with isolated, mild elevation of either the unconjugated bilirubin or the -glutamyltransferase value usually do not have liver disease and generally do not require extensive evaluation. Overall hepatic function can be assessed by applying the values for albumin, bilirubin and prothrombin time in the modified Child-Turcotte grading system.

The commonly used liver function tests (LFTs) primarily assess liver injury rather than hepatic function. Indeed, these blood tests may reflect problems arising outside the liver, such as hemolysis (elevated bilirubin level) or bone disease (elevated alkaline phosphatase [AP] level). Abnormal LFTs often, but not always, indicate that something is wrong with the liver, and they can provide clues to the nature of the problem. However, normal LFTs do not always mean that the liver is normal. Patients with cirrhosis and bleeding esophageal varices can have normal LFTs. Of the routine LFTs, only serum albumin, bilirubin and prothrombin time (PT) provide useful information on how well the liver is functioning. The general subject of LFTs and the differential diagnosis of abnormal LFTs in asymptomatic
1,2

patients

35

have been well reviewed. This article discusses some common pitfalls in the interpretation

of LFTs. Hints for interpreting these tests are presented in Table 1.

TABLE 1

Helpful Hints for Interpreting Liver Function Tests

Situation Comments Mildly elevated ALT level (less ALT value could be normal for gender, ethnicity or body mass index. than 1.5 times normal) Alcoholic hepatitis Consider muscle injury or myopathy. Laboratory values can appear cholestatic, and symptoms can mimic cholecystitis. Minimal elevations of AST and ALT often occur. AST level greater than 500 U per L The AST elevation is unlikely to result from alcohol intake alone. In a heavy drinker, consider acetaminophen toxicity.

Situation Common bile duct stone

Comments Condition can simulate acute hepatitis AST and ALT become elevated immediately, but elevation of AP and GGT is delayed.

Isolated elevation of GGT level Isolated elevation of AP level (asymptomatic patient with normal GGT level) Isolated elevation of unconjugated bilirubin level Low albumin level

This situation may be induced by alcohol and aromatic medications, usually with no actual liver disease. Consider bone growth or injury, or primary biliary cirrhosis. AP level rises in late pregnancy. Consider Gilbert syndrome or hemolysis. Low albumin is most often caused by acute or chronic inflammation, urinary loss, severe malnutrition or liver disease; it is sometimes caused by gastrointestinal loss (e.g., colitis or some uncommon small bowel disease). Normal values are lower in pregnancy.

Blood ammonia level

Blood ammonia values are not necessarily elevated in patients with hepatic encephalopathy. Determination of blood ammonia levels is most useful in patients with altered mental status of new onset or unknown origin.

ALT = alanine aminotransferase; AST = aspartate aminotransferase; AP = alkaline phosphatase; GGT = glutamyltransferase.

Markers of Hepatocellular Injury

The most commonly used markers of hepatocyte injury are aspartate aminotransferase (AST, formerly serum glutamic-oxaloacetic transaminase [SGOT]) and alanine aminotransferase (ALT, formerly serum glutamate-pyruvate transaminase [SGPT]). While ALT is cytosolic, AST has both cytosolic and mitochondrial forms. Hepatocyte necrosis in acute hepatitis, toxic injury or ischemic injury results in the leakage of enzymes into the circulation. However, in chronic liver diseases such as hepatitis C and cirrhosis, the serum ALT level correlates only moderately well with liver inflammation. In hepatitis C, liver cell death occurs by apoptosis (programmed cell death) as well as by necrosis. Hepatocytes dying by apoptosis presumably synthesize less AST and ALT as they wither away. This probably explains why at least one third of patients infected with hepatitis C virus have persistently normal serum ALT levels despite the presence of inflammation on liver biopsy. Patients with cirrhosis often have normal or
6,7

only slightly elevated serum AST and ALT levels. Thus, AST and ALT lack some sensitivity in detecting chronic liver injury. Of course, AST and ALT levels tend to be higher in cirrhotic patients with continuing inflammation or necrosis than in those without continuing liver injury. As markers of hepatocellular injury, AST and ALT also lack some specificity because they are found in skeletal muscle. Levels of these aminotransferases can rise to several times normal after severe muscular exertion or other muscle injury, as in polymyositis, or in the presence of hypothyroidism,
8

