Recognition of Neurologic Complications in the PICU Patient

Children who develop neurologic complications of critical illness deserve a specific mention. This is likely an underrecognized population in many PICUs, not least because of the prevalence of nonconvulsive seizures⁴ (NCS) or ICU-acquired paresis.⁵ In these cases, the recognition of new neurologic deficit(s) relies on the ability of the PICU team to recognize changes in the neurologic exam. In general, interventions to treat or attenuate neurologic insults, whether seizures, ischemia, or increasing intracranial pressure (ICP), are more likely to be successful if initiated early in the process of injury. This means that effective neurologic monitoring in the PICU does not rely solely on the availability of a neurologist, electroencephalogram (EEG), or neuroimaging. Rather, the prerequisite to using these monitoring tools is a high index of suspicion for new neurologic injury on the part of the medical team for patients who do not have primary neurologic injuries.

In addition to the physical exam, there are a number of modalities that can be used for the monitoring of neurologic function in critically ill children and adults.^6,7 This includes continuous EEG monitoring, which is well-established as a modality essential for the detection of NCS in the adult and PICU populations,^8,9 ICP monitoring, transcranial Doppler,¹⁰ and metabolic¹¹ and tissue oxygen monitoring.^6,12 Thresholds for detection of cerebral ischemic injury have been proposed for brain oxygen tension¹³ and near-infrared spectroscopy (NIRS).¹⁴ With respect to pediatric neuromonitoring, as yet there is no consensus on normal values or the age-dependence of these endpoints.

Establishing a Baseline Neurologic Examination and Anticipatory Planningfor New Deficits

The examiner should discuss the findings with the bedside nurse and other members of the medical team. If possible, abnormal findings should be demonstrated to the members of the team. There should be a consensus on the key findings and the approach to be taken for evaluation and management if the neurologic exam changes. Important features of the approach to the neurologic examination of the child in the PICU include the need for serial examinations and the need to establish a plan to manage any changes in this exam before they occur. This is the most important message of this chapter.

History and Assessment of Risk Factors

For children in the PICU, the medical history may provide essential information about the mechanisms and timing of neurologic insult. This information is needed for the interpretation of the neurologic exam and assessment of potential mechanisms of neurologic injury. These data are then used to assess the risk for progression of neurologic injury, to prioritize therapeutic interventions, and to determine the need for, and timing of, additional evaluations, including imaging, EEG, and laboratory studies. In many cases, the decisions to treat (or not treat) a neurologic injury in the PICU will have to be empiric and based on a careful assessment of risk factors and potential mechanisms of injury as well as a weighing of the risks and benefits of such intervention. For example, the approach to evaluation of altered mental status presenting to the PICU will be different with children with an established complex partial seizure disorder, sickle cell disease, or a stem cell transplant recipient on immunosuppressant drugs. Before examining the patient, knowledge of salient details of the medical history allows the examiner to begin to formulate a diagnostic approach and to prepare the early steps in management. The patient with a history of epilepsy in the example above would likely require an emergent EEG to rule out NCS, while the patient with sickle cell disease and risk factors of vascular injury with the same symptoms may require an imaging study, and the immunosuppressed patient a lumbar puncture. The presenting neurologic symptoms may be similar, but the history dictates the assessment based on the likely mechanisms producing the deficits found on neurologic examination.

Iatrogenic Complications of Pharmacotherapy

It is not uncommon to overlook the side effects of drugs in the PICU. A review of medications should be part of the neurologic assessment of any patient with new neurologic symptoms. The initial assessment of the patient should include attention to medications that cross the blood-brain barrier, or that may interfere with the renal or hepatic metabolism of centrally active drugs. In patients in renal or liver failure, or requiring dialysis, the side effects of a centrally acting drug must be considered in the differential diagnosis for any neurologic symptom. This may occur even in the presence of “normal” doses of drugs or levels of anticonvulsants, as off-target drug toxicities may be due to metabolites, not the drug itself. An example is shown of a child with a mild static encephalopathy and spasticity, on dialysis as the sequela of perinatal renal injury, who developed progressive encephalopathy (Figure 55-1). Treatment with flumazenil with immediate improvement in mental status, confirmed that the accumulation of baclofen in this dialsysis-dependent patient was the source of a waxing and waning encephalopathy. Common examples in the PICU include immunophilin-associated seizures, encephalopathy, hypertension; oculogyric crisis, delirium during reduction of sedation, and persistent paralysis from prolonged neuromuscular blockade. In many cases, the attribution of the symptoms to a drug side effect (excess level, too-rapid withdrawal, idiosyncratic reaction, impairment of clearance) will be a diagnosis of exclusion. Nevertheless, a review of medications and recent changes should be a routine component of the neurologic assessment.

