The diagnosis of minimal hepatic encephalopathy requires neuropsychological testing by definition. Older tests include the "numbers connecting test" A and B (measuring the speed at which one could connect randomly dispersed numbers 1–20), the "block design test" and the "digit-symbol test". In 2009 an expert panel concluded that neuropsychological test batteries aimed at measuring multiple domains of cognitive function are generally more reliable than single tests, and tend to be more strongly correlated with functional status. Both the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) and PSE-Syndrom-Test may be used for this purpose. The PSE-Syndrom-Test, developed in Germany and validated in several other European countries, incorporates older assessment tools such as the number connection test.

The diagnosis of hepatic encephalopathy can only be made in the presence of confirmed liver disease (types A and C) or a portosystemic shunt (type B), as its symptoms are similar to those encountered in other encephalopathies. To make the distinction, abnormal liver function tests and/or ultrasound suggesting liver disease are required, and ideally liver biopsy. The symptoms of hepatic encephalopathy may also arise from other conditions, such as cerebral haemorrhage and seizures (both of which are more common in chronic liver disease). A CT scan of the brain may be required to exclude haemorrhage, and if seizure activity is suspected an electroencephalograph (EEG) study may be performed. Rarer mimics of encephalopathy are meningitis, encephalitis, Wernicke's encephalopathy and Wilson's disease; these may be suspected on clinical grounds and confirmed with investigations.

The diagnosis of hepatic encephalopathy is a clinical one, once other causes for confusion or coma have been excluded; no test fully diagnoses or excludes it. Serum ammonia levels are elevated in 90% of people, but not all hyperammonaemia (high ammonia levels) is associated with encephalopathy. A CT scan of the brain usually shows no abnormality except in stage IV encephalopathy, when cerebral oedema may be visible. Other neuroimaging modalities, such as magnetic resonance imaging (MRI), are not currently regarded as useful, although they may show abnormalities. Electroencephalography shows no clear abnormalities in stage 0, even if minimal HE is present; in stages I, II and III there are triphasic waves over the frontal lobes that oscillate at 5 Hz, and in stage IV there is slow delta wave activity. However, the changes in EEG are not typical enough to be useful in distinguishing hepatic encephalopathy from other conditions.

Once the diagnosis of encephalopathy has been made, efforts are made to exclude underlying causes (such as listed above in "causes"). This requires blood tests (urea and electrolytes, full blood count, liver function tests), usually a chest X-ray, and urinalysis. If there is ascites, diagnostic paracentesis (removal of a fluid sample with a needle) may be required to identify spontaneous bacterial peritonitis (SBP).

All patients with clinical or laboratory evidence of moderate to severe acute hepatitis should have an immediate measurement of prothrombin time and careful evaluation of mental status. If the prothrombin time is prolonged by ≈ 4–6 seconds or more (INR ≥ 1.5),

and there is any evidence of altered sensorium, the diagnosis of ALF should be strongly suspected, and hospital admission is mandatory. Initial laboratory examination must be extensive in order to evaluate both the etiology and severity.

- Initial laboratory analysis

- Prothrombin time/INR

- Complete blood count

- Chemistries

- Liver function test: AST, ALT, alkaline phosphatase, GGT, total bilirubin, albumin

- Creatinine, urea/blood urea nitrogen, sodium, potassium, chloride, bicarbonate, calcium, magnesium, phosphate

- Glucose

- Amylase and lipase

- Arterial blood gas, lactate

- Blood type and screen

- Paracetamol (acetaminophen) level, toxicology screen

- Viral hepatitis serologies: anti-HAV IgM, HBSAg, anti-HBc IgM, anti-HCV

- Autoimmune markers: ANA, ASMA, LKMA, immunoglobulin levels

- Ceruloplasmin level (when Wilson's disease suspected)

- Pregnancy test (females)

- Ammonia (arterial if possible)

- HIV status (has implication for transplantation)

History taking should include a careful review of possible exposures to viral infection and drugs or other toxins. From history and clinical examination, the possibility of underlying chronic disease should be ruled out as it may require different management.

