Article Text

Download PDFPDF

Encephalopathy in children: an approach to assessment and management
  1. Emily Davies1,
  2. Daniel J Connolly2,
  3. Santosh R Mordekar1
  1. 1Department of Paediatric Neurology, Sheffield Children's Hospital, NHS Foundation Trust, Sheffield, UK
  2. 2Department of Paediatric Neuroradiology, Sheffield Children's Hospital, NHS Foundation Trust, Sheffield, UK
  1. Correspondence to Santosh R Mordekar, Department of Paediatric Neurology, Sheffield Children's Hospital, NHS Foundation Trust, Ryegate Children's Centre, Tapton Crescent Road, Sheffield, S10 5DD, UK; santosh.mordekar{at}


Childhood encephalopathy is an uncommon but significant paediatric presentation and is associated with significant mortality and long-term morbidity in survivors. By definition it is a somewhat non-specific presentation with a wide differential diagnosis and long list of possible investigations. Choice of investigation, including neuroimaging modality, can be a daunting prospect for the clinician faced with the encephalopathic child and it is important to select appropriately for a high diagnostic yield. Initial management centres on good emergency care irrespective of cause. More specific therapies however vary enormously, and appropriate treatment is important and influences outcome. Evidence exists for mana©gement of many of the individual conditions causing encephalopathy. This review aims to outline a clinical approach to selecting investigations to identify a specific cause and provides an overview of the treatment for the commoner causes of encephalopathy that a general paediatrician may reasonably expect to be faced with.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


The encephalopathic child is a paediatric emergency and presents a considerable challenge. The child may present acutely to general paediatricians and emergency medicine specialists but certain specialities (eg, neurology, hepatology) may see subacute presentations. Encephalopathy is not a diagnosis but a descriptive term for a syndrome of global brain dysfunction with a wide differential. Morbidity and mortality in these children remains high but prompt recognition of the encephalopathic state and appropriate and timely investigations will help identify treatable causes and can minimise further neurological impairment. In this overview we present a structured approach to assessing, investigating and managing such children. We do not include neonatal encephalopathy in this review.

There is inevitably some overlap with the child with encephalitis, on which a helpful review has recently been published,1 but it is important to recognise encephalopathy as a different clinical entity to encephalitis.


A child can present with encephalopathy at any age. There may be an acute or more insidious onset. The child may have previously been neurologically intact, or they may present with an encephalopathic crisis on the background of longstanding neurological impairment. The defining feature is an altered mental state and the parent/carer's assessment of such children is crucial, particularly for the paediatrician seeing the child for the first time.

Depending on the age of the child there may be a change in personality or behaviour, deterioration in cognitive functioning, developmental regression/stasis, reduction in conscious level and specific localising features such as seizures, ataxia, tremor or other focal motor signs. There may also be systemic features such as fever, vomiting, lethargy, loss of appetite and headache.

Causes of encephalopathy

The differential is wide and best considered systematically. Box 1 lists the causes we feel should be considered in the general paediatric acute setting.

Box 1 Causes of encephalopathy

Infection and parainfectious

  • Meningitis

  • Encephalitis

  • Intracerebral abscess

  • Systemic infection leading to altered mental state

  • Acute disseminated encephalomyelitis


  • N-methyl-D-aspartate receptor antibody encephalitis

  • Voltage gated potassium antibody encephalitis

  • Hashimoto's encephalopathy


  • Accidental

  • Non-accidental

Seizure related

  • Non-convulsive status

  • Post ictal

  • Epileptic encephalopathy


  • Drugs (therapeutic and recreational)

  • Heavy metal poisoning

  • Carbon monoxide


  • Uraemia

  • Hyperammonaemia

  • Hyper/hypoglycaemia

  • Lactic acidosis

  • Liver failure

  • Leucoencephaloapathies, for example, mitochondrial disorders, organic acidopathies


  • Renal disease

  • Cardiac disease


  • Neonatal hypoxic-ischaemic encephalopathy

  • Near drowning

  • Near miss sudden infant death syndrome

  • Following prolonged resuscitation/cardiorespiratory arrest

  • Vascular (stroke, venous thrombosis, migraine)


