Although harmless in adults the visible nature of hyperbilirubinaemia makes this an obvious marker of liver disease. Jaundice may, however, result from three distinct processes:

These causes can be inter-related. For example, cirrhosis with some liver cell necrosis may be accompanied by intrahepatic cholestasis and haemolysis.

Patients with carotenaemia from excess dietary carotene or associated with hypothyroidism may also develop yellow skin, but unlike hyperbilirubinaemia there is no conjunctival colouring.

Jaundice is usually detectable when serum bilirubin rises above 50 μmol/I and can often be seen at lower levels. Variable tissue levels in fluctuating jaundice may mean that skin and conjunctival appearances do not correlate. Laboratory measurements of serum bilirubin are based on a diazo colour reaction which forms the purple azobilirubin. Conjugated (direct) bilirubin reacts quickly. Unconjugated (indirect) bilirubin reacts slowly and requires the addition of alcohol for complete reaction. There are considerable technical problems associated with fractionation of bilirubin, and at very low levels of total bilirubin, as well as when there is considerable elevation, the ratio of conjugated to unconjugated bilirubin is unreliable. An alternative assay using alkaline methanolysis and high performance liquid chromatography yields lower and different results, which could be useful in assessing the significance of marginal hyperbilirubinaemia.

Conjugated, but not unconjugated bilirubin is excreted in the urine.

Urobilinogens are formed from bilirubin by bacterial action in the intestine. Most are excreted in the faeces but some are absorbed and excreted in the urine. The excretion is maximal between 2 and 4 pm, and is enhanced by an alkaline urine. On exposure to air the urobilinogen is oxidized to urobilin, which darkens the urine.

A number of intracellular enzymes appear in the serum when liver cells are damaged. Different patterns of elevation suggest different disorders, but none is pathognomic. Elevated serum levels of enzymes are due to their leakage from cells linked with the increased synthesis of enzymes because of induction prior to necrosis.

Multiple coagulation defects are not uncommon in patients with acute and chronic liver disease. Combined deficiencies of factors II (prothrombin), V, VII and X contribute to an abnormally prolonged prothrombin time. The determination of the one-stage prothrombin time is a useful simple test of liver function.

Because vitamin K is a co-factor of hepatic prothrombin synthesis there may be a prolonged prothrombin time in cholestatic jaundice from any cause. The ability of parenteral vitamin K (10 mg vitamin K₁ given IV/IM for 3 days) to convert the prothrombin time to normal values has been used as a diagnostic test for the aetiology of jaundice. Patients with extrahepatic biliary obstruction respond to vitamin K_I injections, but in severe hepatocellular disease the prothrombin time remains unchanged. This is not a reliable diagnostic test.

Other haematological defects which may be found in liver disease include deficiencies of factors IX (plasma thromboplastin component), XI (plasma thromboplastin antecedent) and platelets.

Liver disease may be accompanied by diffuse intravascular coagulation (DIC), in which fibrin degradation products appear in the serum (>40 mg/l), the platelet count falls sharply and there is evidence of haemolysis.

Ammonia levels rise in hepatic coma. Methods for measuring the blood ammonia concentration are complex and this is not performed routinely in the management of patients with liver failure. While venous or arterial blood can be sampled the latter is favoured. The normal arterial blood ammonia concentration is <I mg/l.

Elevated concentrations may be found in hepatocellular failure or when there is shunting of blood from the liver. Arterial blood ammonia levels do not correlate well with the clinical severity of hepatic coma and this correlation is even poorer if venous samples are measured. Elevations of blood ammonia concentration are also found in a variety of rare congenital defects of urea synthesis.

It is possible to measure the concentration of serum bile acids by enzyme fluorimetry or gas liquid chromatography. Both total and the major individual bile acids can be quantitated accurately. Normal fasting levels are <4.5 μmol/l, and postprandial levels <6.5 μmol/l: these are increased in a wide variety of hepatobiliary diseases. Serum bile acids vary with fasting and feeding, during the menstrual cycle, and with vitamin C status in liver disease. Patients on ursodeoxycholic acid (UDCA) therapy regularly have fasting serum UDCA levels of 4+ μmol/l: lower values indicate non-compliance.