A liver biopsy done via the transjugular route because of coagulopathy is not usually necessary, other than in occasional malignancies. As the evaluation continues, several important decisions have to be made; such as whether to admit the patient to an ICU, or whether to transfer the patient to a transplant facility. Consultation with the transplant center as early as possible is critical due to the possibility of rapid progression of ALF.

Acute liver failure is defined as "the rapid development of hepatocellular dysfunction, specifically coagulopathy and mental status changes (encephalopathy) in a patient without known prior liver disease".

The diagnosis of acute liver failure is based on physical exam, laboratory findings, patient history, and past medical history to establish mental status changes, coagulopathy, rapidity of onset, and absence of known prior liver disease respectively.

The exact definition of "rapid" is somewhat questionable, and different sub-divisions exist which are based on the time from onset of first hepatic symptoms to onset of encephalopathy. One scheme defines "acute hepatic failure" as the development of encephalopathy within 26 weeks of the onset of any hepatic symptoms. This is sub-divided into "fulminant hepatic failure", which requires onset of encephalopathy within 8 weeks, and "subfulminant", which describes onset of encephalopathy after 8 weeks but before 26 weeks. Another scheme defines "hyperacute" as onset within 7 days, "acute" as onset between 7 and 28 days, and "subacute" as onset between 28 days and 24 weeks.

This remains a challenge in clinical practice due to a lack of reliable markers. Many other conditions lead to similar clinical as well as pathological pictures. To diagnose hepatotoxicity, a causal relationship between the use of the toxin or drug and subsequent liver damage has to be established, but might be difficult, especially when idiosyncratic reaction is suspected. Simultaneous use of multiple drugs may add to the complexity. As in acetaminophen toxicity, well established, dose-dependent, pharmacological hepatotoxicity is easier to spot. Several clinical scales such as CIOMS/RUCAM scale and Maria and Victorino criteria have been proposed to establish causal relationship between offending drug and liver damage. CIOMS/RUCAM scale involves a scoring system that categorizes the suspicion into "definite or highly probable" (score > 8), “probable” (score 6-8), “possible” (score 3-5), “unlikely” (score 1-2) and “excluded” (score ≤ 0). In clinical practice, physicians put more emphasis on the presence or absence of similarity between the biochemical profile of the patient and known biochemical profile of the suspected toxicity (e.g., cholestatic damage in amoxycillin-clauvonic acid ).

In people with alcoholic hepatitis, the serum aspartate aminotransferase (AST) to alanine aminotransferase (ALT) ratio is greater than 2:1.AST and ALT levels are almost always less than 500. The elevated AST to ALT ratio is due to deficiency of pyridoxal-6-phosphate, which is required in the ALT enzyme synthetic pathway. Furthermore, alcohol metabolite–induced injury of hepatic mitochondria results in AST isoenzyme release. Other laboratory findings include red blood cell macrocytosis (mean corpuscular volume > 100) and elevations of serum γ-glutamyl transferase, alkaline phosphatase, and bilirubin levels. Folate level is reduced in alcoholic patients due to decreased intestinal absorption, increased bone marrow requirement for folate in the presence of alcohol, and increased urinary loss.The magnitude of leukocytosis reflects severity of liver injury. Histologic features include Mallory bodies, giant mitochondria, hepatocyte necrosis, and neutrophil infiltration at the perivenular area. Mallory bodies, which are also present in other liver diseases, are condensations of cytokeratin components in the hepatocyte cytoplasm and do not contribute to liver injury.Up to 70% of patients with moderate to severe alcoholic hepatitis already have cirrhosis identifiable on biopsy examination at the time of diagnosis.