  • Traumatic

  • Spontaneous (eg, arteriovascular malformation, coagulopathy)


  • Primary brain tumour

  • Metastatic disease

Clinical assessment

In any child with reduced conscious level the ABC approach (airway, breathing and circulation) is the appropriate first line assessment. Once airway protection, respiratory adequacy and circulatory sufficiency are established then more detailed assessment of the neurological status can be performed. It is important to obtain a history from someone who knows the child well. Salient points are the child's pre-existing neurological status, development and medical history, as well as birth history. The timing and nature of the deterioration, any recent febrile illnesses, symptoms of headache, vomiting and/or diarrhoea and a history of seizures should be sought. A history of travel or contact with animals or insects is also important. Questions to assess the likelihood of drug/toxin ingestion, including prescribed drugs/chemotherapeutic agents are necessary and need broaching sensitively. A history of trauma is normally obvious but it is important to ask specifically about any minor, seemingly insignificant head or neck trauma (carotid artery dissection can be a cause of stroke in children). A family history of neurological, metabolic, vascular or haematological disease is important, as is a history of early childhood or infant deaths. Parental consanguinity is also relevant. The social history may or may not give clues when considering the possibility of non-accidental injury and the housing situation is important when thinking about lead poisoning.

Formal examination of the encephalopathic child is difficult. Opportunistic examination and observation is a key. Watching if and how the child communicates, whether they are visually alert, and their pattern and quality of movement can tell you more than a formal examination in a combative child. Assessment of orientation and memory can be made with simple questions. A more detailed neurological examination may give vital clues as to the underlying diagnosis and the cardiovascular, respiratory and gastrointestinal examinations should not be overlooked.

A set of basic observations (temperature, heart rate, respiratory rate, blood pressure and saturations) should be carried out as soon as possible. A bedside blood glucose measurement is crucial.


Investigations should be undertaken to identify a cause that is treatable, but identifying a cause that is not treatable is also important. Families are keen to have an explanation for their child's condition, this allows a more informed prognosis to be given and stops further unnecessary investigations.

There is no one list of investigations for the encephalopathic child as these should be tailored to the history and investigation. Recent evidence-based guidelines do, however, state a list of investigations (box 2)2 that should be undertaken at presentation in children with reduced conscious level to identify an immediately treatable cause. Two further stages of investigation are suggested if initial tests are unrevealing. A review paper from 2001 gives an extensive list of investigations for children with non-traumatic coma3 and the recent review paper on encephalitis in children details investigations when this is the cause of encephalopathy.1 Diagnostic imaging plays a pivotal role in establishing a diagnosis in the encephalopathic child.4

Box 2 Initial investigations in a child with encephalopathy

Capillary glucose and laboratory glucose

Blood gas

Urea and electrolytes

Liver function tests


Full blood count and film

Blood cultures

Plasma to save (1–2 ml)

Serum to save (1–2 ml)

Urine dipstick

Urine to save (10 ml)

Here, we propose a clinical approach that incorporates these second line investigations and neuroimaging strategies, by planning investigations based on a differential diagnosis. Below we set out investigations that are indicated for each of the causative categories listed in box 1.


Inflammatory markers would be expected to be raised in meningitis and cerebral abscesses. One should not be reassured by unremarkable values when there is good clinical suspicion of central nervous system infection. Blood cultures and viral serology (including mycoplasma and Borrelia) are useful, as is blood PCR for suspected viruses. Viral throat, urine and stool samples can also help identify a precipitant. In demyelinating conditions, it would also be important to send blood oligoclonal bands (paired with cerebrospinal fluid (CSF)). Aquaporin 4 antibodies are becoming increasingly useful but are not as sensitive for neuromyelitis optica as in adults5. Their presence however can imply a risk of relapse.6 Lumbar puncture (LP) is the gold standard for diagnosis and should be performed as soon as is safe.7,,10 Microscopy, culture and sensitivity, protein and glucose should be sent in addition to viral samples. The list of viruses causing encephalitis/acute demyelinating encephalomyelitis (ADEM) is long and pointers in the history, and results from throat/stool/urine samples will allow targeted testing of the CSF.1 In certain children (eg, patients who are immunocompromised) fungal causes will need to be looked for.

