Coma

Coma is a common presentation to the Emergency Department. However there are many different causes of coma and it is essential that the ED physician is able to quickly and reliably differentiate between the major differential diagnoses.

This article will review the following aspects of the patient in coma:-

• the key features relevant to the history of acute coma
• the common, traumatic, metabolic and infective aetiologies of coma
• the important causes and distinguishing features of coma in travellers
• the investigation of the patient presenting in acute coma
• the generic aspects of coma management
• the key differences in the management of coma according to the cause

The word coma is derived from the Greek k ma meaning deep sleep. Various definitions exist. The Merriam-Webster dictionary defines it as a state of profound unconsciousness caused by disease, injury, or poison The key point common to all definitions is that coma is a state of unconsciousness in which the patient is unable to react to people, or events within their environment. Patients can have different levels of unconsciousness and unresponsiveness depending upon how much or how little of the brain is functioning and the intensity of the stimulus.

It is important to note that there is no single examination finding, score or test that distinguishes between all patients in coma and those who are not.

Basic Science and Pathophysiology

Consciousness requires two key components of the Central Nervous System to be functioning, the Reticular Activating System (RAS) and at least one cerebral hemisphere. For a person to be unconscious, either the RAS has ceased functioning, or both cerebral hemispheres have.

There are two main causes of failure of the reticular activating system

• Brain stem stroke: either ischemic or hemorrhagic
• A pre-death event: increased swelling in the brain pushes down on the brain stem and causes it to fail.

For both cerebral hemispheres to fail requires

• failure of an adequate blood supply
• inadequate substrate for normal metabolism e.g oxygen or glucose
• direct or indirect trauma to the cerebrum
• exposure of the brain to a toxic insult including
  • infection,
  • toxic metabolites
  • exogenous poisons

Coma occurs when either there is failure of the Reticular Activating System or failure of both cerebral hemispheres

Learning Bite

In consequence acute cerebral hemisphere stroke does not cause coma but brain stem haemorrhage or infarction may do.

Initial assessment of the patient should focus on ensuring the patient is neither hypoxic nor hypotensive, both situations can cause coma. Where coma is due to other causes both hypoxia and hypotension can worsen the outcome.

Learning Bite

Hypoxia and Hypotension must be identified and treated in all patients in coma. Failure to do so increases morbidity and mortality.

General examination should focus on the function of specific organs and stigmata of certain toxic causes of coma Fig. 1

Evidence of injury

• Scalp laceration
• Scalp haematoma
• Signs of basal skull fracture
• Bitten tongue suggesting seizure

Evidence of fever

• Bacterial meningitis
• Encephalitis
• Brain injury (Pontine or hypothalamic)
• Hypothermia may indicate sepsis, exposure to cold, intoxication or hypothyroidism

Evidence of organ failure

• hypoxia and hypercapnia associated with hypoventilation
• Hypotension associated with cardiac failure
• examine the breath for the characteristic odours of
• ketones (pancreatic failure)
• uraemia (renal failure)
• fetor hepaticus (liver failure)
• garlic (Organophosphate /insecticide poisoning)

Evidence of toxic ingestion / inhalation

• needle marks associated with i.v. drug abuse,
• bullae associated with barbiturate overdose
• cherry red appearance of carbon monoxide poisoning
• Dry skin (especially in normally moist areas) suggests Tricyclic overdose and other anticholinergic agents
• Profuse sweating suggests hypoglycaemia or organophophate poisoning

Following the general examination of the patient attention can be directed at the detailed neurological assessment.
The depth of coma is most reliably determined using the Glasgow Coma Score (GCS). This is the most valid measure of coma (Table 1)
Patients in the deepest level of coma:

• do not respond with any body movement to pain,
• do not have any speech, and
• do not open their eyes.

Patient with a GCS of less than 10 who are not expected to recover quickly should be assessed with a view to intubation to protect the airway and optimise respiratory function. Patients who are likely to recover quickly e.g post ictal patients do not ordinarily require intubation.

