Guidelines Before American Academy of Neurology

In 1968, the Harvard Criteria, driven by the advances in critical care medicine, the advent of mechanical ventilation, and issues surrounding organ donation, were introduced in a landmark publication examining the definition of irreversible coma as a new criterion for death. Coma in an individual with no discernible central nervous system activity was characterized by the following 4 features: unreceptivity and unresponsiveness, absent movements or breathing, absent reflexes, and a flat electroencephalogram.

More than a decade later, the President’s Commission put forth the Uniform Determination of Death Act (UDDA), which stated “an individual who has sustained either (1) irreversible cessation of circulatory and respiratory functions, or (2) irreversible cessation of all functions of the entire brain, including the brain stem, is dead. The determination of death must be made in accordance with accepted medical standards.” However, no definite criteria or algorithmic approach were delineated, thus allowing for continued variation in interpretation and practice.

It was not until 1995, when the American Academy of Neurology (AAN) set forth practice parameters on the determination of brain death, that there came a standard by which brain death could be algorithmically assessed. Using principles from the definition provided by the UDDA, the AAN proposed “accepted medical standards” for the determination of brain death. Brain death was defined as “the irreversible loss of function of the brain, including the brainstem.” This was the first publication to not only give precise definitions but actually recommend a methodical approach to the clinical diagnosis of brain death, apnea testing, the use of ancillary tests, and the documentation of brain death in the medical record. The prerequisites for proceeding with the clinical diagnosis consisted of clinical and/or neuroimaging evidence of an acute central nervous system catastrophe compatible with the clinical diagnosis of brain death, the exclusion of complicating medical conditions (electrolyte, acid-base, or endocrine derangements), absence of drug intoxication or poisoning, and a core temperature of at least 32°C. This publication is the foundation for the current practice of declaring brain death in neurologically devastated patients. See Table 1 for a summary of the guidelines leading up to the AAN’s practice parameter.

Current AAN Guidelines

In 2010, the AAN published an update to the 1995 guideline focusing on several clinical questions regarding the potential for misdiagnosis when using the criteria, adequate observation times, the implications of complex motor movements, and new ancillary tests. The guidelines incorporated new evidence to date since the prior guidelines and provided a step-by-step approach to brain death determination ( Box 1 ) that emphasized the 3 clinical findings necessary to declare brain death: coma (with a known irreversible cause), absence of brainstem reflexes, and apnea. The AAN guidelines do not differentiate between a primary brainstem lesion (“brainstem” death) and a lesion of the cerebrum and brainstem (“whole brain” death) as long as the examination is consistent with brain death. Guidelines in other countries vary in how they deal with patients with primary brainstem lesions ( Table 2 ).

Variability Among Institutions

Because state law drives brain death policies on an institutional level, there remains significant practice variation despite the new detailed AAN Practice Parameter. Although most US state laws have adopted the UDDA and reference the AAN Practice Parameter, many have amendments addressing physician qualifications and the need for a second examination, as well as religious exemptions. Practice inconsistencies are evidenced both anecdotally through the personal experiences of providers who perform such declarations and systematically in relatively recent reviews of policies at esteemed institutions of neurology and neurosurgery across the country. The disparity is widespread, ranging from which prerequisites should be met, what the lowest acceptable core temperature should be, how the apnea test should be performed, how many examiners should be required, and what and under which circumstances an ancillary test should be used.

The absence of federal or national standards and the vague policies adopted on a state and institutional level allow for tremendous liberty in interpretation, which contributes to the confusion for providers and mistrust among the public. There is room for misinterpretation and a grave potential for mistakes, of which the media never hesitates to take full advantage. Most recently, a New York Post article highlighted several cases in which the motives of declaring brain death were called into question. It is the variation in practice as well as on a state law level that in many ways adds further fuel to the fire. For example, New York State law requires that physicians consider accommodating families who, on religious or moral grounds, desire the maintenance of mechanical ventilation once a patient has been declared dead by neurologic criteria. However, it is up to individual hospitals to establish written procedures on what constitutes reasonable accommodations in such circumstances. Because objections to the brain death standard based solely on psychological denial that death has occurred or on an alleged inadequacy of the brain death determination are not based on the individual’s moral or religious beliefs, “reasonable accommodation” is not required in such circumstances. Unfortunately, such accommodations can send mixed messages to the public, particularly when there are publications of brain dead patients “surviving” for more than 14 years. The ethical and legal controversies surrounding brain death will be touched on later in the article, but a comprehensive discussion is beyond the scope of this article.

With brain death guidelines falling under the jurisdiction of individual states’ legislatures and ultimately at the discretion of institutions to formalize and implement, it will be difficult to reach conformity. This has led to an outcry from experts in the field to push for a national standard for the declaration of brain death, as well as advocate for a certification process, akin to Advanced Cardiac or Trauma Life Support (ACLS) training, for providers involved in the sensitive assessment of these patients.

