PREDICTING OUTCOME IN HYPOXIC-ISCHEMIC BRAIN INJURY

PEDIATRIC CRITICAL CARE: A NEW MILLENNIUM

0031-3955/01 $15.00

+ .OO

PREDICTING OUTCOME IN HYPOXIC-ISCHEMIC BRAIN INJURY Sergio J. Jacinto, MD, Maria Gieron-Korthals, MD, and Jose A. Ferreira, MD

Predicting the neurologic outcome of a child who has sustained a severe hypoxic-ischemic event is crucial for several reasons: Families want to know whether their children will survive. If they survive, is there going to be any permanent damage? If there is any neurologic damage, how severe will it be? If physicians have data to support their efforts in predicting outcomes, then they would be able to help families to make treatment decisions. The hardest decision a parent can make is whether to withdraw life support from his or her child when there might be the slightest likelihood of survival, no matter how poor the neurologic outcome might be. Physicians are obligated to be educated on and keep abreast of clinical, laboratory, and neurophysiologic markers that could help to determine how much neurologic damage a child might have sustained in a hypoxic-ischemic event. This article focuses on hypoxic-ischemic encephalopathy (HIE) in children less than 12 years of age caused by neardrowning. DEFINITION OF HYPOXIC-ISCHEMIC ENCEPHALOPATHY

Hypoxic-ischemic encephalopathy implies the interruption of supply of vital nutrients to the brain, mainly oxygen and glucose, sufficiently substantial to cause irreversible damage. When the brain is depleted of oxygen, the result is hypoxic encephalopathy. When the brain is depleted of blood flow, the result is cerebral ischemia. Blood flow could be interrupted regionally, within a specific

From the Departments of Pediatrics (SJJ, MGK, JAF) and Neurology (MGK), University of South Florida College of Medicine, Tampa, Florida

PEDIATRIC CLINICS OF NORTH AMERICA VOLUME 48 * NUMBER 3 * JUNE 2001

647

648

JACINTOet a1

vascular distribution (as with an embolic event causing a stroke), or globally (as with a cardiopulmonary arrest [CPA]) with cessation of cerebral blood flow (CBF), resulting in severe hypoxia and generalized ischemia. This occurrence causes a complex cascade of events that leads to excitotoxic damage from excitatory amino acids, a derangement in calcium homeostasis, and further neuronal damage from oxygen-derived free radi~als.4~ PREVALENCE AND CAUSE OF HYPOXIC-ISCHEMIC ENCEPHALOPATHY

Most cases of HIE result from injury in the prenatal period secondary to intrauterine asphyxia, with disturbance of gas exchange across the placenta and with respiratory failure at birth. Postnatally, respiratory insufficiency secondary to acute respiratory distress syndrome, recurrent apneic spells, or severe rightto-left shunt secondary to congenital heart disease or persistent fetal circulation can cause HIE. Because neonatal brain tolerance to severe hypoxia and ischemia may be different from that of older children, this article focuses on HIE from causes other than those of the perinatal period. Beyond the neonatal period, primary causes of HIE are airway obstruction caused by drowning; choking; suffocation; aspiration; severe, acute asthma; inhalation injury; and obstruction of CBF, such as that occurring in hanging and strangulation. Asphyxia is the most common cause of CPA in Swaiman and Ashwal listed major causes of CPA by system affected31: Upper airway obstruction Croup Foreign body obstruction Supraglottitis Angioedema Trauma (blunt and penetrating) Abscess, bacterial infection Lower airway disease Drowning Inhalation injury Asthma Pneumonia Bronchiolitis Bronchopulmonary dysplasia Chest trauma Foreign body obstruction Neurologic Head trauma (diffuse axonal injury, subdural hematoma, epidural hematoma, subarachnoid hemorrhage) CNS infection Botulism Cervical spine trauma Increased intracranial pressure Ventriculoperitoneal shunt obstruction Status epilepticus Cardiovascular Hypovolemic shock Septic shock Congenital heart disease