Learning Bite

Patients with a GCS of less than 10 should be considered for intubation.

It should be noted that the Glasgow Coma Score is not of itself a prognostic tool.
At the same time as the GCS is determined it is essential that the pupillary responses to light are recorded as well ocular abnormalities and any posturing of the limbs. In cases where head injury is not suspected the presence or absence of neck stiffness should be determined.

1. Evidence of Head Injury

In patients with head injury the priority is to maintain adequate cerebral perfusion pressures and cerebral oxygenation. Hypotension and hypoxia, which may arise as a result of injuries other than to the head must be aggressively treated. Thereafter the patient should undergo CT scanning without delay, in order to assess whether there is a neuro-surgically remediable cause such as an extradural haematoma.

2. No head injury but focal neurological signs

Most of these patients will have suffered an acute cerebrovascular event. However it is important to remember that most patients who have had a cerebrovascular event do not present in coma. Another important cause of such a presentation, is acute hypoglycaemia. As such a BM test and laboratory blood sugar analysis should be rapidly obtained.

Learning Bite

Hypoglycaemia is a reversible cause of coma. Prompt recognition and treatment reduces morbidity and mortality

3. No head injury, no neurological signs, infection probable

The differential diagnosis for such patients is meningitis or cerebral malaria. The priority is to commence antibiotic therapy appropriate for acute meningitis whilst undertaking appropriate tests for malaria (in patients who could conceivably have contracted the condition) and arrange urgent CNS imaging.

4. No head injury, no neurological signs, infection unlikely

These patients can be regarded as having been poisoned. The toxin may be exogenous e.g. Carbon monoxide, Tricyclic antidepressants, Narcotics etc. or endogenous arising as a result of organ failure e.g. Diabetic Ketacidosis, Myxoedema coma, Respiratory failure. As such the most useful tests are arterial blood gas analysis, serum anion gap and serum electrolytes

Learning Bite

Metabolic and Toxic causes of coma are usually identified by routine blood tests including arterial blood gas analysis

Alcohol can cause coma in a number of direct and indirect ways. Acute intoxication will cause coma; the amount required depending on the patients tolerance to alcohol. Acute withdrawal can lead to delirium and seizures. Following a seizure the patient may be comatose. Head injuries sustained as a result of intoxication can cause coma and the chronic alcoholic is particularly susceptible as long term alcohol abuse impairs clotting and causes cerebral atrophy with a consequent increased risk of sub-dural haemorrhage.

Patients with arms flexed and legs extended are said to have a decorticate posture. This indicates injury above the midbrain. Those with legs and arms extended and adducted are said to have a decerebrate posture. This indicates a brainstem lesion. Both are poor prognostic signs

Following the above focused examination coma patients can be assigned to one of four groups which then guides further assessment, investigation and management

1. Evidence of Head Injury
2. No head injury but focal neurological signs
3. No head injury, no focal neurological signs, infection probable (history, temp, WBC)
4. No head injury, no focal neurological signs, infection unlikely (history, temp, WBC)

From each of the above groups it is possible to determine the most likely causes and commence an appropriate strategy of investigation and treatment In many patients investigation and treatment should proceed simultaneously.

Examples of CT scan images associated with various causes of coma

A = Intracerebral abscess. This scan shows a large intracerebral abscess with an air fluid level. There is associated oedema. There is no significant midline shift or evidence of compression of the left hemisphere. Planned surgical drainage is indicated

B = Large cerebellar haemorrhage. There is a large cerebellar haematoma. Patients with cerebellar haematomas > 3cm in size have better outcomes if the haematoma is surgically drained. Intervention before the onset of hydrocephalus and/or loss of brainstem reflexes improves the prognosis

C = Small extradural haematoma and frontal lobe neoplasm. Although there is a left temporal extradural haematoma if the patient is fully conscious it may not be sufficiently large to warrant evacuation. The confounding factor is the presence of the right frontal lesion which is either a primary cerebral neoplasm or solitary metastasis.