The prognosis for people with ALD depends on the liver histology as well as cofactors, such as concomitant chronic viral hepatitis. Among patients with alcoholic hepatitis, progression to liver cirrhosis occurs at 10–20% per year, and 70% will eventually develop cirrhosis. Despite cessation of alcohol use, only 10% will have normalization of histology and serum liver enzyme levels. As previously noted, the MDF has been used to predict short-term mortality (i.e., MDF ≥ 32 associated with spontaneous survival of 50–65% without corticosteroid therapy, and MDF 11) and 90-day (MELD > 21) mortality. Liver cirrhosis develops in 6–14% of those who consume more than 60–80 g of alcohol daily for men and more than 20 g daily for women. Even in those who drink more than 120 g daily, only 13.5% will suffer serious alcohol-related liver injury. Nevertheless, alcohol-related mortality was the third leading cause of death in 2003 in the United States. Worldwide mortality is estimated to be 150,000 per year.

The severity of cirrhosis is commonly classified with the Child-Pugh score. This scoring system uses bilirubin, albumin, INR, the presence and severity of ascites, and encephalopathy to classify patients into class A, B, or C. Class A has a favourable prognosis, while class C is at high risk of death. This system was devised in 1964 by Child and Turcotte, and modified in 1973 by Pugh and others.

More modern scores, used in the allocation of liver transplants but also in other contexts, are the Model for End-Stage Liver Disease (MELD) score and its pediatric counterpart, the Pediatric End-Stage Liver Disease (PELD) score.

The hepatic venous pressure gradient, (difference in venous pressure between afferent and efferent blood to the liver) also determines the severity of cirrhosis, although it is hard to measure. A value of 16 mm or more means a greatly increased risk of death.

Ultrasound is routinely used in the evaluation of cirrhosis. It may show a small and nodular liver in advanced cirrhosis along with increased echogenicity with irregular appearing areas. Other liver findings suggestive of cirrhosis in imaging are an enlarged caudate lobe, widening of the fissures and enlargement of the spleen. An enlarged spleen (splenomegaly), which normally measures less than 11–12 cm in adults, is suggestive of cirrhosis with portal hypertension, in the right clinical context. Ultrasound may also screen for hepatocellular carcinoma, portal hypertension, and Budd-Chiari syndrome (by assessing flow in the hepatic vein).

Cirrhosis is diagnosed with a variety of elastography techniques. Because a cirrhotic liver is generally stiffer than a healthy one, imaging the liver's stiffness can give diagnostic information about the location and severity of cirrhosis. Techniques used include transient elastography, acoustic radiation force impulse imaging, supersonic shear imaging and magnetic resonance elastography. Compared to a biopsy, elastography can sample a much larger area and is painless. It shows a reasonable correlation with the severity of cirrhosis.

Other tests performed in particular circumstances include abdominal CT and liver/bile duct MRI (MRCP).

The diagnosis is made in a patient with history of significant alcohol intake who develops worsening liver function tests, including elevated bilirubin and aminotransferases. The ratio of aspartate aminotransferase to alanine aminotransferase is usually 2 or more. In most cases, the liver enzymes do not exceed 500. The changes on liver biopsy are important in confirming a clinical diagnosis.

In most cases, liver function will return to normal if the offending drug is stopped early. Additionally, the patient may require supportive treatment. In acetaminophen toxicity, however, the initial insult can be fatal. Fulminant hepatic failure from drug-induced hepatotoxicity may require liver transplantation. In the past, glucocorticoids in allergic features and ursodeoxycholic acid in cholestatic cases had been used, but there is no good evidence to support their effectiveness.

An elevation in serum bilirubin level of more than 2 times ULN with associated transaminase rise is an ominous sign. This indicates severe hepatotoxicity and is likely to lead to mortality in 10% to 15% of patients, especially if the offending drug is not stopped (Hy's Law). This is because it requires significant damage to the liver to impair bilirubin excretion, hence minor impairment (in the absence of biliary obstruction or Gilbert syndrome) would not lead to jaundice. Other poor predictors of outcome are old age, female sex, high AST.