PREDICTING OUTCOME IN HYI’OXIC-ISCHEMIC BRAIN INJURY

649

Cardiogenic shock Dysrhythmia Myocarditis Pericardial effusion Coronary artery disease Other Lightning and electric injury Sudden infant death syndrome Burns Intoxication Metabolic disorders Hickey et all6 provide the commonest causes of CPA and their prevalence in Table 1. PHYSIOPATHOLOGY OF HYPOXIC-ISCHEMIC ENCEPHALOPATHY Hypoxia is a common pathway in the pathogenesis of HIE regardless of cause. CPA, for example, produces a complete cessation of CBF and thus global hypoxia and ischemia. Neuronal cell injury and death begins within minutes of hypoxia or ischemia when the supply of nutrients, such as oxygen and glucose, is interrupted. If resuscitation follows an ischemic event, the cell death is not immediate but rather occurs hours or days later. Substrate depletion disrupts cellular production of adenosine triphosphate, and production of energy ceases. The longer the duration of ischemia, the larger and more diffuse the areas of the brain that are affected. Ischemia sets into motion a cascade of events that first involve the most vulnerable areas of the brain, such as the brainstem, hippocampus, and cerebral cortex. Injury progresses exponentially with duration of ischemia and eventually becomes irreversible.I4 It seems that some cell death in HIE is a result of reperfusion of the ischemic tissue after restoration of circulation.3l The mechanisms of cell injury during reperfusion are multifaceted and include a release of oxygen-free radicals from the systemic circulation. The brain is particularly vulnerable to free radical injury, mainly nitric acid, peroxynitrite, hydrogen peroxide, and superoxide. The radicals are highly reactive in the presence of brain polyunsaturated fatty acids, such as arachidonic acid.30They peroxidize cell membranes, alter the bloodbrain barrier, and disrupt DNA synthesis. Ultimately, the cellular metabolism ceases, and neuronal infrastructure is destroyed. The ability of the cell to defend

Table 1. CAUSES OF CARDIAC ARREST IN CHILDREN Cause

Patients (%)

Submersion Sudden infant death syndrome Trauma Respiratory Sepsis Heart disease or arrhythmia Other or unknown

27 20

15 9 9 6 14

From Hickey RW, Cohen DM, Shausbaugh S, et a1 Pediatric patients requiring CPR in the prehospital setting. Ann Emerg Med 25:495, 1995; with permission.

650

JACLNTO et a1

itself against the free radicals is limited and worsens after an ischemic area is reperfu~ed.~' Hypoxemia is the most important consequence of drowning and leads to the development of HIE by the same mechanisms as described earlier. Most of the pathology associated with submersion is related to the duration of asphyxia from the time of the submersion until adequate respiration is restored. NEUROLOGIC APPROACH TO A CHILD FOLLOWING A HYPOXIC-ISCHEMIC EVENT History

History taking and physical examination are important parts of the overall evaluation of infants and children who have sustained a hypoxic-ischemic event. A history should be obtained from a first-hand witness and should include a quick review of systems. History should include recent symptoms or illnesses, such as headaches, vomiting or fever, recent injuries (especially head trauma), and psychiatric disorders. In addition, the possibility of accidental (by toddlers) or purposeful (by teenagers, i.e., a suicide attempt) ingestion of medication must be determined. In the event of a CPA or a submersion accident, the family should be asked the following questions: How much time elapsed between someone last seeing the child and finding the child with asystole? How much downtime or asystolic time was there? (This information is vital to know, although it frequently is not known on arrival of the child in the emergency department.) What was the interval between finding the child and arrival of emergency medical services (EMS), or when was effective cardiopulmonary resuscitation (CPR) instituted? Was CPR started immediately after the child was found? How many rounds of medications (e.g., epinephrine, atropine, and sodium bicarbonate) did this child receive before return of spontaneous circulation was obtained? Were there any pertinent findings on physical examination on the arrival of EMS, including fixed, dilated, or midsized pupils or seizures? What was the initial arterial pH value? In the event of a submersion accident, in what type of water was the child found, and what was the approximate temperature of the water? With this information, physicians have a relatively good idea of the degree of anoxic brain injury that the child has sustained, keeping in mind that some of this information might be inaccurate. General Physical Examination

The physician evaluating a comatose child should have a systematic approach that allows for appropriate diagnostic and therapeutic endeavors and not irrelevant ~onsiderations,3~ and this should be performed after the ABCs (airway, breathing, circulation) have been restored and secured. The physical examination should include: Vital signs: Patient temperature, pulse, respiratory rate, and blood pressure