EEG can also be helpful in diagnosing encephalitis. There is usually generalised or focal slowing. The classically described focal temporal changes and periodic lateralising epileptiform discharges are no longer thought to be pathognomonic of herpes simplex virus encephalitis.11

Neuroimaging is not necessary in straightforward meningitis at presentation but may be required if the child does not improve as expected. CT pre/post contrast can diagnose complications of meningitis (eg, hydrocephalus, abscess, infarction) (figure 1) but in suspected encephalitis and demyelinating conditions MRI is more sensitive. (figure 2A,B) MRI is also superior when parainfectious cerebellitis is suspected.

Figure 1

CT of the head with contrast. Hydrocephalus, marked periventricular low attenuation with effacement of surface sulci consistent with ventriculitis in a child with meningitis.

Figure 2

(A) Brain MRI (coronal T2 fluid attenuation inversion recovery (FLAIR)) showing bilateral multiple foci signal chance consistent with acute demyelinating encephalomyelitis (ADEM). (B) Brain MRI (axial fast spin echo) showing bilateral multiple foci T2 signal change with enhancement consistent with ADEM.


It has become clear that autoantibodies are associated with encephalitis, particularly limbic encephalitis,12,,15 with promising treatment options with immunomodulatory therapies. N-methyl-D-aspartate (NMDA) receptor antibodies, voltage gated potassium channel antibodies and glutamic acid decarboxylase antibodies have all been implicated in children with limbic encephalitis and should be considered in children presenting with memory disturbance, affective change and seizures or movement disorders. Testing for these antibodies can be performed on blood. NMDA receptor antibody encephalitis is associated with ovarian teratomas and any girl in whom this is a suspected diagnosis should have a pelvic ultrasound. MRI of the brain may show changes in the limbic region and it may be this that first prompts further investigation with antibody testing.

Hashimoto's encephalopathy is a rare cause of encephalopathy in children but needs to be considered and thyroid function tests as well as thyroid autoantibodies are simple tests that should be performed in unexplained encephalopathy.16 (figure 3A,B).

Figure 3

(A) Brain MRI (coronal T2 fluid attenuation inversion recovery (FLAIR)) showing high T2 signal in the hippocampi bilaterally in a child with limbic encephalitis. (B) Brain MRI (coronal T2 HiRes) showing high signal in the hippocampi bilaterally in the same child as figure 3A.


Neuroimaging is the pivotal investigation. Non-contrast CT is often sufficient, identifying bleeds, cerebral contusion and in some case showing hypoxic-ischaemic damage. In non-accidental injury there are characteristically subdural bleeds of differing appearances on CT and MR, fractures and evidence of hypoxic-ischaemic damage. (figure 4) Bleeding for which there is no obvious causes can also raise the possibility of metabolic disease such as glutaric aciduria type 1 and Menke's. Organic and amino acids, copper and caeruoplasmin and in some cases a skin biopsy may be necessary. Basic investigations such as full blood count, clotting and liver function tests (LFTs) are also needed and crossmatching blood should be considered. MRI is more sensitive at detecting subtle hypoxic-ischaemic changes and may be useful in determining the extent of irreversible brain injury. Rarely, infection can produce subdural fluid collection with secondary haemorrhage and parenchymal injury. Inflammatory markers, blood and CSF analysis may be required to exclude infection for medicolegal purposes. (figure 5A,B).

Figure 4

CT of the head showing subdural blood with a bright cerebellum in a child with non-accidental head injury.

Figure 5

(A) CT of the head with contrast showing swollen brain with loss of grey/white matter differentiation. (B) Brain MRI (axial T2 fast spin echo) showing diffuse high signal in the basal ganglia in the same patient as figure 5A.

