Once a diagnosis of acute liver failure is suspected or established, patients should undergo concurrent investigation and treatment with the initial aim being to identify the aetiology and institute any disease specific therapy, which may be appropriate. Depending on the initial presentation, resuscitative measures may have to be instigated simultaneously. An appreciation of factors that make for a poor prognosis also enables the early identification of patients who are likely to fail supportive medical management and require liver transplantation.

Clearly not every patient where a diagnosis of acute liver failure is made or suspected will have a metabolic derangement that would warrant intensive care unit (ICU) admission or indeed be referred to ICU. For example, patients with subacute failure may remain for long periods with low-grade encephalopathy on a medical ward.

However, it is important to bear in mind that even in the presence of hepatic regeneration patients will succumb to complications such as sepsis or multi-organ failure and will need a close level of monitoring. Furthermore, patients presenting with hyperacute or acute liver failure can rapidly progress to multi-organ failure.

As a result it is recommended that, a low threshold for admission to critical care be adopted, especially until the aetiology is established.

The main indications for ICU referral and admission are as follows [3, 8]:

The ICU clinician should anticipate the severity of presentation and likelihood of supportive and medical therapy failing despite optimal care. The most widely accepted prognostic tool in acute liver failure is the King’s College criteria.

The King’s College criteria were originally used to predict survival without liver transplantation. It is now used to refer patients to a specialist liver unit and select potential liver transplant recipients.

The King’s College criteria differ between paracetamol and non-paracetamol causes. It has a good specificity (82–92 %) but limited sensitivity (68 %). Its positive predictive value for ICU death without transplantation is 0.98 and the negative predictive value is 0.82 [7] (Tables 9.4 and 9.5).

The use of arterial lactate is used to improve the sensitivity of the King’s Criteria in the context of paracetamol-induced liver failure [9]. Lactate concentrations of >3.5 mmol/L on admission and >3.0 mmol/L after adequate fluid resuscitation, when used in combination had similar predictive ability as the King’s criteria but identified non-surviving patients earlier. The use of the post resuscitation lactate increases the sensitivity of the King’s criteria to 91 % [9]

All patients with any degree of encephalopathy, with acute liver failure should be referred to a specialist liver unit.

It should be noted that once a decision to transfer a patient with encephalopathy has been made, then sedation and artificial ventilation should be considered for safe transfer. Specifically, good practice advocates sedation and ventilation even in grade I and II encephalopathy. These patients can deteriorate rapidly with potential for associated cardiovascular collapse.

A careful history must be taken including collateral histories from family members. Clinicians in referring hospitals often have a window of opportunity to meticulously enquire for aetiology especially regarding exposure to viral infections and drugs before the onset of encephalopathy. A common reason behind indeterminate aetiologies lies in inadequate history at the time of presentation and this can complicate further management [10], thus the taking of a thorough history in identifying the aetiology cannot be overstated.

Enquiry into recent foreign travel to endemic areas, risk factors for immunosuppression (e.g. immunosuppressant drugs, history of carcinomas and chemotherapy), high-risk sexual activity and medication history including alternative medicines should be made. Intravenous and recreational drug use particularly but not limited to cocaine and amphetamines and the ingestion of poisonous wild mushrooms, usually due to misidentification, should also be excluded.

Questioning of psychiatric history and neurological disorders especially in the context of renal tubular acidosis (especially Fanconi’s syndrome) are particularly relevant, indicating Wilson’s disease, whilst history of pregnancies, miscarriages and amenorrhea could indicate autoimmune hepatitis or HELLP syndrome. If a Budd-Chiari syndrome is diagnosed, then a search for an underlying cause should be made (to include malignancy, antiphospholipid syndrome, protein S, C and antithrombin III deficiency and factor V Leiden).

It is important to remember that even paracetamol taken within recommended daily doses have been known to cause acute liver failure, thus the potential for inadvertent overdose exists, especially in high risk groups (e.g. anorexia nervosa and alcohol abuse). Particular attention should be paid to interpreting paracetamol levels in these groups and familiarity with guideline updates is advised [11].