PREDICTING OUTCOME IN HYPOXIC-ISCHEMIC BRAIN INJURY

651

Head, eyes, ears, nose, throat: A search for evidence of head trauma (i.e., hematomas, scalp lacerations, otorrhea, rhinorrhea, "raccoon sign," Battle's sign, and skull depressions or irregularities) Neck: Assessment of meningeal irritation, although physicians should beware of cervical spine injuries; a patient with head trauma never should be manipulated unless the cervical spine has been cleared radiographically Cardiovascular: Assessment for murmurs, gallops, and arrhythmias Lungs: Assessment for unequal breath sounds, hyperresonance or rib irregularities Abdomen: Assessment for distention or absent bowel sounds Extremitieshkin: Assessment for fractures, bone irregularities, hematomas, and abrasions Neurologic Examination

The neurologic examination in a comatose child can be performed quickly and effectively if a systematic approach is used. The neurologic examination should include:

Mental status: Awake, confused (delirium), lethargic, stuporous, comatose Cranial nerves: Discs on funduscopic examination, pupillary size and light reflex, oculocephalic reflex, corneal reflex, oculovestibular reflex, gag reflex, respiratory pattern Motor examination: In response to noxious stimuli, check for purposeful movement, localizing, withdrawal, decorticate posture, decerebrate posture or no movement at all (flaccid) Deep tendon reflexes: Normal, increased, asymmetric, absent The Glasgow Coma Scale (GCS) was developed in 1974 by Teasdale and JennetPo"as a quick, reproducible method of assessment of comatose patients with prognostic implications (Table 2). A modified scale for infants is shown in Table 3. This scale is used more by EMS technicians and by trauma surgeons than by other physicians on a regular basis. After the initial general and neurologic assessment, the physician should be able to determine the patient's clinical

Table 2. GLASGOW COMA SCALE Activity

Eye opening

Best verbal response

Best motor response

Patient's Response

Score

Spontaneous To speech To pain None Oriented Confused Inappropriate Incomprehensible None Obeying Localizing Withdrawing Flexing Extending None

4 3 2 1 5 4 3 2 1 6 5 4 3 2 1

652

JACINTO et a1

Table 3. MODIFIED COMA SCALE FOR INFANTS Activity ~~~~

Patient’s Response ~

Eye opening

Best verbal response

Best motor response

~

Score

~

Spontaneous To speech To pain None Coos and babbles Irritable cries Cries to pain Moans to pain None Normal spontaneous movements Withdraws to touch Withdraws to pain Abnormal flexion Abnormal extension None

4 3 2 1 5

4 3 2 1 6 5 4 3 2 1

From Davis RJ, Tan VF, Dean JM, et a1 Head and spinal cord injury. In Rogers MC (ed): Textbook of Pediatric Intensive Care. Baltimore, Williams and Wilkins, 1992, pp 805-867

status, including noncoma, coma with preserved brainstem reflexes, or brain death.

Brain Death Criteria

In 1968, the Report of the Ad Hoc Committee of the Harvard Medical School to Examine the Definition of Brain Death and, shortly thereafter, the Collaborate Study of the National Institutes of Neurologic Diseases and Stroke included the criteria for brain death.’=These criteria have been endorsed by the American Academy of Neurology, the American Electroencephalographic Society, the American Bar Association, the American Medical Association, the National Conference of Commissioners on Uniform State Laws, and the President’s Commission for the Study of the Ethical Problems in Medicine and Biomedical and Behavioral Research. More recently, the American Academy of Neurology published standards for the diagnosis of brain death. After ensuring that the patient has coma that is irreversible or reversible causes have been excluded, with the temperature more than 33°C and no confounding drug intoxication, clinical testing must reveal no evidence of cortical or brainstem function for a diagnosis of brain death. Pupils may be midsized and fixed and not necessarily fixed and dilated as previously thought. Complementary studies include an isoelectric EEG and an apnea test (that includes at least a 10-minute preoxygenation with 100°/~oxygen) allowing the Paco2 to increase to more than 60, with no spontaneous breaths being consistent with brain death. A perfusion nuclear scan may be of help in confirming absence of CBF. The ancillary studies are confirmatory and not necessarily required for diagnosis. Two physicians must be present, one of whom is the treating physician, and the second, a board-eligible or certified neurologist, neurosurgeon, internist, pediatrician, surgeon or anesthesiologist.