Seizure related

Seizure activity can be a primary cause of encephalopathy or can be a consequence of another diagnosis and is important to recognise promptly. Several studies have linked outcome to seizure activity in patients with encephalopathy of various causes.17,,19

No investigation may be necessary in a child with known epilepsy who is encephalopathic following a prolonged seizure/seizure cluster but recovers as expected. EEG however can help diagnose seizure activity when it is difficult clinically (eg, non-convulsive status (NCS), electrical status epilepticus in sleep). It can also help classify epileptic encephalopathies. In newly presenting epileptic encephalopathies there are a number of other investigations indicated and seeking the advice of a paediatric neurologist is advised. It is beyond the scope of this review to detail investigation of epileptic encephalopathy. Having said that, paired fasting blood and CSF studies for lactate and glucose can be useful and given that LP may be performed before consultation with a paediatric neurologist, ensuring these samples are taken with a sample to save may avoid the need for repeat LP. Taking blood for karyotyping and DNA for storage is also appropriate.


The history will hopefully give some clues but urine for toxicology should be obtained as soon as possible. Blood should also be taken for paracetamol and salicylate levels, and a heavy metal screen where poisoning could be a possibility. Drug levels of anticonvulsants in known epileptic children should be tested in situations of acute deterioration in conscious level, as toxicity may be the cause.

An MRI scan can be useful and show particular patterns of brain injury with particular drugs. Discussion with a neuroradiologist is advised.


This is a broad category and children with suspected metabolic disease are at risk of being overinvestigated while missing the one crucial investigation necessary for diagnosis. Blood, urine and CSF results are important in many cases and ammonia, lactate, venous blood gas, plasma amino acid, urine organic and amino acid and sometimes CSF lactate, glucose and amino acid tests should be done acutely. Beyond this, discussion with a metabolic specialist or neurologist is advised.

MRI is also an important diagnostic tool in metabolic disease, and again it is outside the scope of this review to cover this in detail. When metabolic disease is suspected it is important to consider the use of spectroscopy. Metabolic causes of encephalopathy can be suspected when there are symmetrical changes in white or grey matter or both on MRI and magnetic resonance spectroscopy shows a lactate peak.

Endocrine causes should also be addressed in this category and blood glucose, blood gas and urine dipstick tests will identify diabetic ketoacidosis and blood and urine electrolytes will point to adrenal insufficiency.


This diagnosis is largely made on initial observations. The underlying cause (eg, haemolytic uraemic syndrome, Henoch–Scohlein purpura) should be investigated by renal function, urinalysis, renal ultrasound cardiac echo and endocrine investigations (eg, cortisol, renin, aldosterone, catecholamines).


Hypoxic-ischaemic damage secondary to trauma, prolonged resuscitation or multiorgan failure may be suggested on initial CT by loss of grey/white matter differentiation. MRI is more sensitive at detecting damage and can show changes within hours of injury, although there is often an optimum timing for scanning. If the child survives, repeat imaging when myelination is complete (after 2 years of age) is recommended to look at the permanent changes once acute oedema and haemorrhage have resolved. In such children LFTs, urea and electrolyte, clotting and urine dipstick tests are essential to look for evidence of multiorgan damage.

In children with stroke the first investigation is usually CT, which reliably detects haemorrhagic causes of stroke. Studies have shown however that CT alone will miss a large number of acute ischaemic stroke cases20 (CT detection in 10% compared to MRI detection in 46%). Brain imaging is recommended within 24 h of onset of symptoms suggestive of ischaemic stroke, along with imaging of the cervicocranial vascular tree, most conveniently performed with MRI with magnetic resonance angiography (MRA)20 (figure 6). MRI is also the follow-up imaging modality of choice.21 A full blood count and clotting investigation are needed and the erythrocyte sedimentation rate and an autoimmune screen are helpful in looking for vasculitic causes. Homocysteine, cholesterol and triglyceride levels should be checked and prothrombotic states should also be looked for, in the family history and by checking protein C, protein S, factor V Leiden, anti-phospholipid and anti-cardioplipin antibodies. Discussion with haematologists is advised at this stage. An echocardiogram is also indicated where a cause for infarction is not clearly demonstrated.

Figure 6

MRI of the brain showing right middle cerebral artery infarct.


This is normally secondary to trauma but may be spontaneous in children with abnormal clotting/platelets or in children with arteriovenous malformations (AVM). Platelet count and clotting screen are essential. Specific factor assays can be discussed with a paediatric haematologist. In the first instance CT will show most haemorrhages but underlying AVM may need to be considered with MRI/MRA, CT angiography or catheter angiography. AVM are often not visible until the acute bleed has resolved or been neurosurgically managed22 and therefore delayed angiographic imaging may be advised.