which can cause mild muscle injury and the release of aminotransferases. In fact, AST and ALT were once used in the diagnosis of myocardial infarction. Slight AST or ALT elevations (within 1.5 times the upper limits of normal) do not necessarily indicate liver disease. Part of this ambiguity has to do with the fact that unlike the values in many other biochemical tests, serum AST and ALT levels do not follow a normal bell-shaped distribution in the population. Instead, AST and ALT values have a skewed distribution characterized by a long
9

tail at the high end of the scale (Figure 1). For example, the mean values for ALT are very similar
5

from one population to another, but the degree to which the distribution is skewed varies by gender and ethnicity. The ALT distributions in males and nonwhites (i.e., blacks and Hispanics) tend to have a larger tail at the high end, so that more values fall above the upper limits of normal set for the average population.
10,11

Typical ALT or AST Distribution

FIGURE 1. Typical population distribution of serum alanine aminotransferase (ALT) or aspartate aminotransferase (AST) levels. The population distributions for these aminotransferases do not follow a bell-shaped curve but are skewed, with a long tail at higher values. By convention, values above the 97.5th percentile are considered elevated (shaded area). Although average values for ALT and AST are similar in different populations, the distributions are more skewed in males and in nonwhite ethnic groups (i.e., blacks and Hispanics). Figure 1 adapted with permission from Quinn PG, Johnston DE. Detection of chronic liver disease: costs and benefits. Gastroenterologist 1997;5:5877.

AST and ALT values are higher in obese patients, probably because these persons commonly have fatty livers. ALT levels have been noted to decline with weight loss. Depending on the physician's
12 13

point of view, the upper limits of normal for AST and ALT levels could be set higher for more obese persons. Rare individuals have chronically elevated AST levels because of a defect in clearance of the enzyme from the circulation. For both AST and ALT, the average values and upper limits of normal in
14

patients undergoing renal dialysis are about one half of those found in the general population. Mild
15

elevations of ALT or AST in asymptomatic patients can be evaluated efficiently by considering alcohol abuse, hepatitis B, hepatitis C and several other possible diagnoses (Table 2).
5

TABLE 2

Causes of Elevated ALT or AST Values in Asymptomatic Patients*

A Autoimmune hepatitis B Hepatitis B C Hepatitis C D Drugs or toxins E Ethanol F Fatty liver G Growths (i.e., tumors) H Hemodynamic disorder (congestive heart failure) I Iron (hemochromatosis), copper (Wilson's disease) or alpha -antitrypsin deficiency
1

M Muscle injury
ALT = alanine aminotransferase; AST = aspartate aminotransferase. *The differential diagnosis of elevated aminotransferase values is presented as a mnemonic, with the disorders not necessarily listed in the order of incidence or importance. Alcohol, hepatitis B and hepatitis C account for more than three fourths of all cases of cirrhosis. Adapted with permission from Quinn PG, Johnston DE. Detection of chronic liver disease: costs and benefits. Gastroenterologist 1997;5:5877.

Various liver diseases are associated with typical ranges of AST and ALT levels (Figure 2). ALT levels often rise to several thousand units per liter in patients with acute viral hepatitis. The highest ALT levelsoften more than 10,000 U per Lare usually found in patients with acute toxic injury subsequent to, for example, acetaminophen overdose or acute ischemic insult to the liver. AST and ALT levels usually fall rapidly after an acute insult.

Typical AST or ALT Values in Disease

FIGURE 2. Typical serum aspartate aminotransferase (AST) or alanine aminotransferase (ALT) values for various diseases. Note that the horizontal axis is a log scale. As shown, the amino-trans-ferases are often normal in patients with cirrhosis. In patients with uncomplicated alcoholic hepatitis, the AST value is rarely greater than 500 U per L and

is usually no more than 200 to 300 U per L. The highest peak aminotransferase values are found in patients with acute ischemic or toxic liver injury.

Lactate dehydrogenase (LDH) is less specific than AST and ALT as a marker of hepatocyte injury. However, it is worth noting that LDH is disproportionately elevated after an ischemic liver injury.
16

It is especially important to remember that in patients with acute alcoholic hepatitis, the serum AST level is almost never greater than 500 U per L and the serum ALT value is almost never greater than 300 U per L. The reasons for these limits on AST and ALT elevations are not well understood. In typical viral or toxic liver injury, the serum ALT level rises more than the AST value, reflecting the relative amounts of these enzymes in hepatocytes. However, in alcoholic hepatitis, the ratio of AST to ALT is greater than 1 in 90 percent of patients and is usually greater than 2. The higher the AST-to17