D = Large metastasis.. This scan shows a neoplastic lesion in the right temporal lobe. Thi may be a primary or secondary lesion and surgical intervention may or may not be appropriate

E = Intracerebral bleed. This scan shows an intracerebral haemorrhage. There is no evidence that surgical evacuation of such heamorrhages improves outcome.

The necessity of treating hypoxia and hypotension to prevent further neurological damage has already been stated. Administer oxygen to all coma patients as required to maintain normoxia, and ensure that ventilation rate and depth are adequate. Urgent blood gas analysis will allow determination of the following:-

• Hypoxia
• Hypercapnia
• Acidosis respiratory or metabolic

All of which can cause coma or occur as a result of coma.
Further analysis of those patients with a metabolic acidosis will enable calculation of the anion gap which will indicate probable causes for the coma .
A serum lactate is helpful in establishing the presence and degree of tissue hypo-perfusion and is particularly useful as a marker in sepsis
Patients suspected of taking overdoses of benzodiazepines or opiates should be considered for reversal of these agents using flumazenil or naloxone respectively, bearing in mind the potential to induce seizures or an acute withdrawal state it may be better simply to support the airway and ventilation until these agents wear off. Guidance on this matter, and on the management of other poisonings, can be found via TOXBASE.

Wernickes encephalopathy is a rare cause of coma, however indiscriminate infusion of glucose in (thiamine deficient) alcoholics can precipitate further acute neurological damage. In consequence all malnourished and alcoholic patients in coma should receive 100 mg thiamine slowly over 5 minutes prior to the administration of glucose.

Learning Bite

In cases of non-traumatic coma the attending doctor should consider specific treatment with naloxone, flumazenil, thiamine and glucose

Learning Bite

Surgical evacuation of cerebellar haematomas is proven to improve outcome; surgical evacuation of intracerebral haematomas is not

Learning Bite

Coma following cardiac arrest is not of itself an indication to withdraw therapy

Learning Bite

All patients who present in coma with pyrexia should receive broad spectrum antibiotic therapy urgently

The prognosis of patients in coma is related principally to the direct cause of the coma. The 3 main CNS causes of death are:-

• Raised intracranial pressure
• Herniation
• Encephalitis / Meningitis

Two other prognostic determinants have been identified including

• Duration of coma
• Presence/ Absence of brainstem reflexes

Cause of coma

For patients and their relatives and friends the outcome to be wished for is at least a moderate to good recovery. Overall only 15 % of patients in non-traumatic coma lasting more than 6 hrs will make such a recovery. This figure rises to 35% in patients whose coma is due to infection, organ failure and biochemical derangement and falls to 11% in patients whose coma is due to hypoxic injury. The outcome following a cerebrovascular insult is even worse with only 7% likely to achieve a moderate to good recovery.
One particularly unfortunate group of patients are those who have suffered an hypoxic brain injury. 20% will enter a persistent vegetative state as hypoxia is particularly likely to cause irreversible bi-hemispheric damage with relative sparing of the brainstem.
The outcome of coma following trauma is significantly better. Of those with a GCS of 8 or less 35 % had a good recovery at one month.

Specific research has been done, and guidelines produced, for prognostication amongst comatose survivors of cardiac arrest. These can be found in the Resuscitation Council (UK) guidelines¹¹.

Duration of coma

Those patients not opening their eyes in response to pain after 6hrs of coma have only a 10% chance of making a moderate to good recovery compared to 20% of those whose eyes do open to pain. In those who remain in coma for 1 week only 3% will make a good recovery

Corneal reflexes and pupillary responses

In a prospective study by Levey et al of 500 patients in coma death occurred in all those with absent corneal reflexes or absent pupillary response to light at 24 hrs.

Key learning point
The Prognosis of a patient in coma is determined primarily by the cause of the coma.

The Differential diagnosis of any patient presenting with coma of unknown cause must include patients who have the locked in syndrome and those who are malingering.

