Pitfalls in the diagnosis of death using neurological criteria

In the UK, human death is defined as irreversible loss of the capacity to breathe combined with irreversible loss of the capacity for consciousness. Irreversible cessation of brainstem function produces this clinical state and equates with the death of the individual.¹ Previously called ‘brainstem death’, the preferred terminology is now ‘diagnosis of death using neurological criteria’ (DNC) or neurological determination of death, thereby moving away from an anatomically-based definition of death to one that is based on permanent loss of function. It is more than 50 years since the Harvard criteria for the DNC were published, yet there remains significant international variation.^2,3 Despite this variation, DNC universally requires the presence of an irreversible structural brain injury that can cause the condition followed by the completion of two essential steps:

• 1)Confirming that the patient meets certain preconditions that ensure the clinical state is not treatable or reversible.
• 2)Undertaking a series of clinical tests to confirm the absence of brainstem function.

Whilst the tests in step 2 demonstrate absence of function, its irreversibility is dependent on the assessment in step 1. Rare errors in diagnosis are invariably made during step 1.

Step 1

Who: Whilst guidance exists as to who is eligible to undertake the two steps, none exists as to how competence can be maintained or assessed.¹ Knowledge of the basis of the tests does not indicate practical competence.

Where: Whilst there is no requirement, DNC will almost always take place within the ICU. Given recent national guidance, testing outside of the ICU will be unusual and should be avoided.⁴

When: The patient must be deeply comatose and apnoeic as a result of an irreversible, untreatable brain injury of known aetiology. Caution is advised if the aetiology is restricted to the posterior fossa (when waiting longer or ancillary testing should be considered) or the causative lesion is of an inflammatory nature, as this may be associated with reversibility (e.g. a brain abscess). It is vital a structural cause is established as loss of brainstem reflexes is not restricted to DNC. Case reports of reversible loss of brainstem reflexes after overdose exist.^5,6 Core temperature should be >34°C, as profound hypothermia is also associated with the absence of brainstem reflexes.

Before testing, the patient must be stable physiologically. Other causes of apnoea and coma must be excluded, including extreme metabolic derangement (Table 1). If uncertainty exists as to whether any such abnormality is attributable to or contributing to the clinical state, the process should be paused and reevaluated once the abnormality has been corrected. Myxoedema coma, Addisonian crisis, and severe neuromuscular weakness can all cause apnoea and coma; if suspected, they should be tested for formally.

Table 1.

Cardiorespiratory, biochemical, and physiological targets to be met before clinical testing¹

Variable	Target
HR	Sinus rhythm 60–100 beats min−1
MAP	60–80 mmHg
CVP	4–10 mmHg
Pulmonary artery pressure	<12 mmHg
Cardiac index	>2.1 L min−1 m−2
Mixed venous oxygen saturation	>60%
Pao2	>10 kPa
Paco2	<6 kPa
Arterial pH/[H+]	7.35–7.45/[45–35 nmol L−1]
Serum Na+	115–160 mmol L−1
Serum K+	>2 mmol L−1
Serum Mg2+	0.5–3 mmol L−1
Serum PO43−	0.5–3 mmol L−1
Serum glucose	3.0–20.0 mmol L−1

Completion of step 1 requires confidence in the absence of sedative drugs or neuromuscular blocking agents contributing to the coma or apnoea. Absence of neuromuscular blocking agent activity can be confirmed by using a peripheral nerve stimulator. Excluding any residual effect of sedatives can be challenging, particularly after prolonged drug infusion, in the presence of impaired renal or hepatic function or where therapeutic hypothermia has been used. Hypothermia decreases liver metabolism and hypothermia-induced reductions in clearance have been shown for a number of sedatives commonly used in the ICU.⁷ Measuring serum concentrations can be useful; midazolam (<5mcg L⁻¹) and thiopental (<10mg L⁻¹).¹ Particular care is needed if prolonged infusions of fentanyl have been used, as the rate of elimination is very context-sensitive. Waiting additional time or undertaking ancillary investigations is recommended where uncertainty exists. Specific antagonists such as naloxone or flumazenil have a limited role.

Once the clinical and structural criteria for testing have been established, prolonged observation is not usually necessary. However, it is recommended that clinical testing is not undertaken:⁸

• (i)<6 h after the loss of the last brainstem reflex and discontinuation of sedative infusions
• (ii)<24 h after the loss of the last brainstem reflex where the primary aetiology is hypoxic brain injury
• (iii)<24 h after restoration of normothermia after therapeutic hypothermia.

