Clinical testing for neurological determination of death

Clinical testing for neurolifgcal determination of death is a Level 1 topic from Section 2.1.15 in the CICM Syllabus for the Second Part Examination (2nd ed), where all the topics are Level 1 topics.

It is therefore nor surprising as to how frequently this comes up in the SAQs and vivas. The most "examinable" aspects are the preconditions to testing, the precise sequence of testing, which cranial nerves are involved, and the expected findings of the apnoea test.

The ANZIC Statement on Death and Organ Donation is the main resource for this. It has been revised many times and trainees reading back to questions from the distant past should not be surprised to see comments or answers which do not seem relevant to the modern version of the Statement. Throughout the process, over many years, the one thing that has remained consistent is the college expectation of near-verbatim recall of this document, and a thorough familiarity with all of the specific criteria. This should cut down the cognitive load of making summaries and agonising over whether something is important enough to include or not: in this specific case everything is important and the trainee should take a detailed knowledge of this statement into the exam. One possible caveat to this is the gradual transition of focus away from memorised branstem reflex testing routines (as they are documented on all  the standard forms anyway) and towards things like preconditions and the validity of radiological tests. 

Previous SAQs on this subject include the following:

Definition of brain death

  • Unresponsive coma
  • Absence of brain stem reflexes
  • Absence of respiratory centre functions
  • Clinical setting which suggests that these findings are IRREVERSIBLE.

OR:

  • The demonstrated absence of intracranial blood flow

Preconditions for clinical brain death testing

The below preconditions are taken directly from the Statement.

  • Evidence of sufficient intracranial pathology to deteriorate to loss of all brain function
  • AND all of the following:
    • Normothermia ≥35 C
    • Normotension: SBP ≥90 or MAP ≥60
    • Exclusion of effects of sedative medications
    • Absence of severe electrolyte, metabolic or endocrine disturbances:
      • BSL <3 mmol/L or >25 mmol/L
      • Na+ <125 mmol/L or >160 mmol/L
      • PO4 <0.5 mmol/L
      • Mg2+ <0.5 mmol/L
      • urea >40 mmol/L
      • Untreated severe hypothyroidism
      • Untreated severe hypoadrenalism
    • Absence of acute liver failure or decompensated chronic liver disease.
    • Absence of neuromuscular-blocking drugs
    • At least one intact ear and one intact eye
    •  Enough spinal cord function to assess the motor response in the facial nerve (VII) to painful stimulus in the upper limbs and to assess the motor response in the upper limbs to painful stimulus in the trigeminal (V) sensory region.
    • Ability to perform apnoea testing

Four hours of observation during which the patient remains in unresponsive coma

  • GCS of 3
  • Unresponsive pupils
  • Absent cough reflex
  • No spontaneous breathing efforts

Preconditions for apnoea testing

  • Absence of concomitant high cervical cord injury
  • Normoxia; satisfactory gas exchange
  • Haemodynamic stability
  • Absent brain stem reflexes

Timing of brain death testing

  • The clinical testing takes place at least 24 hrs after a cardiac arrest.
    • Or, 24 hours after therapeutic hypothermia has been ceased (i.e. after the patient has rewarmed)
  • Prior to 24 hrs, an arrested patients’ brain death can be determined by demonstration of absent cerebral blood flow.
  • After 30 days of age, the same rules apply to children and to adults. In early neonates, younger than 30 days, 48 hours is the minimal period of observation (not 24).
  • In premature newborns (before 36 weeks) begin death testing cannot be performed with certainty.

The process of clinical brain death testing

Testing for unresponsive coma

  • Painful stimulus in cranial nerve distribution, eg. supraorbital nerve pressure
  • Painful stimulus in all 4 limbs, eg. nailbed pressure
  • There should be no response
    • There may be spinal reflexes; these will only be triggered by painful stimulus in the 4 limbs.
    • The spinal reflexes will NOT be triggered by painful stimulus in the cranial nerve distribution.

Brain Stem  Reflex Testing:

These are tested in sequence. All reflexes must be absent.

  • Pupil Light Reflex: CN II,  III
    • Pupil constricts in response to light. Cataract surgery is no contraindication
  • Corneal  Reflex: CN V,  VII
    • Eye blinks in response to the cornea being touched. The cornea, not the sclera.
  • Trigeminal Pain: CN V, VII
    • Painful stimulus over the supraorbital nerve. There shouldn’t be any grimacing.
  • Vestibulo-ocular reflex: CN III, IV, VI, VIII
    • Examine the ear: auditory canal must not be blocked.
    • Put the head at 30 degrees.
    • Put ice-cold water into the ear
    • Watch the eyes for 60 seconds: in brain death, they will remain midline.
      •  The “dolls eye test” is a sub-maximal stimulus of the same reflex
  • Gag reflex: CN IX, X
    • Poke the posterior pharynx, both sides.
  • Cough reflex: CN X
    • Stimulate the trachea with a soft suction catheter…this wont work in people with a high spinal cord injury, as the efferent limb is severed

Apnoea testing

Apnoea testing must be carried out only after the brainstem reflexes have been tested, and if any of them were found to be positive any further brain death testing cannot continue.