PREDICTING OUTCOME IN HYPOXIC-ISCHEMIC BRAIN INTURY

653

PREDICTING NEUROLOGIC OUTCOME Different outcomes can develop from a coma after a sustaining hypoxicischemic insult. Different factors play important roles as to whether a child may have no neurologic deficits and whether he or she will survive or, worse, survive to live in a persistent vegetative state (PVS). Data to help predict as early and as accurately as possible the neurologic outcome of a child who has sustained a severe hypoxic-ischemic event are shown in Table 4. Predicting the outcome of a child who has sustained a hypoxic-ischemic event is based on different factors, including the cause. Most of the data was collected from the literature about near-drowning because of its high prevalence in the state of Florida. Approximately 68% to 90% of children who have sus34 tained a submersion accident have good A study by Jacobsen et all7of 26 pediatric near-drowning victims concluded that all patients with spontaneous respirations immediately after CPR survived, with minimal or no neurologic impairment. Biggart and Bohn4reported on the effect of hypothermia on outcome of drowning in children who arrived in an emergency department following warm-water drowning. Children with significant hypothermia and children without spontaneous respirations and with asystole died or had severe neurologic damage.* Duration of Anoxia Sufficient literature supports that there is a better outcome in children in whom CPR was started at the accident scene.18From this comes the importance of the duration of asystole or the duration of submersion in the case of a drowning.18,32, 35, 41 In general, children who were submerged for more than 25 minutes or who needed prolonged resuscitation (>25 minutes) have unfavorable outcomes. In one study,12 of patients who required CPR for 10 minutes or less, 95% survived with no or minimal neurologic impairment. Patients who arrived in the emergency department with a spontaneous pulse had good neurologic outcomes,12whereas children who had ongoing CPR on arrival to the emergency department died or survived but with severe anoxic encephalopathy.2,26, 28, 32, 38 Water Temperature In general, warm-water submersion accidents are associated with poor outcome, whereas icy-water submersion accidents of the same duration of submersion are associated with a better outcome?, 26, 27, 34

Table 4. GRADES OF OUTCOME FROM COMA Grade

Good recovery Moderate disability Severe disability Persistent vegetative state No recovery

Definition

Back to baseline Not back to baseline, with mental or physical disabilities Depend on others for activities of daily living; regain some cognitive function Awake, no cognitive awareness Coma until death

654

JACINTO et a1

Glasgow Coma Scale Score on Presentation

Children who present to the emergency department with a GCS score of 6 or more have an excellent likelihood of full neurologic recovery, whereas a GCS score of less than 5 is associated with a high risk for death or severe CNS damage8,38,34; however, cases of children with GCS scores as low as 3 or 4 surviving with full neurologic recoveries have been reported.s pH on Presentation

Patients with an initial pH value of less than 7.1 rarely have a favorable outcome.12 Hyperglycemia

Several studies have looked at hyperglycemia in relation to poor outcomes.20, In general, blood glucose levels of more than 250 mg/dL were related to death and persistent vegetative states, especially when coupled with a GCS score of 3 or less.23The higher the blood glucose level, the worse the hypoxicischemic injury.= 23, 24

Neurologic Examination

Between 1973 and 1977,500 patients with nontraumatic coma were studied by Levy and Bates,19 who looked at neurologic examination on admission and days 1,3, and 7 after a nontraumatic hypoxic-ischemic event (Fig. 1).Significant findings were identified, especially in patients who did not have two or three of the reflexes, including corneal, pupillary reaction to light, or oculovestibular, on admission and day 1. Only 1%and 2%, respectively, of patients recovered with moderate disability or good recovery. On day 3, corneal reflex and oculovestibular or motor reflex had to be present. If not, 4% had severe disability, and the rest remained in a PVS. On day 7, eye opening had to be present, at least in response to pain. If not, 100% had no recovery, PVS, or severe disability. This study identified clinical features in comatose patients that present within the first week following a hypoxic-ischemic event that are important in predicting neurologic recovery. This study’s population included patients aged more than 12 years, with a mean age of 59 years. In the authors’ experience, similar outcomes are found in children, although they are not aware of a study of this caliber having been performed in children. Duration of Coma

Bratton and Jardine5performed a retrospective chart review of 24 children admitted to a pediatric ICU during a 5-year period. One conclusion was that comatose drowning victims who do not recover consciousness within 24 hours after submersion have a uniformly poor outcome. Table 5 indicates the summary of the variables indicating a poor neurologic outcome. Neuroimaging

A CT scan of the head is the initial study of near-drowning victims and usually is obtained in the emergency department to rule out other causes of coma, particularly an intracranial hemorrhage. In most cases, initial CT scans