Figure 55–1 Bipolar EEG montage (A) of a child with a mild static encephalopathy and spasticity, on dialysis as the sequela of perinatal renal injury, who developed progressive encephalopathy with high-amplitude epileptiform discharges. Treatment with flumazenil resulted in immediate improvement in mental status and sustained improvement in the EEG (B), confirming that the accumulation of baclofen in this dialsysis-dependent patient was the source of a waxing and waning encephalopathy.

Vital Signs

Therapies that target specific mechanisms of cell injury for acute CNS disease are lacking in adult and pediatric neurocritical care. Nevertheless, one simple intervention, the prevention of hyperthermia, may significantly reduce neurologic injury after stroke or traumatic brain injury (TBI). The neuroprotective benefit of induced hypothermia remains to be proven for pediatric TBI.^15–17 While hypothermia may remain as yet an unproven therapy for neurologic insults, the prevention of hyperthermia can be regarded as a safe, and probably effective, intervention.¹⁸ Therefore, as a general principal for ICU neurology in patients with known or suspected central nervous system (CNS) injury, the neurologic exam should first include vital signs and fever should be treated aggressively. The range of temperatures determined to be “normothermic” may vary but a consensus should be reached and temperatures above the accepted range treated quickly.

In contrast to adults, there are limited data on the optimum ranges for blood pressure, ICP, and cerebral perfusion pressure (CPP) for children. In adults, there are goal-directed protocols aimed at improving outcome following TBI. The targets of ICP less than 20 mm Hg and CPP above 60 mm Hg are based on a number of studies.^19–23 In contrast, very limited data exist for children with TBI.^24,25 Guidelines suggest an ICP threshold of 20 mm Hg for all ages and a CPP threshold of 40 to 65 mm Hg as an “age-related continuum.”²⁶ In children younger than 15 years of age, a mean CPP less than 40 mm Hg during the first 48 hours following injury has been associated with increased mortality.²⁷ Although children with TBI younger than 2 years of age have a high mortality rate compared with older children, no thresholds for ICP or CPP have been established for this age group.²⁸ Accordingly, blood pressure, temperatures, and oxygenation must be interpreted in the context of the underlying neurologic insult or risk for further injury. As a general principle, hyperthermia with the attendant increase in cerebral metabolic demand should be treated aggressively whatever the primary injury.²⁹

General Physical Exam

An essential component of the neurologic examination is obtaining an accurate head circumference. This may serve as a baseline for following the development of hydrocephalus in the at-risk infant. In older children, an abnormal (large or small) head circumference may be a previously overlooked sign of pathology. The fontanelle should be palpated and the findings on the exam agreed upon with the infant in a quiet, resting state. A bulging fontantelle is an important finding of meningeal irritation or increased ICP. Examination of skin for the cardinal features of the phakomatoses may identify café au lait spots characteristic of neurofibromatosis, or shagreen patches, hypopigmented macules, and angiomyofibromas characteristic of tuberous sclerosis. The physical findings associated with inflicted trauma deserve particular mention as these may be subtle³⁰; they include unusual patterns of bruising, blood in the oropharynx, and burn or belt marks.

Importance of Observation in the Neurologic Exam

The elements of the neurologic examination in the PICU comprise the same features of the neurologic exam as for patients who are not critically ill. The assessment of mental status, cranial nerves, motor function, reflexes, sensation, and cerebellar function must be adapted to each patient, but the structure of the neurologic exam in the PICU is no different from the outpatient examination. If possible, sedating drugs and paralytic agents should be reduced prior to the exam. If this is not feasible, the exam must be interpreted in the context of these confounds.

The assessment of mental status begins with observation. The examiner should first confirm the drugs used for sedation if any, and recent changes in dosing. It is not possible to give precise dose ranges for typical sedating agents associated with a mental status exam. For commonly used agents in the PICU (midazolam, morphine, fentanyl, dexmedetomidine) the effects of the drug on arousal and responsiveness will vary with the age of the patient, duration of exposure, nature of the neurologic insult, effect of other drugs on metabolism, and genetically determined ability to clear these drugs. This complex set of interactions underscores the importance of experience in the examination of these patients to determine what is an acceptable level of arousal, as well as the need for serial neurologic examinations.