Brain tumours may present with encephalopathy. Children with known brain tumours can also become encephalopathic as a result of raised intracranial pressure, tumour spread, or chemotherapeutic agents. Imaging is the first line investigation. CT will detect many tumours and in an emergency is useful at showing hydrocephalus, cerebral oedema and haemorrhage. Most paediatric brain tumours are in the midline and therefore midline shift, a marker of cerebral mass effect, may not be clearly evident. MRI may be useful for surgical planning or for smaller lesions or tumours in specific sites. Once imaging has suggested a tumour then more detailed neuro-oncological investigations and management need to be planned with specialist teams. Basic full blood count, clotting, group and save/crossmatch, full electrolyte investigations are certainly needed and pituitary function tests should be considered dependent on the site of the tumour.


Emergency management

In the first instance a child with altered mental state needs to be managed with good emergency care, irrespective of the cause. This undoubtedly improves outcome. Consideration needs to be given to use of a paediatric intensive care unit (PICU) care if there is a suggestion of an unstable airway or organ support is required.

Management of raised intracranial pressure is important and neuroprotective strategies include nursing with head in midline, 20° up and maintaining normocapnea.

Treatment of raised intracranial pressure can involve use of hyperosmolar substances. Evidence for use of mannitol in children is mainly extrapolated from adult studies and there is emerging but non-conclusive evidence that hypertonic saline is more effective than mannitol.23,,25 To date use of either in traumatic brain injury with life threatening intracranial pressure is supported by the literature as is use in the intensive care setting where intracranial monitoring is occurring. For other situations individual judgement is required.

The use of steroids in bacterial meningitis reduces hearing loss and neurological sequelae in Haemophilus influenzae meningitis but does not affect mortality26 and, if used, should be given with or within 4 h of the first dose of antibiotics.10 Dexamethasone can be helpful for cerebral oedema associated with intracranial mass lesions. Treatment of acutely reversible causes such as hypoglycaemia and electrolyte disturbances should be promptly recognised and treated, although care should be taken not to correct some electrolyte disturbances too rapidly. Management of the child with diabetic ketoacidosis is a prime example and there are national guidelines that should be followed.27 28 Seizures should be treated by following the Advanced Paediatric Life Support guideline for status epilepticus.29

Where metabolic disease is suspected emergency treatment is crucial to prevent further damage and the first step is stopping feeds and starting intravenous dextrose. If hyperammonaemia is present, sodium benzoate, sodium phenylbutyrate and arginine intravenously will help reduce plasma ammonia levels and discussion with the local dialysis unit/PICU is required.30

In many circumstances it is appropriate to initiate antimicrobial treatment acutely, particularly if there is any suggestion of fever or prodromal illness. Cefotaxime is the appropriate first line antibiotic, with the addition of aciclovir where there is a reasonable suspicion of herpes encephalitis. Mycoplasma is another treatable organism so addition of a macrolide should be considered.

Specific management

It is outside the scope of this article to detail the specific treatment of each individual cause of encephalopathy but table 3 lists some specific treatments and supportive care. Some of these can be started soon after presentation but some should be reserved until discussion with a specialist.

Table 1

Specific management of encephalopathy


Following acute management the rehabilitation phase is also an extremely important part of management and should not be overlooked. Involvement of the hospital multidisciplinary team should be instigated as soon as possible. Consideration should be given to education and psychological input for the child as well as family support as there may be a huge change to individual and family circumstances.


The encephalopathic child is a paediatric emergency that can present a diagnostic conundrum, and carries with it significant morbidity and mortality. Good management can limit the extent of longer-term damage. It is important that all clinicians who may be involved with children with altered mental state use a systematic approach that allows them to form a differential diagnosis, instigate useful and appropriate investigations and initiate good emergency care. Clinicians also need to consider specific treatments for individual diagnoses. This review has described such an approach, starting with the patient's presentation rather than a specific diagnosis and pulling together biochemical, metabolic, microbiological, neuroradiological and electrophysiological investigations, based on recent available evidence. It therefore provides a review of current clinical practice.


View Abstract


  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.