PREDICTING OUTCOME IN HYPOXIC-ISCHEMIC BRAIN INJURY

655

Table 5. POOR NEUROLOGIC OUTCOME Characteristic

Value

Duration of anoxia / asystole Water temperature Glasgow Coma Scale score on presentation pH on presentation Initial blood glucose Duration of coma

> 25 min Warm ( 2 10°C) (4 < 7.10 > 250 mg/dL > 24 h

are normal or show nonspecific changes not helpful in predicting outcome. At times, it takes up to 24 hours for some hypoxic-ischemic changes, which include poor gray and white differentiation secondary to edema, to appear on CT scanning. This occurrence leads to decreased ventricular size and effacement of the sulci with decrease in the basal cistern area. This finding can be appreciated in serial CT scans. MR imaging of the brain seems to be much more sensitive to early changes and therefore, after initial stabilization of the patient, an MR image should be obtained if clinically indicated. Of the different sequences, diffusion-weighted MR imaging seems to be the most sensitive, especially when performed during the early subacute phase (24 hours to 13 days after the hypoxic-ischemic event).' NEUROPHYSIOLOGIC TESTING

In addition to the physical and neurologic examinations, neurophysiologic testing may add valuable information to the evaluation of patients with HIE that, in turn, may help in predicting outcome. The most valuable methods include visual or flash evoked potentials, somatosensory evoked potentials (SSEPs), brainstem auditory evoked responses (BAERs), and electroencephalography (EEG). These have been studied extensively and when combined with the overall neurologic evaluation may be helpful in predicting the outcome in cases of HIE. SSEP studies of comatose children may be of prognostic value, as demonstrated in multiple studies. DeMeirleir and Taylor'O recorded SSEPs in 73 comatose children with multiple causes of coma admitted to an ICU. Fourteen children who recovered normally had increased interpeak latencies or normal SSEPs. Patients with normal outcomes eventually developed normal interpeak latencies. Absent cortical SSEP unilaterally in 5 patients was predicative of a hemiparesis. Of 31 children with bilaterally absent cortical responses, 19 died and 12 developed severe neurologic sequelae. Lutschg et a122studied 43 comatose children with median nerve SSEPs. Coma was secondary to head trauma in 26 patients and to HIE in 17 patients. Of the 15 fatal cases, initial SSEP studies showed bilateral absence of the cortical potentials in 12, unilateral absence in 1, and delayed latency in 1, and 1 had a normal study. The cause of death of these latter three cases was not given. Follow-up SSEP studies of patients who eventually died were described as severely abnormal. All five patients with brain death confirmed by cerebral angiography had bilaterally absent cortical components. Patients who experienced normal outcomes had initially normal or mildly delayed cortical SSEPs. In another study by Frank et al,13 SSEPs proved more reliable than the clinical examination or EEG for predication of outcome. In a study by White et a1,44of nine patients who had absent waves N19-P22 bilaterally, four died and five progressed to PVS. Similar to adults, the bilateral

656

JACINTO et a1

Patient Examination 1

Best 1-Year Recovery (%)

Any 2 Reactive? Corneal Pupil Oculovestibular

r

Number of Patients

p

Verbal Moans?

I

No Recovery Vegetative State

Severe Disability

Moderate Disability Good Recovery

46

13

41

58

19

23

69

14

17

80 97

8 2

12 1

56

4 No Motor Withdrawal?

I

106

I

L

A

9

Patient Examination

I

Any3;&ive? Corneal

Best 1-Year Recovery (%)

Oculovestibular (Motor)

Number of Patients

No Recovery Vegetative State

Severe Disability

Moderate Disability Good Recovery

Verbal At Least Inamrominate

24

0

33

67

42

21

31

104

16

13

11

36

84 98

11 0

4 2

I

r

Motor Withdrawal? At Least

I

I36

-

Any 1 presents Oculcephalic: NL Oculovestibular: NL Soon1 Eve Movement: NL Motor Extension or Flexion .

k I No

-

L Q7

3

0 ,

Figure 1. Neurologic examination of patients with nontraumaticcoma. Finding on admission to hospital: 500 patients (A), 387 patients at 1 day (B),261 patients at 3 days (C), and 179 patients at 7 days (0)NL . = normal. (From Levy DE, Bates D: Prognosis in nontraumatic coma. Ann Intern Med 94:293-301, 1981; with permission.) Illustration continued on opposite page