If a liver biopsy is needed for diagnosis of the condition, the mother should be appropriately stabilized and treated to reduce bleeding related complications. The diagnosis can be made by a frozen-section (as opposed to a specimen in formalin) that is stained with the Oil red O stain, that shows microvesicular steatosis (or small collections of fat within the liver cells). The microvesicular steatosis usually spares zone one of the liver, which is the area closest to the hepatic artery. On the regular trichrome stain, the liver cell cytoplasm shows a foamy appearance due to the prominence of fat. Necrosis is rarely seen. The diagnosis can be enhanced by electron microscopy which can be used to confirm the presence of microvesicular steatosis, and specifically the presence of megamitochondria and paracrystalline inclusions. Liver diseases with similar appearances include Reye's syndrome, drug-induced hepatitis from agents with mitochondrial toxicity, including nucleoside reverse transcriptase inhibitors used to treat HIV, and a rare condition known as Jamaican vomiting sickness which is caused by the eating of the unripened Ackee fruit.

The diagnosis of acute fatty liver of pregnancy is usually made on clinical grounds by the treating physician or midwife, but differentiation from other conditions affecting the liver may be difficult. The diagnosis of acute fatty liver of pregnancy is suggested by jaundice with a lesser elevation of liver enzymes, elevated white blood cell count, disseminated intravascular coagulation, and a clinically unwell patient.

A liver biopsy can provide a definitive diagnosis, but is not always done, due to the increased chance of bleeding in acute fatty liver of pregnancy. Often testing will be done to exclude more common conditions that present in a similar fashion, including viral hepatitis, pre-eclampsia, HELLP syndrome, intrahepatic cholestasis of pregnancy, and autoimmune hepatitis.

Blood tests, cerebrospinal fluid examination by lumbar puncture (also known as spinal tap), brain imaging studies, electroencephalography (EEG), and similar diagnostic studies may be used to differentiate the various causes of encephalopathy.

Diagnosis is frequently clinical. That is, no set of tests give the diagnosis, but the entire presentation of the illness with nonspecific test results informs the experienced clinician of the diagnosis.

Most patients presenting with jaundice will have various predictable patterns of liver panel abnormalities, though significant variation does exist. The typical liver panel will include blood levels of enzymes found primarily from the liver, such as the aminotransferases (ALT, AST), and alkaline phosphatase (ALP); bilirubin (which causes the jaundice); and protein levels, specifically, total protein and albumin. Other primary lab tests for liver function include gamma glutamyl transpeptidase (GGT) and prothrombin time (PT).

Some bone and heart disorders can lead to an increase in ALP and the aminotransferases, so the first step in differentiating these from liver problems is to compare the levels of GGT, which will only be elevated in liver-specific conditions. The second step is distinguishing from biliary (cholestatic) or liver (hepatic) causes of jaundice and altered laboratory results. The former typically indicates a surgical response, while the latter typically leans toward a medical response. ALP and GGT levels will typically rise with one pattern while aspartate aminotransferase (AST) and alanine aminotransferase (ALT) rise in a separate pattern. If the ALP (10–45 IU/L) and GGT (18–85) levels rise proportionately about as high as the AST (12–38 IU/L) and ALT (10–45 IU/L) levels, this indicates a cholestatic problem. On the other hand, if the AST and ALT rise is significantly higher than the ALP and GGT rise, this indicates an hepatic problem. Finally, distinguishing between hepatic causes of jaundice, comparing levels of AST and ALT can prove useful. AST levels will typically be higher than ALT. This remains the case in most hepatic disorders except for hepatitis (viral or hepatotoxic). Alcoholic liver damage may see fairly normal ALT levels, with AST 10x higher than ALT. On the other hand, if ALT is higher than AST, this is indicative of hepatitis. Levels of ALT and AST are not well correlated to the extent of liver damage, although rapid drops in these levels from very high levels can indicate severe necrosis. Low levels of albumin tend to indicate a chronic condition, while it is normal in hepatitis and cholestasis.