ALT ratio, the greater the likelihood that alcohol is contributing to the abnormal LFTs. In the absence of alcohol intake, an increased AST-to-ALT ratio is often found in patients with cirrhosis. The elevated AST-to-ALT ratio in alcoholic liver disease results in part from the depletion of vitamin B (pyridoxine) in chronic alcoholics. ALT and AST both use pyridoxine as a coenzyme, but the
6 18

synthesis of ALT is more strongly inhibited by pyridoxine deficiency than is the synthesis of AST. Alcohol also causes mitochondrial injury, which releases the mitochondrial isoenzyme of AST. Patients with alcoholic hepatitis can present with jaundice, abdominal pain, fever and a minimally elevated AST value, thereby leading to a misdiagnosis of cholecystitis. This is a potentially fatal mistake given the high surgical mortality rate in patients with alcoholic hepatitis.
19

Markers of Cholestasis
Cholestasis (lack of bile flow) results from the blockage of bile ducts or from a disease that impairs bile formation in the liver itself. AP and -glutamyltransferase (GGT) levels typically rise to several times the normal level after several days of bile duct obstruction or intrahepatic cholestasis. The highest liver AP elevationsoften greater than 1,000 U per L, or more than six times the normal valueare found in diffuse infiltrative diseases of the liver such as infiltrating tumors and fungal infections. Diagnostic confusion can occur when a patient presents within a few hours after acute bile duct obstruction from a gallstone. In this situation, AST and ALT levels often reach 500 U per L or more in the first hours and then decline, whereas AP and GGT levels can take several days to rise. Both AP and GGT levels are elevated in about 90 percent of patients with cholestasis. The elevation
20

of GGT alone, with no other LFT abnormalities, often results from enzyme induction by alcohol or aromatic medications in the absence of liver disease. The GGT level is often elevated in persons who take three or more alcoholic drinks (45 g of ethanol or more) per day. Thus, GGT is a useful marker
21

for immoderate alcohol intake. Phenobarbital, phenytoin (Dilantin) and other aromatic drugs typically cause GGT elevations of about twice normal. A mildly elevated GGT level is a typical finding in patients taking anticonvulsants and by itself does not necessarily indicate liver disease.
22,23

Serum AP originates mostly from liver and bone, which produce slightly different forms of the enzyme. The serum AP level rises during the third trimester of pregnancy because of a form of the enzyme produced in the placenta. When serum AP originates from bone, clues to bone disease are often present, such as recent fracture, bone pain or Paget's disease of the bone (often found in the elderly). Like the GGT value, the AP level can become mildly elevated in patients who are taking phenytoin.
22,23

If the origin of an elevated serum AP level is in doubt, the isoenzymes of AP can be separated by electrophoresis. However, this process is expensive and usually unnecessary because an elevated liver AP value is usually accompanied by an elevated GGT level, an elevated 5-nucleotidase level and other LFT abnormalities. In one study, isolated AP elevations were evaluated in an unselected group of patients at a Veterans
24

Affairs hospital. Most mild AP elevations (less than 1.5 times normal) resolved within six months, and almost all greater elevations had an evident cause that was found on routine clinical evaluation. Persistently elevated liver AP values in asymptomatic patients, especially women, can be caused by primary biliary cirrhosis, which is a chronic inflammatory disorder of the small bile ducts. Serum antimitochondrial antibody is positive in almost all of these patients.

Indicators of How Well the Liver Functions

BILIRUBIN

Bilirubin results from the enzymatic breakdown of heme. Unconjugated bilirubin is conjugated with glucuronic acid in hepatocytes to increase its water solubility and is then rapidly transported into bile. The serum conjugated bilirubin level does not become elevated until the liver has lost at least one half of its excretory capacity. Thus, a patient could have obstruction of either the left or right hepatic duct without a rise in the bilirubin level. Because the secretion of conjugated bilirubin into bile is very rapid in comparison with the conjugation step, healthy persons have almost no detectable conjugated bilirubin in their blood. Liver disease mainly impairs the secretion of conjugated bilirubin into bile. As a result, conjugated bilirubin is rapidly filtered into the urine, where it can be detected by a dipstick test. The finding of bilirubin in urine is a particularly sensitive indicator of the presence of an increased serum conjugated bilirubin level. In many healthy persons, the serum unconjugated bilirubin is mildly elevated to a concentration of 2 to 3 mg per dL (34 to 51 mol per L) or slightly higher, especially after a 24-hour fast. If this is the only LFT abnormality and the conjugated bilirubin level and complete blood count are normal, the diagnosis is usually assumed to be Gilbert syndrome, and no further evaluation is required. Gilbert syndrome was recently shown to be related to a variety of partial defects in uridine diphosphateglucuronosyl transferase, the enzyme that conjugates bilirubin.
25