  • Preoxygenate with 100% FiO2 for 5 minutes, and then turn off the ventilator.
  • Continue supplying oxygen via a manual resuscitator with connected PEEP valve (they used to recommend a T-piece). Watch for absent breaths.
  • After 10 minutes, take an ABG to demonstrate that the CO2 is rising and pH is falling (PaCO2 rises by 3mmHg every minute of apnoea).
  • To qualify for  brain death, "apnoea must persist in the presence of an adequate stimulus to spontaneous ventilation, i.e. an arterial PaCO₂ >60 mmHg (8 kPa) and an arterial pH <7.30"

To get through this without making the patient desperately hypoxic, the college permit mild hypoventilation (to a CO2 of 45). If the patient does become hypoxic, a couple of mandatory breaths are permitted to restore oxygenation, before carrying on with the process. 

The ANZICS statement reports that there has never been a documented case of anybody who fulfilled the above brain death criteria, and the preconditions for brain death, who has ever recovered any brain function.

Findings compatible and incompatible with brain death

Observations compatible with brain death:

  • Spinal reflexes in response to noxious stimulus:
    • Extension-pronation movements of the upper limbs
    • Nonspecific flexion of the lower limbs
    • Undulating toe reflex
    • Lazarus sign
    • Deep tendon reflexes
    • Plantar responses (flexor or extensor)
    • Head turning
  • Sweating
  • Blishing
  • Tachycardia
  • Normal blood pressure in absence of vasopressors
  • Absence of diabetes insipidus

Observations incompatible with brain death:

  • Extensor posturing (decorticate)
  • Flexor posturing (decerebrate)
  • True extensor or flexor responses to painful stimuli
  • Seizures
  • Attempt at breathing, defined as any respiratory muscle activity that results in abdominal or chest excursions or activity of accessory respiratory muscles

For a more comprehensive overview, a good (ancient) article from the Acta Neurochirurgica describes what the authors have quaintly termed "Spinal Man", a species of human bereft of higher cortical function, which is a creature reliant purely on spinal reflexes.

Additionally, a more recent article discusses the various physiological responses to apnoea testing, including all the various cardiovascular derangements which occur.

Thus:

a) - a seizure - rules out brain death

b) - a positive caloric reflex - is a brainstem reflex which is still working, and it rules out brain death

c) - arm flexion to ipsilateral painful stimulus - could be a spinal reflex, and does not rule out brain death. 

d) - a Lazarus sign - does not rule out brain death

e) - a hypercapnea-associated catecholamine surge - can occur with zero cerebral input, and does not rule out brain death.

Jurisdictional differences

Of the pages in Deranged Physiology, for some reason, this page generates the greatest amount of geographical confusion from readers who remain unpersuaded by the earlier reference to the ANZICS statement and assume that the content is regionally specific to their own region. This page is of course specific for Australia and New Zealand, as this is the area where the author practices, having no knowledge beyond the purely theoretical when it comes to brain death diagnosis in other countrie.  As such, the guidance and values in this page will not be representative of everybody's local guidelines. For example, the UK FICM have their own processes which are slightly different and reported in kPa (the 2025 AoMRC Code requires a starting PaCO₂ of at least 5.3 kPa, at least five minutes of apnoea, and a rise in PaCO₂ of at least 2.7 kPa from baseline, with an end PaCO₂ of at least 8.0 kPa and arterial pH below 7.3).  Something entirely different happens in the US. 

References

Australian and New Zealand Intensive Care Society. “The Statement on Death and Organ Donation.” Edition 4.1, 2021.

McNair, N. L., and K. J. Meador. "The undulating toe flexion sign in brain death." Movement disorders 7.4 (1992): 345-347.

Jørgensen, E. O. "Spinal man after brain death." Acta neurochirurgica 28.4 (1973): 259-273.

Ropper, Allan H. "Unusual spontaneous movements in brain‐dead patients."Neurology 34.8 (1984): 1089-1089.

Heytens, Luc, et al. "Lazarus sign and extensor posturing in a brain-dead patient: case report." Journal of neurosurgery 71.3 (1989): 449-451.

Lang, C. J. G., and J. G. Heckmann. "Apnea testing for the diagnosis of brain death." Acta neurologica scandinavica 112.6 (2005): 358-369.