A thorough social history is important in lieu of any potential transplant assessment.

A history or clinical stigmata suggesting underlying chronic liver disease should be sought as this would alter management.

Imaging by computerised tomography is useful in cases where there is a history of cancer, if patency of portal vessels is queried or if Budd-Chiari syndrome is suspected. It is also indicated where intracranial hypertension and cerebral oedema is suspected, but clearly the risks in transportation in this cohort of systemically unstable patients needs to be balanced against the potential management benefits of imaging.

Initial laboratory investigations should be extensive to evaluate aetiology and severity of acute liver failure [8]. It is recommended that a full liver screen is requested on first presentation, even if it is intended for the patient to be transferred to a specialist liver unit (Box 9.1). To avoid delays, these results, once known should be communicated to the specialist liver unit to expedite the commencement of any specific aetiology-based therapy or decision for listing for transplantation.

The aetiology and incidence of acute liver failure varies worldwide. Overall the incidence is significantly lower in the developed world when compared to developing countries where viral infections (hepatitis A, B and E) are the main aetiologies [12]. In the United States and Western Europe drug induced aetiologies, especially paracetamol predominates, followed by non A-E hepatitis where no cause is found (Table 9.6).

As previously stated, acute liver failure can be diagnosed in a previously well but undiagnosed patient with Wilson’s disease, hepatitis B or autoimmune hepatitis where compensated cirrhosis may have been present, provided the disease has been recognised for less than 26 weeks.

Acute liver failure results in a systemic inflammatory response and has multi-systemic manifestations. Table 9.7 illustrates the clinical features and therapeutic options according to the organ system affected.

There are a number of points in the general management of liver failure patients that are relevant irrespective of the underlying aetiology.

Patients with acute liver failure should be managed according to standard best practices on the intensive care unit.

Since serial evaluation of laboratory coagulation variables, like INR and PT, are important elements of prognostic evaluation, the administration of coagulation factors should be avoided except if the patient is bleeding or prior to invasive procedures [12].

Intubation of the trachea is recommended for patients with severe metabolic disturbance refractory to adequate fluid resuscitation and in patients with grade III and IV encephalopathy, typically for airway protection and carbon dioxide control. Propofol is the sedating agent of choice. Use of suxamethonium as part of a ‘rapid sequence induction’ is debatable in view of effects on intracranial pressure (ICP). Drugs with hepatic metabolism are avoided in favour of those with extra-hepatic metabolism (e.g. Atracurium – Hoffman degradation, Remifentanil – plasma esterase). The routine use of infusions of neuromuscular blockade is not recommended due to associations with neuromyopathy and ventilator associated pneumonia even in cases of raised ICP. Tracheal intubation and artificial ventilation for encephalopathy requires standard neuroprotection strategies to be employed to counter lability in ICP.

Respiratory care is based on the use of lung protective strategies with low tidal volumes, judicious use of PEEP and avoiding high peak airway pressures. Intrapulmonary shunts are uncommon in acute liver failure in contrast to chronic liver disease where the hepato-pulmonary syndrome can occur. Physiotherapy and respiratory toileting should be undertaken with caution due to the risk of bleeding and increasing ICP.

Cardiovascular effects of hypotension due to a low systemic vascular resistance in association with a high cardiac output are to be expected, often worsened by concurrent infections and hypovolaemia. Patients presenting with acute liver failure typically require fluid resuscitation to correct hypovolaemia and resulting hypoperfusion. It is not atypical to see a profound metabolic derangement fulfilling the King’s criteria upon admission to ICU to normalise with adequate resuscitation. The choice of fluid is dependent on clinical preference, with the caveat that lactate containing fluids and 5 % dextrose should be avoided. The use of 5 % dextrose in acute liver failure risks hyponatraemia, cerebral oedema and worsening intracranial hypertension [14]. The livers’ inability to efficiently clear lactate and the likelihood of a type 1 lactic acidosis precludes the use of the former.