Mild hemolysis, such as that caused by hereditary spherocytosis and other disorders, can also result in elevated unconjugated bilirubin values, but hemolysis is not usually present if the hematocrit and blood smear are normal. The presence of hemolysis can be confirmed by testing other markers, such as haptoglobin, or by measuring the reticulocyte count. Severe defects in bilirubin transport and conjugation can lead to markedly elevated unconjugated bilirubin levels, which can cause serious neurologic damage (kernicterus) in infants. However, no serious form of liver disease in adults causes elevation of unconjugated bilirubin levels in the blood without also causing elevation of conjugated bilirubin values. When a patient has prolonged, severe biliary obstruction followed by the restoration of bile flow, the serum bilirubin level often declines rapidly for several days and then slowly returns to normal over a period of weeks. The slow phase of bilirubin clearance results from the presence of delta-bilirubin, a form of bilirubin chemically attached to serum albumin. Because albumin has a half-life of three
26

weeks, delta-bilirubin clears much more slowly than bilirubin-glucuronide. Clinical laboratories can measure delta-bilirubin concentrations, but such measurements are usually unnecessary if the physician is aware of the delta-bilirubin phenomenon.
ALBUMIN

Although the serum albumin level can serve as an index of liver synthetic capacity, several factors make albumin concentrations difficult to interpret. The liver can synthesize albumin at twice the
27

healthy basal rate and thus partially compensate for decreased synthetic capacity or increased albumin losses. Albumin has a plasma half-life of three weeks; therefore, serum albumin concentrations change slowly in response to alterations in synthesis. Furthermore, because two thirds of the amount of body albumin is located in the extravascular, extracellular space, changes in distribution can alter the serum concentration. In practice, patients with low serum albumin concentrations and no other LFT abnormalities are likely to have a nonhepatic cause for low albumin, such as proteinuria or an acute or chronic inflammatory state. Albumin synthesis is immediately and severely depressed in inflammatory states such as burns, trauma and sepsis, and it is commonly depressed in patients with active rheumatic disorders or severe end-stage malnutrition. In addition, normal albumin values are lower in pregnancy.
PROTHROMBIN TIME

The liver synthesizes blood clotting factors II, V, VII, IX and X. The prothrombin time (PT) does not become abnormal until more than 80 percent of liver synthetic capacity is lost. This makes PT a relatively insensitive marker of liver dysfunction. However, abnormal PT prolongation may be a sign of serious liver dysfunction. Because factor VII has a short half-life of only about six hours, it is sensitive to rapid changes in liver synthetic function. Thus, PT is very useful for following liver function in patients with acute liver failure. An elevated PT can result from a vitamin K deficiency. This deficiency usually occurs in patients with chronic cholestasis or fat malabsorption from disease of the pancreas or small bowel. A trial of

vitamin K injections (e.g., 5 mg per day administered subcutaneously for three days) is the most practical way to exclude vitamin K deficiency in such patients. The PT should improve within a few days.
BLOOD AMMONIA

Measurement of the blood ammonia concentration is not always useful in patients with known or suspected hepatic encephalopathy. Ammonia contributes to hepatic encephalopathy; however, ammonia concentrations are much higher in the brain than in the blood and therefore do not correlate well. Furthermore, ammonia is not the only waste product responsible for encephalopathy. Thus,
28

blood ammonia concentrations show only a mediocre correlation with the level of mental status in patients with liver disease. It is not unusual for the blood ammonia concentration to be normal in a patient who is in a coma from hepatic encephalopathy. Blood ammonia levels are best measured in arterial blood because venous concentrations can be elevated as a result of muscle metabolism of amino acids. Blood ammonia concentrations are most useful in evaluating patients with stupor or coma of unknown origin. It is not necessary to evaluate blood ammonia levels routinely in patients with known chronic liver disease who are responding to therapy as expected.

Grading Liver Function by Child-Turcotte Class

In communicating among themselves, many physicians use the Child-Turcotte class as modified by Pugh, often termed the Child class, to convey information about overall liver function and prognosis (Table 3). This grading system can be used to predict overall life expectancy and surgical
29

mortality in patients with cirrhosis and other liver diseases.