There can be a seven-fold fluctuation of total bile acid values through the day. Levels may also be elevated in bacterial colonization of the small bowel and associated with hyperlipidaemia. Serum bile acids are, therefore, not entirely specific to hepatobiliary disease, and although they can be used for screening and monitoring liver disease they do not add further information to the conventional screen of ‘liver function’ tests.

Bile acid tolerance tests and clearance studies in which serum levels are measured after oral or intravenous administration of unlabelled or radio-labelled bile acids, give results which are too variable to be helpful in individual diagnosis.

Serum levels of 7-alpha-hydroxycholesterol (the essential precursor in hepatic bile acid synthesis) correlate with liver cirrhosis.

One ‘normal’ upper limit for serum cholesterol is 7.8 mmol/l, and for triglycerides is 2.5 mmol/I; higher values are definitely abnormal. The serum level of total cholesterol rises in both intra- and extrahepatic cholestasis. This results from the presence of an abnormal lipoprotein (LPX) in the serum which can be measured immunochemically. Very low levels of high-density lipoprotein (HDL) cholesterol are characteristic of cholestasis: the lower limit of normal is about I mmol/l.

The presence of altered or abnormal lipoprotein components can be associated with many liver diseases. Raised levels of cholesterol, triglycerides, low-density (LDL) and very low-density (VLDL) lipoprotein in various combinations is seen.

This may be important since markedly elevated levels of serum triglycerides (> 10 mmol/I) cause turbidity and interfere with most other biochemical measurements. Alcoholism is the most common cause of secondary hyperlipidaemia, and may itself cause cirrhosis and pancreatitis.

Fluid-electrolyte disturbances including secondary aldosteronism are encountered in liver disease: hyponatraemia (Na <130 mmol/l) and hypokalaemia (K < 3.5 mmol/l) are common. Although low serum sodium levels are often well tolerated, low serum potassium can potentiate hepatic encephalopathy. Urea is synthesized in the liver: low levels (<3.3 mmol/l) may indicate severe hepatocellular dysfunction but can also reflect dilution with fluid retention. Blood urea levels may be apparently normal in patients with liver disease and associated renal impairment, and serum creatinine (normal range 45–150 μmol/l) is a much better index of renal failure.

Vitamin B_I2 is normally present in liver cells and levels are elevated in metastatic liver disease, liver abscess and hepatitis. Levels also rise in patients receiving hydroxocobalamin therapy. Plasma glucose levels may be informative as both diabetes mellitus and hypoglycaemia occur in liver disease.

Alcohol is the most common cause of liver disease in Western societies. The main hurdle in diagnosis is to suspect the cause, and the patient’s general demeanour may give clues.

The ‘CAGE’ questionnaire is a simple 4-point system to assess alcohol abuse. A patient who answers ‘yes’ to all four questions is an alcoholic, and a score of two or three out of four is suspicious.

Patients may be teetotal at the time when they are suffering the effects of previous heavy drinking, and an assessment of the amount drunk needs to take into account changing patterns.

Patients are not always honest about excess alcohol intake, and laboratory tests are often valuable in establishing diagnoses. They are not infallible, and all can produce normal results in people with severe alcoholic liver disease. In addition, a significant minority of alcohol abusers have non-alcohol-related liver disease.

Measurement of the alcohol level in the blood is extremely useful. If there is any alcohol at all in a morning sample then the patient is probably drinking to excess.

The assessment of long-term heavy drinking is helped by various tests. The most useful are raised gamma-GT levels (>50 lU/l) and mean corpuscular volumes (>95 fl). The alkaline phosphatase level may also be raised to a lesser extent, and the platelet count reduced. Chest radiology may reveal old or recent rib fractures in binge drinkers. Specialized tests such as measurement of glutamate dehydrogenase, mitochondrial AST, carbohydrate-deficient transferrin and apolipoprotein A₁ may prove useful but are not generally available.

Useful in vitro migration and transformation tests may be performed by culture and challenge of lymphocytes with drugs suspected of causing toxic reactions.These should be considered if a patient has suffered a serious drug reaction: they may establish the diagnosis without the need for potentially hazardous in vivo challenge tests.

The peripheral T-cell population is reduced in alcoholic liver disease, chronic acute hepatitis and primary biliary cirrhosis. Human leukocyte antigens HLA-B40 and HLA-B8 are said to be more frequent in patients with alcoholic cirrhosis than in non-cirrhotic alcoholics and cirrhosis of other causes.