Lab results for liver panels are frequently compared by the magnitude of their differences, not the pure number, as well as by their ratios. The AST:ALT ratio can be a good indicator of whether the disorder is alcoholic liver damage (above 10), some other form of liver damage (above 1), or hepatitis (less than 1). Bilirubin levels greater than 10x normal could indicate neoplastic or intrahepatic cholestasis. Levels lower than this tend to indicate hepatocellular causes. AST levels greater than 15x tends to indicate acute hepatocellular damage. Less than this tend to indicate obstructive causes. ALP levels greater than 5x normal tend to indicate obstruction, while levels greater than 10x normal can indicate drug (toxic) induced cholestatic hepatitis or Cytomegalovirus. Both of these conditions can also have ALT and AST greater than 20× normal. GGT levels greater than 10x normal typically indicate cholestasis. Levels 5–10× tend to indicate viral hepatitis. Levels less than 5× normal tend to indicate drug toxicity. Acute hepatitis will typically have ALT and AST levels rising 20–30× normal (above 1000), and may remain significantly elevated for several weeks. Acetaminophen toxicity can result in ALT and AST levels greater than 50x normal.

"Acute on chronic liver failure" is said to exist when someone with chronic liver disease develops features of liver failure. A number of underlying causes may precipitate this, such as alcohol misuse or infection. People with ACLF can be critically ill and require intensive care treatment, and occasionally a liver transplant. Mortality with treatment is 50%.

Diagnostic methods for hypertensive encephalopathy include physical examination, blood pressure measurement, blood sampling, ECG, EEG, chest X-ray, urinalysis, arterial blood gas analysis, and imaging of the head (CAT scan and/or MRI). Since decreasing the blood pressure is essential, anti-hypertensive medication is administered without awaiting the results of the laboratory tests. Electroencephalographic examination detects the absence of alpha waves, signifying impaired consciousness. In people with visual disturbances, slow waves are detected in the occipital areas.

Clinical practice guidelines by the American College of Gastroenterology have recommended corticosteroid treatment. Patients should be risk stratified using a MELD Score or Child-Pugh score.

- Corticosteroids: These guidelines suggest that patients with a modified Maddrey's discriminant function score > 32 or hepatic encephalopathy should be considered for treatment with prednisolone 40 mg daily for four weeks followed by a taper. Models such as the Lille Model can be used to monitor for improvement or to consider alternative treatment.

- Pentoxifylline: A randomized controlled trial found that among patients with a discriminant function score > 32 and at least one of the following symptoms (a palpable, tender enlarged liver, fever, high white blood cell count, hepatic encephalopathy, or hepatic systolic bruit), 4.6 patients must be treated with pentoxifylline for 4 weeks to prevent one patient from dying. Subsequent trials have suggested that pentoxifylline may be superior to prednisolone in the management of acute alcoholic hepatitis with discriminant function score >32. Advantage of pentoxifylline over prednisolone was better tolerability, lesser side effects, with decreased occurrence of renal dysfunction in patients receiving pentoxifylline.

- Potential for combined therapy: A large prospective study of over 1000 patients investigated whether prednisolone and pentoxifylline produced benefits when used alone or in combination. Pentoxifylline did not improve survival alone or in combination. Prednisolone gave a small reduction in mortality at 28 days but this did not reach significance, and there were no improvements in outcomes at 90 days or 1 year.

Rapid diagnosis is important to attempt to prevent further damage to the brain and further neurologic deficits. It is a diagnosis of exclusion, so a full work up for other possible etiologies (hepatic, uremic, infectious, oncologic) should be performed. Screening for heavy metals, as well as other toxins, should be done immediately as those are some of the most common causes and the patient can then remove themselves from the dangerous environment. In addition, a full examination of blood (CBC) and metabolites (CMP) should be done.

Treatment varies according to the type and severity of the encephalopathy. Anticonvulsants may be prescribed to reduce or halt any seizures. Changes to diet and nutritional supplements may help some patients. In severe cases, dialysis or organ replacement surgery may be needed.