Guidelines in Other Countries

Just as there is great variability across North America, there is great variability across the world with regard to the diagnosis of brain death. Until Wijdicks published on the subject in 2002, the degree to which countries differed had not been formally characterized. Wijdicks reviewed original brain death documents for 80 countries throughout the world; the differences were astounding. Although brainstem reflexes were consistently evaluated, there are marked differences in apnea test performance. A Pa co ₂ target value for the confirmation of apnea was used in only 59% of the guidelines, whereas 28% felt disconnecting a patient from the ventilator for 10 minutes after preoxygenation with 100% oxygen was sufficient. Additionally, the number of physicians required to diagnose brain death varied, with 44% of countries requiring 1 physician, 34% requiring 2 physicians, and 16% requiring more than 2 physicians. Forty percent of countries required ancillary testing, although the type varied considerably and appeared arbitrary. In one country surveyed, a cerebral angiogram was performed twice, separated by an “adequate” observation period to document the absence of cerebral blood flow. Half of countries surveyed require 2 or more physicians to confirm brain death. Most of Africa not only did not have legal provisions for organ transplantation but also could not perform brain death criteria testing simply because it was too difficult. In the Middle East, very few countries have official guidelines for determining brain death. Conceptually, brain death by neurologic criteria has not been accepted in Asia. China has no legal criteria for the determination of brain death. In countries that do have guidelines, there are often panels of doctors who corroborate in order for the declaration of death to be made. Canada published criteria in 1999 that were quite similar to the AAN guidelines, differing only in that they did not require oculocephalic reflex testing, permitted hypothermia (Temperature ≥32.2°C) during apnea testing, as well as a variable interval between examinations depending on the etiology.

Countries vary not only in their neurologic criteria for brain death but also in the way they approach primary brainstem lesions, global hypoxic-ischemic injury, and drug intoxications, including sedatives and analgesics. See Table 2 for a comparison of the approach in Canada, the United Kingdom, and Germany to common clinical situations surrounding brain death.

Common Pitfalls

To our knowledge, there are no peer-reviewed reports in medical journals of conditions mimicking brain death that have detailed a complete brain death examination. The most-cited potential mimics are fulminant Guillain-Barre syndrome, baclofen overdose, barbiturate overdose, delayed vecoronium clearance, and hypothermia. However, when the criteria for brain death are used correctly and corroborated with ancillary testing when necessary, there should be no concern for the misdiagnosis of brain death in such conditions.

On the other hand, there are several clinical signs or “red flags” that should caution one from moving forward with the assessment of brain death. These include, but are not limited to, a normal computed tomography (CT) scan, unsupported blood pressure, absence of diabetes insipidus, marked heart rate variations, fever or shock, marked metabolic acidosis, hypothermia lower than 32°C as this is often accidental and reversible, marked miosis (opiate or organophosphate toxicity), myoclonus (lithium or selective serotonin reuptake inhibitor [SSRI] toxicity), rigidity (SSRI or haloperidol toxicity), profuse diaphoresis, and positive urine or serum toxicology.

The AAN guidelines recognize that “because there are deficiencies in the evidence base, clinicians must exercise considerable judgment when applying the criteria in specific circumstances” and that “ancillary tests can be used when uncertainty exists about the reliability of parts of the neurologic examination or when the apnea test cannot be performed.” There are conditions in which the diagnosis of brain death cannot be made on clinical grounds alone and confirmatory testing may be required. These include severe facial trauma preventing complete brain stem reflex testing, preexisting pupillary abnormalities, and sleep apnea or severe pulmonary disease resulting in chronic retention of carbon dioxide.

Patients who meet criteria for brain death are by definition motionless. However, there are a host of reflexes and movements that can cast doubt on the diagnosis and be alarming for the inexperienced examiner. Fortunately, many of these observations are well described phenomena in the literature and are entirely consistent with brain death. In brain death, several functions, including temperature regulation, hypothalamic-pituitary-adrenal axis function, as well as spinal reflexes can be maintained for up to several hours/days. The presence of spinal reflexes is not surprising, given early twentieth century work showing retained forward location in cats and dogs with transected spinal cords. It has been postulated that it is the absence of cortical inhibitory and modulatory afferents to spinal cord centers that allows for the activation of basic spinal cord sequences, causing the reflex movements commonly seen in brain death. In addition to spinal reflexes, movements after death have been described during apnea testing and during organ procurement, as well as in the morgue. Movements can manifest in the head, neck, upper and lower extremities, and the trunk, and have been named “Lazarus signs” for their biblical connotation. These movements range from neck, limb, and trunk flexion to facial twitches and finger jerks.

In a recent publication describing a “new” spinal reflex observed in brain death, Mittal and colleagues outline the following 5 aspects of the movements after brain death that may assist the clinician in differentiating spinal from postural responses:

1. There is no resemblance of a spinal response to the classic postural motor responses. These responses are recognized by synchronized decorticate (thumb folded under flexed fingers in a fist, pronated forearm, flexed elbow, and extended lower extremity with inverted foot) or decerebrate responses (pronated and extended upper and lower extremity).

2. Most often, the spinal responses are slow and short in duration. However, there can be some exceptions as follows: finger flexion can be seen as quick jerks with minimal excursions, and lower extremity responses are often more complex and can be wavy or shocklike.

3. The most common spinal response is triple flexion response (flexion in foot, knee, and hip) which may have variations, such as undulating toe sign or a Babinski sign.

4. Most movements are provoked and not spontaneous. The provocation can be movement during nursing care procedures of the patient, such as turning in bed or transfer from bed to a transport cart.

5. In some patients, spinal responses can be elicited by forceful neck flexion and by noxious stimuli below cervicomedullary junction. They are not seen with pressure at the supraorbital ridge or temporomandibular joint.