1. Locked in Syndrome. In this syndrome, an acute pontine lesion causes paralysis of all 4 limbs and the head and neck including the facial muscles. The patient is however conscious (the reticular activating formation is intact) and they are able to blink and move their eyes because the oculomotor pathways are spared. It is essential therefore, before a painful stimulus is applied, to ask the patient to blink and move their eyes up and down.

2. Malingering. Patients who are malingering may or may not be initially convincing. Often the history raises some suspicion, similarly the neurological findings are contradictory or inconsistent. It is however a diagnosis initially of exclusion. Normally the condition lasts only a day or so; there is a charachteristic sudden awakening and associated retrograde amnesia.

3. Stroke. Most patient s who have had a CVE do not present in coma. Other causes should be sought.

1. Acute cerebral hemisphere CVE does not cause coma but brain stem haemorrhage or infarction may do. Grade A recommendation. Evidence level 1c
2. Hypoxia and Hypotension must be identified and treated in all patients in coma. Failure to do so increases morbidity and mortality Grade B recommendation. Evidence level 2a
3. Patients with a GCS of less than 10 should be considered for intubation Grade D recommendation. Evidence level 5
4. Hypoglycaemia is a reversible cause of coma. Prompt recognition and treatment reduces morbidity and mortality Grade B recommendation. Evidence level 2b
5. Surgical evacuation of cerebellar haematomas is proven to improve outcome; surgical evacuation of intracerebral haematomas is not. Grade B recommendation. Evidence level 2a
6. Metabolic and Toxic causes of coma are usually identified by routine blood tests including arterial blood gas analysis Grade D recommendation. Evidence level 5
7. Adrenal insufficiency may mimic sepsis and coexist with sepsis and other endocrine failure states. Grade D recommendation. Evidence level 5
8. In cases of non-traumatic coma the attending doctor should consider specific treatment with naloxone, flumazenil, thiamine and glucose Grade D recommendation. Evidence level 5
9. All patients who present in coma with pyrexia should receive broad spectrum antibiotics Grade D recommendation. Evidence level 5
10. Coma following cardiac arrest is not of itself an indication to withdraw therapy Grade B recommendation. Evidence level 2a
11. The Prognosis of a patient in coma is determined primarily by the cause of the coma. Grade B recommendation. Evidence level 2a

1. Dubey A, et al: Volume of extradural hematoma and prognosis. Neurology India December 2004 Vol 52 Issue 4
2. Baloh RW. Dizziness: neurologic emergencies. Neurol Clin 1998;2:305-321
3. St Louis EK, Wijdicks EFM, Li H et al: Predictors of poor outcome in patients with spontaneous cerebellar haematoma. Can J Neurol Sci 27:32 2000
4. Donauer E, Loew F, Faubert C et al: Prognostic features in the treatment of cerebellar haemorrhage. Acta Neurochir (Wien) 131:59, 1994
5. Bassetti C, Bromio F, Mathis J et al. Early prognosis in coma after cardiac arrest: a prospective clinical, electrophysiological and biochemical study of 60 patients. J Neurol Neurosurg Psychiatry 996;61:610615.
6. Mendelow AD, Gregson BA, Fernandes HM, et al. Early surgery versus initial conservative treatment in patients with spontaneous supratentorial intracerebral haematomas in the International Surgical Trial in Intracerebral Haemorrhage (STICH): a randomised trial. Lancet. Jan 29-Feb 4 2005;365(9457):387-97.
7. Friedrich J. O., Sheth K. N., Mayer S. A., Brun N. C. Recombinant Activated Factor VII for Acute Intracerebral Hemorrhage N Engl J Med 2005; 352:2133-2134, May 19, 2005.
8. Pal J, Brown R, Fleiszer D. The value of the Glasgow Coma Scale and Injury Severity Score: predicting outcome in multiple trauma patients with head injury.J Trauma. 1989 Jun;29(6):746-8.
9. Levey et al
10. Wijdicks EFM. Catastrophic neurologic disorders. 2nd edition. Published by OUP 2004.
11. Resuscitation Council UK, 2015, Guidelines: Post-resuscitation care.