Step 2

This comprises a series of clinical tests, undertaken twice, to confirm absence of brainstem function. Practical guidance on how to perform the tests in adults, children, and neonates is available online.⁸ The family should be invited to observe the tests, after an explanation of what to expect, including the possibility of spinal reflexes. The time of completion of the first set of tests is the time of death.

Pupillary reflex

Check for direct and consensual reflex in each eye. Pupils do not need to be the same size or shape. Inability to examine a pupillary reflex (e.g. after trauma) does not preclude DNC, but consider ancillary testing.

Corneal reflex

Ensure the cornea is stimulated, not the sclera. Avoid corneal damage.

Oculovestibular reflex

Visualise each tympanic membrane before injecting ice cold water. Remove any obstructing material (e.g. wax). Look for any eye movement during injection, not just nystagmus.

No motor response to pain

There must be no response in the cranial nerve distribution to supraorbital pressure or a stimulus applied to the periphery. Spinally-mediated reflexes may occur in response to a peripheral stimulus, but differentiating spinal reflexes from retained motor responses requires experience. The process should be paused, and testing repeated if there is uncertainty.

Cough reflex

Absence of a cough in response to stimulation of the carina by passing a tracheal suction catheter via the tracheal tube. The suction catheter must be long enough to stimulate the carina.

Gag reflex

Absence of gag in response to stimulation of the posterior pharynx.

The apnoea test—general points

The apnoea test is only undertaken if all other reflexes are absent. Ensure cardiovascular stability, preoxygenate with 100% O₂, and reduce minute volume to achieve a Paco₂>6 kPa before disconnecting from the ventilator. Observe the patient for 5 min to ensure that there is no spontaneous respiratory effort. Use end-tidal CO₂ monitoring. Check that Paco₂>6.5 kPa and increased by >0.5 kPa before reconnecting.

The apnoea test—pitfalls

The patient remains attached to the ventilator during the apnoea test

Failure to disconnect can lead to ventilator autotriggering where minor pressure or volume changes in the circuit, such as those due to cardiac pulsations, are interpreted as a spontaneous effort and support provided to that interpreted ‘breath effort.’

The patient with a high cervical cord injury

The apnoea test may not be reliable. Ancillary testing should be undertaken.

The patient with chronic obstructive pulmonary disease

Allow CO₂ to increase to >6.5 kPa with a starting pH<7.40.

The patient with cardiorespiratory instability

Stabilise before testing. Preoxygenate the lungs with 100% O₂. Insufflate O₂ via a suction catheter at 5 L min⁻¹. Apply 10 cm H₂O PEEP via a Mapleson C circuit. Adjust vasoactive infusions as required. If unable to complete, consider ancillary testing.

The patient on extracorporeal membrane oxygenation

Drug sequestration in the extracorporeal circuit and difficulties in undertaking the apnoea test are challenging, but do not preclude DNC. Recently published guidance describes how to conduct testing whilst on extracorporeal membrane oxygenation.⁹

Ancillary testing

Ancillary tests are not compulsory in the UK, but are recommended if confounding factors cannot be confidently excluded. Computed tomography angiography, EEG, four vessel cerebral angiography, MRI, transcranial Doppler, and nuclear studies are examples. The choice of investigation depends on local availability and expertise to correctly interpret the results. The presence of cerebral blood flow does not equate to the presence of function nor of reversibility.

Education

In the UK, DNC is undertaken in approximately 1,600 patients annually. Many ICU consultants will do this less than once a year. Forms endorsed by ICU professional bodies (being updated in 2019) are available to guide doctors though all aspects of DNC and highlight the pitfalls.¹⁰ These forms help increase confidence and consistency in DNC. Online educational videos are also available for teaching and training purposes.⁸

Conclusion

Completion of the two steps means that DNC can be confirmed with certainty. Ancillary tests are rarely required, but are useful when the tests cannot be completed in full or when confounding factors cannot be excluded.

Declaration of interest

The authors declare that they have no conflicts of interest.

Biographies

Ian Thomas MRCP FRCA FFICM DICM is a consultant in anaesthesia and intensive care medicine, and clinical lead for organ donation at Southmead Hospital Bristol.

Alex Manara FRCP FRCA is a consultant in anaesthesia and intensive care medicine at Southmead Hospital. He is also regional clinical lead for organ donation for the South West of England and UK national lead for quality and safety in organ donation.

Matrix codes: 1A01, 2C06, 3C00