Sympathomimetic drugs can increase motivation, cognition, motor performance and alertness in patients with encephalopathy caused by brain injury, chronic infections, strokes, brain tumors.

Ultrasonography (US) is the first-line imaging technique for the diagnosis and follow-up of portal hypertension because it is non-invasive, low-cost and can be performed on-site.

Signs of portal hypertension on ultrasound include dilatation of the portal vein of over 13 mm in diameter, a portal flow mean velocity of less than 12 cm/s on Doppler ultrasound, porto–systemic collateral veins (patent paraumbilical vein, spleno–renal collaterals and dilated left and short gastric veins), splenomegaly and signs of cirrhosis (including nodularity of the liver surface).

The hepatic venous pressure gradient (HVPG) measurement has been accepted as the gold standard for assessing the severity of portal hypertension. Portal hypertension is defined as HVPG greater than or equal to 5 mm Hg and is considered to be clinically significant when HVPG exceeds 10 to 12 mm Hg.

Diagnosis of Wernicke's encephalopathy or disease is made clinically. Caine et al. in 1997 established criteria that Wernicke's encephalopathy can be diagnosed in any patient with just two or more of the main symptoms noted above. The sensitivity of the diagnosis by the classic triad was 23% but increased to 85% taking two or more of the four classic features. This criteria is challenged because all the cases he studied were alcoholics.

Some consider it sufficient to suspect the presence of the disease with only one of the principal symptoms. Some British hospital protocols suspect WE with any one of these symptoms: confusion, decreased consciousness level (or unconsciousness, stupor or coma), memory loss, ataxia or unsteadiness, ophthalmoplegia or nystagmus, and unexplained hypotension with hypothermia. The presence of only one sign should be sufficient for treatment.

As a much more diverse range of symptoms has been found frequently in patients it is necessary to search for new diagnostic criteria, however Wernicke's encephalopathy remains a clinically-diagnosed condition. Neither the MR, nor serum measurements related to thiamine are sufficient diagnostic markers in all cases. Non-recovery upon supplementation with thiamine is inconclusive.

The sensitivity of MR was 53% and the specificity was 93%. The reversible cytotoxic edema was considered the most characteristic lesion of WE. The location of the lesions were more frequently atypical among non-alcoholics, while typical contrast enhancement in the thalamus and the mammillary bodies was observed frequently associated with alcohol abuse. These abnormalities may include:

- Medial thalami, periaqueductal gray matter, mamillary bodies, and brainstem nuclei edema (Zuccoli G.). Involvement is always bilateral symmetrical. Value of DWI in the diagnosis of WE is minimal. Axial FLAIR MRI images represent the best diagnostic MRI sequence. Contrast material may highlight involvement of the mamillary bodies.

There appears to be very little value for CT scans.

Thiamine can be measured using an erythrocyte transketolase activity assay, or by activation by measurement of in vitro thiamine diphosphate levels. Normal thiamine levels do not necessarily rule out the presence of WE, as this may be a patient with difficulties in intracellular transport.

Chronic liver failure usually occurs in the context of cirrhosis, itself potentially the result of many possible causes, such as excessive alcohol intake, hepatitis B or C, autoimmune, hereditary and metabolic causes (such as iron or copper overload, steatohepatitis or non-alcoholic fatty liver disease).

Blood testing usually shows high levels of the liver transaminase enzymes, AST and ALT, which may exceed 10,000 U/L. It has been found that those who suffer from ischemic hepatitis had significant cardiac disease as well.

As a measure of precaution, paracetamol levels and a toxicology screening should be completed, it is also imperative to be able to exclude the possibility of viral hepatitis.

Hepatic doppler ultrasound is typically utilized to confirm or suggest the diagnosis. Most common findings on liver doppler ultrasound include increased phasicity of portal veins with eventual development of portal flow reversal. The liver is usually enlarged but maintained normal echogenicity. A liver biopsy is required for a definitive diagnosis.