PREDICTING OUTCOME IN HYl'OXIC-ISCHEMIC BRAIN INJURY

657

Patient Examination ~

~~~

Best 1-Year Recovery (%) Corneal

Number of Patients

No Recovery Vegetative State

Severe Disability

Moderate Disability Good Recovery

At Least Inappropriate Words?

0

26

14

At Least Withdrawal?

40

21

33

16

16 4

8 0

Verbal

62 56

96

C Patient Examination Best 1-Year Recovery (%) At Least To Pain

Motor At Least Localizing

Number of Patients

No Recovery Vegetative State

Severe Disability

Moderate Disability Good Recovery

99

1

24

15

54

63 92

28 8

10 0

D Figure 1 (Continued).

absence of waves N19-P22 in children in coma is highly predictive of a poor outcome. Information on the value of BAERs in neonates, infants, and children with brain death is more limited. Steinhart and we is^^^ examined 10 clinically brain-dead children (median age, 8.5 years). Nine of them lacked all components of BAERs, and only one patient had wave I. In contrast, 13 comatose but not brain-dead children (median age, 3 years) had at least waves I and V present. The usefulness of BAERs in HIE remains controversial because they can be altered by peripheral and central auditory disorders. Furthermore, some evidence suggests that BAERs are especially vulnerable to hypoxia in young patients. Following hypoxic episodes, two neonates studied by Dear and Godfrey9 and one 33-month-old child studied by Taylor et a140lost all BAER waves or all components after wave I. Responses returned after several days. None of these patients fulfilled criteria for brain death. One of them survived in PVS, and the other was delayed de~elopmentally.~~ In some cases, BAERs may disappear transiently in neonates, infants and small children who survive, although not necessarily with good outcome. In early studies of evoked potentials in comatose states, Bergamasco et aP and Walter and Arfe143performed serial flash or visual evoked potentials to follow evolving comatose states. Subsequent investigators usually combined

658

JACINTO et a1

visual evoked potentials with BAERs, SSEPs, or both. There is no consensus as to which tests are the best indicators of outcome, but, when combined with clinical and EEG data, they may be useful. Patients who are in a coma or who are otherwise unresponsive show a wide spectrum of EEG patterns of variable reactivity. In rare instances, a single EEG can indicate the nature of the underlying disorder. In most instances, however, a single EEG provides little, if any, useful information on the cause and reversibility of coma. This lack of specificity of cause is reflected best in many patterns, including burst suppression, periodic bilateral epileptiform discharges, widespread nonreactive rhythms, irregular frequency, and low-voltage nonreactive records that can occur in patients with HIE or drug intoxication uncomplicated by CPA. This limitation is overcome only partly by serial EEGs. Whether an EEG can predict the outcome of patients in a comatose state is debatable. Lack of specificity of EEG patterns significantly diminishes the prognostic value of individual EEGs. A pattern such as burst suppression, for example, predicts poor outcome in patients with HIE, whereas satisfactory recovery commonly occurs with the same pattern in properly treated patients with toxic encephalopathies uncomplicated by CPA. Even when the cause of coma is known, the prognostic use of a single EEG is far from e~tablished.~ Few studies address the significance of electrocerebral inactivity in comatose infants and children. Harden15reported on results in 60 infants, aged 1 day to 6 months, who were resuscitated after cardiac arrest. She found that interpreting EEG findings was complicated by various factors, including congenital anomalies, severe systemic metabolic derangements, and repeated circulatory disturbances. Effects of these conditions, together with transient cerebral ischemia and anoxia, generated more varied EEG patterns than did CPA in older patients. Despite this, the analysis of 2180 EEGs recorded over 20 years on 363 resuscitated children enabled Pampiglione and Harden29to conclude that there was ”little difficulty in predicting that cerebral functions were irreparably lost when electrocerebral silence persisted for 6 to 12 hours in repeated records.” From the currently available information, it seems that evoked potentials are