The observations of the nurses and parents should first be solicited. The examiner should enquire about evidence for changes in arousal or awareness, such as spontaneous eye-opening, evidence of a sleep-wake cycle, change in activity, or response to interventions such as suctioning. If family members are present, their observations of response to their presence or voice are important and may represent the first signs of awareness on the part of the patient. It is quite appropriate to have a family member carry out part of the exam, asking the patient to follow commands, as young children are more likely to respond to their family than to a stranger.

The pattern of breathing rate and rhythm should be observed. Specific patterns may help localize the site of neurologic dysfunction but not the mechanism involved (Table 55-1). The crescendo-decrescendo pattern alternating with periods of apnea, characteristic of Cheyne-Stokes respiration, may be due to dysfunction of the cerebral hemispheres, the thalamus, or the hypothalamus with preserved brainstem function, but can also be found in patients with congestive heart failure or primary respiratory disease. Similarly, the sustained, deep breathing pattern of central neurogenic hyperventilation may be due to either structural injury to the midbrain, sepsis, pulmonary disease, or compensated metabolic acidosis.

Table 55–1 Approach to the ICU Neurologic Examination: Localization and Mechanism

The conventional neurologic examination proceeds from mental status, cranial nerves, motor (bulk, tone, and strength), reflexes, and sensation to cerebellar exam and gait. In children and in the PICU in particular, the exam must be based on good observation and typically performed out of sequence. It is easier to assess tone and reflexes in the asleep, relaxed patient, before waking the patient up to assess mental status. While the exam is discussed in the standard order below, it is often more informative to begin the hands-on part of the exam with assessment of tone and reflexes. Once awake and agitated, it may be difficult to elucidate subtle asymmetries of tone and reflexes, which may be key findings of the exam.

The state of arousal (awake, asleep), responsiveness, (interactive, verbal, nonresponsive), position (tone and asymmetry of limb position), movement (purposeful, spontaneous, dystonic, choreic, asymmetry of movement) can all be reliably assessed by observation. In a nonintubated, nonsedated patient the examiner may proceed directly with a standard mental status exam adjusted for age. In the young child who is able to cooperate, the specific cognitive and language skills expected for age can be assessed (Table 55-2).

Table 55–2 Age-Dependent Motor and Language Patterns

Nonorganic Pathology in the Pediatric Intensive Care Unit

Not all neurologic deficits in the PICU are organic. Conversion disorders also occur in critically ill patients or result in patients being admitted to the PICU. In the latter case, this is most often associated with nonepileptiform seizures. These cases may have been treated with anticonvulsants, resulting in respiratory depression requiring intubation or ICU monitoring. A number of features of the examination and history may help make this distinction (Table 55-3). Seizures are a common complication of critical illness in adults and children³¹ and are also a common reason for neurologic consultation in the PICU. A high index of suspicion for seizures in the PICU is important in order to decrease the risk for secondary brain injury resulting from the metabolic stress imposed by seizures. In such cases, video EEG monitoring is the most efficient means of distinguishing ictal from nonictal events. While continuous EEG monitoring may be needed to determine whether such events are seizures or not, there are features of the physical exam that may aid in making this diagnosis. In general, and particularly in infants, the eyes are open during a seizure. In one series,³² over 90% of cases with electrographically confirmed seizures occurred with eye opening. The eyes may look straight ahead, versively deviate to one side contralateral to the hemisphere from which the seizures originate, or exhibit only nystagmus. Seizures are typically a “positive” phenomenon and will have movement associated with them unless there has been injury to the corticospinal tracts resulting in a paretic limb. Other features of the exam or description of the spells including stereotypy, crescendo-decrescendo behavior, or the presence of automatisms may help to distinguish ictal and nonictal behavior. In most cases, video EEG monitoring is the most efficient means of distinguishing ictal from nonictal events.