TABLE 3

30

Grading Liver Function Using the Child-Turcotte Class as Modified by Pugh*

Points 0 More than 3.5 g per dL (35 g per L) Bilirubin Prolongation of prothrombin time Ascites Encephalopathy None None Controlled Controlled Refractory Refractory Less than 2 mg per dL (34 mol per L) Less than 4 seconds

Feature Albumin

1 2 2.8 to 3.5 g per dL (28 to Less than 2.8 g per dL 35 g per L) 2 to 3 mg per dL (34 to 51 mol per L) 4 to 6 seconds More than 6 seconds (28 g per L) More than 3 mg per dL

*The Child-Turcotte class, as modified by Pugh, often known simply as the Child class, is calculated by adding the points as determined by the patient's laboratory results: class A = 0 to 1; class B = 2 to 4; class C = 5 and higher. The classes indicate severity of liver dysfunction: class A is associated with a good prognosis, and class C is associated with limited life expectancy. Ascites and encephalopathy are graded as none, controlled with routine medical therapy or refractory to medical therapy.

Adapted with permission from Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:6469.

For elective general abdominal surgery, perioperative mortality is in the neighborhood of several percent for patients who fall into the Child class A, 10 to 20 percent for those in class B and approximately 50 percent for those in class C. These percentages must be balanced by prognostic
31

considerations when transplantation becomes an option. The presence of cirrhosis by itself is not an indication for liver transplantation, and transplantation is rarely performed in patients who fall into Child class A. For example, the 10-year survival rate is as high as 80 percent in patients with hepatitis C and cirrhosis who have Child class A liver function and no variceal bleeding. However, once
32

patients with any type of liver disease fall into the Child-Turcotte class B or class C category, survival is significantly reduced and transplantation should be considered.

The Author
DAVID E. JOHNSTON, M.D., is associate professor of medicine in the Division of Gastroenterology at the University of New Mexico School of Medicine, Albuquerque. After graduating from the University of Pittsburgh School of Medicine, Dr. Johnston completed a residency in medicine at Brigham and Women's Hospital, Boston, and a fellowship in gastroenterology at New England Medical Center, also in Boston. Address correspondence to David E. Johnston, M.D., Division of Gastroenterology/ACC-5, University of New Mexico School of Medicine, 2211 Lomas Blvd. NE, Albuquerque, NM 871315271. Reprints are not available from the author.
REFERENCES 1. Kaplan MM. Laboratory tests. In: Schiff L, Schiff ER, eds. Diseases of the liver 7th ed. Philadelphia: Lippincott, 1993:10844. 2. Kamath PS. Clinical approach to the patient with abnormal liver function test results. Mayo Clin Proc. 1996;71:108994. 3. Theal RM, Scott K. Evaluating asymptomatic patients with abnormal liver function test results. Am Fam Physician. 1996;53:21119. 4. Goddard CJ, Warnes TW. Raised liver enzymes in asymptomatic patients: investigation and outcome. Dig Dis. 1992;10:21826. 5. Quinn PG, Johnston DE. Detection of chronic liver disease: costs and benefits. Gastroenterologist. 1997;5:5877. 6. Healey CJ, Chapman RW, Fleming KA. Liver histology in hepatitis C infection: a comparison between patients with persistently normal or abnormal transaminases. Gut. 1995;37:2748. 7. Haber MM, West AB, Haber AD, Reuben A. Relationship of aminotranferases to liver histological status in chronic hepatitis C. Am J Gastroenterol. 1995;90:12507. 8. Helfgott SM, Karlson E, Beckman E. Misinterpretation of serum transaminase elevation in occult myositis. Am J Med. 1993;95:4479. 9. Sherman KE. Alanine aminotransferase in clinical practice. Arch Intern Med. 1991;151:2605. 10. Carter-Pokras OD, Najjar MF, Billhymer BF, Shulman IA. Alanine aminotransferase levels in Hispanics. Ethnic Dis. 1993;3:17680. 11. Manolio TA, Burke GL, Savage PJ, Jacobs DR Jr, Sidney S, Wagenknecht LE, et al. Sex- and race-related differences in liver-associated serum chemistry tests in young adults in the CARDIA study. Clin Chem. 1992;38:18539. 12. Salvaggio A, Periti M, Miano L, Tavanelli M, Marzorati D. Body mass index and liver enzyme activity in serum. Clin Chem. 1991;37:7203. 13. Palmer M, Schaffner F. Effect of weight reduction on hepatic abnormalities in overweight patients. Gastroenterology. 1990;99:140813. 14. Vajro P, Lofrano MM, Fontanella A, Fortunato G. Immunoglobulin complexed AST (macro-AST) in an asymptomatic child with persistent hyper-transaminasemia. J Pediatr Gastroenterol Nutr. 1992;15:45860. 15. Yasuda K, Okuda K, Endo N, Ishiwatari Y, Ikeda R, Hayashi H, et al. Hypoaminotransferasemia in patients undergoing long-term hemodialysis: clinical and biochemical appraisal. Gastroenterology. 1995;109:1295300. 16. Gitlin N, Serio KM. Ischemic hepatitis: widening horizons. Am J Gastroenterol. 1992;87:8316.