sensitive indicators of the clinical outcome of comatose patients and are far superior to EEG7 as presently performed and interpreted. The available data indicate that, unlike EEGs, evoked potentials are not altered by most medications in concentrations commonly used in comatose patients, such as high-dose barbiturates, in the absence of hypothermia or acute alcohol intoxication. Although visually grading EEG features is difficult, evoked potentials provide exquisitely quantitative measures easily related to indices of the patient’s clinical state and are readily amenable to statistical analysis. Because available data on the reliability of SSEPs are solid, SSEPs will continue to provide a noninvasive, complementary, and objective method of evaluating the somatosensory system in infants and children. SUMMARY

Predicting the neurologic outcome of children after a hypoxic-ischemic event continues to be a challenge for intensivists and pediatric neurologists. Nevertheless, with accurate history taking, serial neurologic examination, and some ancillary studies, the clinician can predict accurately whether a child will die or have profound neurologic damage. Aggressive resuscitation should be offered to all children when found in CPA. A simple ingestion might have led to this clinical scenario, and complete neurologic recovery may be possible if effective resuscitation is implemented. In cases of drowning, several factors, if present, are consistent with profound neurologic sequelae or death. These in-

PREDICTING OUTCOME IN HYPOXIC-ISCHEMIC BRAIN INJURY

659

clude prolonged submersions with asystole, delayed onset of CPR, no spontaneous respirations on arrival to the emergency department, and low initial pH value. The options of withdrawal of life support or a DNR status should be offered to families of children who have survived a devastating hypoxicischemic event but who are in a PVS. If brain-death criteria have been fulfilled, the patient must then be disconnected from life support after organ donation has been discussed with the family. ACKNOWLEDGMENTS Dr. Jacinto thanks University Community Hospital, Audio/Visual Department; Rick Ruge and Daniel Baker; Sharon Henrich from the Medical Library; his assistant, Ileana Lopez; his wife and children for being very patient in his absence in preparing the manuscript; and the person who put it all together, his transcriptionist, Carol T. Benton.

References 1. Arebelaez A, Castillo M, Mukherji SK Diffusion-weighted MR imaging of global cerebral anoxia. AJNR Am J Neuroradiol20999-1007, 1999 la. Beecher HK A definition of irreversible coma: Report of the ad hoc committee of the Harvard Medical School to examine the definition of brain death. JAMA 205:85-88, 1968 2. Bell TS, Ellenberg L, et al: Neuropsychological outcome after severe pediatric neardrowning. Neurosurgery 173604-608, 1985 3. Bergamasco B, Bergamini L, et al: Longitudinal study of visual evoked potentials in subjects in posttraumatic coma. Schweiz Arch Neurol Neurochir Psychiatr 97:l-10, 1966 4. Biggart MJ, Bohn DJ: Effect of hypothermia and cardiac arrest on outcome of neardrowning accidents in children. J Pediatr 117179-183, 1990 5. Bratton SL, Jardine DS: Serial neurologic examinations after near-drowning and outcome. Arch Pediatr Adolesc Med 148:167-170, 1994 6. Calder RA, Clay CY Drownings in Florida. J Fla Med Assoc m679-682, 1990 7. Daly DD, Pedley TA: Coma, other states of altered responsiveness and brain death. In Current Practice of Clinical Electroencephalography. New York, Raven Press, 1990, pp 425487 7a. Davis RJ, Tan VF, Dean JM: Head and spinal cord injury. In Rogers MC (ed): Textbook of Pediatric Intensive Care. Baltimore, Williams and Wilkins, 1992, pp 805-867 8. Dean JM, Kaufman ND: Prognostic indicators in pediatric near-drowning: Glasgow Coma Scale. Crit Care Med 9:536-539, 1981 9. Dear BRF, Godfrey DJ: Neonatal auditory brainstem response cannot reliably diagnosis brainstem death. Arch Dis Child 6017-19, 1985 10. DeMeirleir LJ, Taylor MJ: Prognostic utility of SEPs in comatose children. Pediatr Neurol 3:78-82, 1987 11. DeNicola LK, Falk JT, et al: Submersion injuries in children and adults. Crit Care Clin 13~477-502,1997 12. Fisher DH: Near drowning. Pediatr Rev 14:14&151,1993 13. Frank LM, Furgiuele TL, Etheridge GE: Prediction of chronic vegetative state in children using evoked potentials. Neurology 35:931-934, 1985 14. Greenwald BW, et al: Critical care of children with acute brain injury. Adv Pediatr 4247-89, 1995 15. Harden A EEG studies following resuscitation after cardiac arrest in 60 babies [abstract]. Electroencephalogr Clin Neurophysiol 27:333, 1969 16. Hickey RW, Cohen DM, et a 1 Pediatric patients requiring CPR in the prehospital setting. Ann Emere Med 25:495.1995 17. Jacobgen WK, Magon LJ, Briggs BA, et al: Correlation of spontaneous respirations and neurologic damage in near-drownings. Crit Care Med 11:487, 1983 18. Levin DL, et al: Drowning and near-drowning. Pediatr Clin North Am 40:321-336, 1993