Table 55–3 Clinical Features of Seizures and Nonepileptic Spells

Even in the PICU, the examiner should remember that functional deficits may occur, and a number of motor signs may help identify the origin of functional symptoms. Given the complex pathophysiology of many critically ill patients and the multiple potential mechanisms of neurologic injury, the diagnosis of a functional neurologic deficit should be made only after a thorough evaluation for organic causes. Although this is uncommon in the PICU, a functional etiology should not be overlooked and early identification may help to minimize diagnostic testing and unnecessary or high-risk treatment. There are a number of elements of the physical exam that may help distinguish between organic and functional neurologic deficits.³³ An inconsistent examination may be the first clue; for example, a patient with apparent weakness during bedside strength testing may be noted to change position during sleep or another may walk despite apparent paresis on bedside testing. These observations can often be very helpful in the assessment of children with weakness thought to be due to a neuromuscular disease. After first observing for inconsistency, the most useful test for functional weakness is the Hoover sign. This test relies on the principle that when flexing one’s hip, the natural accompanying movement is to extend the contralateral hip.³³ With the patient supine, the examiner places one hand on the weak leg and the other hand under the ankle of the strong leg. The patient is then asked to perform a straight-leg raise of the healthy limb. In a functional pattern of weakness, the examiner will feel no downward pressure from the good leg since there is no effort being applied to raise the ostensibly weak leg. Other tests include the “arm-drop” in which a paretic or plegic arm is dropped over the patient’s face (the patient with organic weakness is more likely to allow the arm to strike the face), or observation of exact splitting of sensory or vibration deficits in the midline. A number of functional gait disturbances are also characteristic, including a monoplegic dragging gait (the whole limb is dragged without the circumduction present in pyramidal hemiparesis), a “walking on ice” pattern, excessive slowness, or sudden buckling at the knees with recovery. Of these, none are definitively diagnostic of a functional pathology, and in children in particular must be interpreted with caution. Importantly, functional and organic deficits may coexist.

Assessment of Level of Consciousness and Mental Status

For children with depressed consciousness, the precise stimulus required to elicit a response and the nature of this response should be specified. It is not helpful to describe the patient as “lethargic” or “obtunded.” First, the child is called by name to determine whether there is a response. If this is not effective, the examiner may ask a family member to speak to the child. Next, the stimulus is increased. It is most helpful to simply describe the patient’s response to specific stimuli and to use the same stimuli for serial examinations. The Glasgow Coma Scale score, although of limited use in preverbal children, is a quantitative measure (Table 55-4). If the patient’s eyes open to voice, the response to commands is next tested. This is discussed in more detail below, but the commands should be increased in complexity from one-step to two- or three-step. If there is no response to voice, a painful stimulus is next applied. It should be noted that in patients with sensory deficits due to neuropathy, spinal cord, or CNS lesions or with focal limb weakness, the extremity or dermatome selected for testing should have intact sensory or motor function. In the sedated or severely impaired patient the response may comprise only increase in heart rate. It is important that the stimulus and the specific response, rather than vague descriptors (“lethargic, drowsy, sleepy”) are documented, as this will be more helpful in assessing serial examinations. The rate of sedating drugs or recent administration of sedating drugs should be documented with this exam.

Table 55–4 Glasgow Coma Scale

The assessment of higher cognitive function, and the early recognition of compromise of cognitive function, is a challenge at any time when evaluating young children. In the ICU, where the additional confounds of sedation, other organ dysfunction, sleep disturbance, and anxiety must be accounted for, this evaluation is somewhat more challenging. Of course, this is usually the most important component of the exam and the technologies used for neurologic monitoring in the ICU (imaging, EEG, NIRS, microdialysis, tissue oxygenation) all serve the same goal of the bedside exam, of detecting compromise of cerebral function to enable directed therapeutic intervention.

In the older, awake child a complete mental status exam can be performed. This must be adjusted for age (see Table 55-2) but should include assessment of language (fluency and comprehension): the ability to name, repeat, write, read, and respond to written commands. Simple mathematical problems should be adjusted to the child’s age-dependent ability. Praxis can be assessed quickly even in the PICU by demonstration of learned behaviors, even in the young child (brush your teeth, brush your hair). Other components of the mental status exam, including memory, fund of knowledge, and reasoning can be assessed by holding a conversation with the patient.

Most importantly, the examiner needs to have an appropriate index of suspicion for these subtle neurologic deficits of attention, memory, praxis, language comprehension, and reading comprehension which may be the early sign of new neurologic injury. This is particularly true for patients with metabolic (often liver or renal failure), infectious, or iatrogenic (sedation, immunophilins) risk factors for CNS dysfunction and may easily be missed if not specifically investigated.