17. Cohen JA, Kaplan MM. The SGOT/SGPT ratioan indicator of alcoholic liver disease. Dig Dis Sci. 1979;24:8358. 18. Diehl AM, Potter J, Boitnott J, Van Duyn MA, Herlong HF, Mezey E. Relationship between pyridoxal 5-phosphate deficiency and aminotransferase levels in alcoholic hepatitis. Gastroenterology. 1984;86:6326. 19. Greenwood SM, Leffler CT, Minkowitz S. The increased mortality rate of open liver biopsy in alcoholic hepatitis. Surg Gynecol Obstet. 1972;134:6004. 20. Whitfield JB, Pounder RE, Neale G, Moss DW. Serum -glutamyl transpeptidase activity in liver disease. Gut. 1972;13:7028. 21. Whitehead TP, Clarke CA, Whitfield AG. Biochemical and hematological markers of alcohol intake. Lancet. 1978;1(8071):97881. 22. Keeffe EB, Sunderland MC, Gabourel JD. Serum -glutamyl transpeptidase activity in patients receiving chronic phenytoin therapy. Dig Dis Sci. 1986;31:105661. 23. Mendis GP, Gibberd FB, Hunt HA. Plasma activities of hepatic enzymes in patients on anticonvulsant therapy. Seizure. 1993;2:31923. 24. Lieberman D, Phillips D. Isolated elevation of alkaline phosphatase: significance in hospitalized patients. J Clin Gastroenterol. 1990;12:4159. 25. Bosma PJ, Chowdhury JR, Bakker C, Gantla S, de Boer A, Oostra BA, et al. The genetic basis of the reduced expression of bilirubin UDP-glucuronosyl-transferase 1 in Gilbert's syndrome. N Engl J Med. 1995;333:11715. 26. Westwood A. The analysis of bilirubin in serum. Ann Clin Biochem. 1991;28:11930. 27. Rothschild MA, Oratz M, Schreiber SS. Serum albumin. Hepatology. 1988;8:385401. 28. Basile AS, Jones EA. Ammonia and GABA-ergic neurotransmission: interrelated factors in the pathogenesis of hepatic encephalopathy. Hepatology. 1997;25:13035. 29. Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg. 1973;60:6469. 30. Villeneuve JP, Infante-Rivard C, Ampelas M, Pomier-Layrargues G, Huet PM, Marleau D. Prognostic value of the aminopyrine breath test in cirrhotic patients. Hepatology. 1986;6:92831. 31. Garrison RN, Cryer HM, Howard DA, Polk HC Jr. Clarification of risk factors for abdominal operations in patients with hepatic cirrhosis. Ann Surg. 1984;199:64855. 32. Fattovich G, Giustina G, Degos F, Tremolada F, Diodati G, Almasio P, et al. Morbidity and mortality in compensated cirrhosis type C: a retrospective follow-up study of 384 patients. Gastroenterology. 1997;112:46372.

Copyright 1999 by the American Academy of Family Physicians. This content is owned by the AAFP. A person viewing it online may make one printout of the material and may use that printout only for his or her personal, non-commercial reference. This material may not otherwise be downloaded, copied, printed, stored, transmitted or reproduced in any medium, whether now known or later invented, except as authorized in writing by the AAFP. Contact afpserv@aafp.org for copyright questions and/or permission requests. AFP Home | About Us | Contact Us | Subscribe | AFP by E-Mail | Permissions About Online Access | Employment Opportunities Information for: Authors | Advertisers