660

JACINTO et a1

19. Levy DE, Bates D, Caronna JJ, et al: Prognosis in nontraumatic coma. Ann Intern Med 94293-301,1981 20. Longstreth WT, Inui T S High blood glucose level on hospital admission and poor neurological recovery after cardiac arrest. Ann Neurol 15:59-63, 1984 21. Longstreth WT, Diehr P: Neurologic outcome and blood glucose levels during out of hospital cardiopulmonary resuscitation. Neurology 361186-1191, 1986 22. Lutschg J, Pfeinger J, Lutin HIP, et al: Brainstem auditory evoked potentials and early somatosensory evoked potentials in neuro-intensively treated comatose children. Am J Dis Child 137421426, 1983 23. Maichaud LJ, Rvara FP, et al: Elevated initial blood glucose levels and poor outcome following severe brain injuries in children. J Trauma 31:1356-1362, 1991 24. Merguerian PA, Perel A: Persistent nonketotic hyperglycemia as a grave prognostic sign in head-injured patients. Crit Care Med 9:838-840, 1981 25. National Safety Council: Accident Facts, 1993. Itasca, IL, National Safety Council, 1993 26. Nichter MA, Everett PB: Childhood near-drowning. Crit Care Med 17993-995, 1989 27. Orlowski JP: Drowning, near-drowning, and ice water submersions. Pediatr Clin North Am 34:75-92,1987 28. ORourke PP: Outcome of children who are apneic and pulseless in the emergency room. Crit Care Med 14466-468, 1986 29. Pampiglione G, Harden A Resuscitation after cardiovascular arrest Prognostic evaluation and early electroencephalographic findings. Lancet 1:1261-1264, 1968 30. Pediatr Rev 14, 1993 31. Perkins RM, Ashwal S: Hypoxic ischemic encephalopathy in infants and older children. In Swaiman KF, Ashwal S (eds): Pediatric Neurology Principles and Practice. St. Louis, Mosby, 1999, pp 916-921 32. Peterson B: Morbidity of childhood near-drowning. Pediatrics 593364-370, 1997 33. Plum F, Posner JB: Prognosis of coma. In the Diagnosis of Stupor and Coma, ed 3. Philadelphia, FA Davis Company, 1980, pp 209,349 34. Quan L, Gore EJ, Wentz K, et al: 10 year study of pediatric drownings and neardrownings in King County, Washingion: Lessons & injury prevention. Pediatrics 83~1035-1040.1989 35. Quan L, Kinder D: Pediatric submersions: Prehospital predictors of outcome. Pediatrics 90:909-913, 1992 36. Sachdeva RC: Near drowning. Crit Care Clin 15:281-296, 1999 37. Schindler MB, Bohn D, et al: Outcome of out-of-hospital cardiac or respiratory arrest in children. N Engl J Med 335:1473, 1996 38. Spack L, Gedeit R Failure of aggressive therapy to alter outcome in pediatric neardrowning. Pediatr Emerg Care 13:98-102, 1997 39. Steinhart CM, Weiss IP: Use of brainstem auditory evoked potentials in pediatric brain death. Crit Care Med 13:560-562, 1985 40. Taylor J, Houston BD, Lowry NJ: Recovery of auditory brainstem response after a severe hypoxic ischemic insult. N Engl J Med 309:1169-1170, 1983 40a. Teasdale G, Jennett B: Assessment of coma and impaired consciousness. A practical scale. Lancet 2:81, 1974 41. Turner, et al: Improvement of neurologic status after pediatric near-drowning accidents. Crit Care Med 131080, 1985 42. Vannucci RL, Perlman JM: Interventions for perinatal hypoxic ischemic encephalopathy. Pediatrics 100:1004, 1997 43. Walter S, Arfel G: Responses aux stimulations visuelles dans les etats de coma, aigu et de coma chronique. Electroencephalog Clin Neurophysiol 3227-41, 1972 44. White LE, Frank LM, Furgiuele TL, et a 1 Childhood coma: Improved prediction of outcome with somatosensory and brainstem auditory evoked potentials. Ann Neurol 18417,1985 45. Wijdicks EFM Anoxic-ischemic encephalopathy. In Gilchrist JM: Prognosis in Neurology. Boston, Butterworth-Heinemann, 1998, pp 7-10

Address reprint requests to Sergio J. Jacinto, MD 4602 N. Armenia Avenue Suite D-1 Tampa, FL 33603