33 

When Is It Safe to Extubate a Newly Quadriplegic Patient? When Is Early Tracheostomy Appropriate?

Ann-Christine Duhaime

BRIEF ANSWER

Background

Inflicted head injury is the most common cause of trauma-related death during infancy. It results in at least 1000 deaths/year in the United States, more than the annual mortality of all pediatric brain tumors combined.^1,2 Furthermore, it has been theorized that the number of fatalities is far exceeded by the number of children who are permanently injured or disabled by this type of trauma (class III data).³ The risk that a child will sustain a nonaccidental injury during the first year of life has been estimated to be one in 4065, or 0.025% (class III data).⁴ Countless families are affected in profound ways by proven or suspected nonaccidental injury. Both false-negative and false-positive diagnoses can have disastrous consequences. Often, the diagnosis rests on the word of the neurosurgeon, and because of its uncertainty and potential implications, many clinicians approach this subject with reluctance.

An evidence-based approach to the diagnosis, management, and outcome of inflicted head injuries in infancy is problematic. Child abuse is by nature cloaked in guesswork and judgment because the facts on which the clinician generally relies for an accurate evaluation of a problem are usually missing. Also, in these cases, the patients cannot speak for themselves. Thus, the diagnosis must be made by inference. Management options and prognostic guidelines are similarly lacking in firm data on which to base decisions.

Despite these difficulties, some data are available to help guide the clinician, and ongoing efforts to collect more solid data may add to the degree of certainty with which opinions are rendered and decisions made. This chapter reviews some of the available studies that are relevant to making the diagnosis of inflicted injury in infants.

Almost all series of this type are retrospective, and the vast majority suffer to a greater or lesser degree from poorly defined inclusion criteria for the diagnosis of abuse (class III data).3,5–7 Thus, most of this information falls into the category of class III data and must be viewed accordingly. Nonetheless, certain common characteristics can be found in most descriptions of inflicted injury, as outlined below.

Patient Characteristics

The vast majority of patients with inflicted head injuries of infancy and early childhood are under 3 years of age, with most being under 1 year (class III data).^2,8,9 In series of children under 2 years of age, 24 to 32% of patients admitted because of head injury have inflicted injuries (class II and class III data).^8,10

History And Presentation

The most commonly provided histories for patients with inflicted head injuries describe a fall from a short height or no history of trauma at all. Commonly described symptoms include lethargy, irritability, vomiting, seizure, breathing difficulty, or unresponsiveness (class III data).^6,7,11 One half of the patients in one series exhibited impairment of mental status at presentation (class III data).¹¹ Seizures have been reported in 40 to 71% of patients, and apnea in ~50% (class III data).^6,12

Intracranial Injury Characteristics

Classic intracranial findings associated with inflicted head injury in infancy include acute subdural hemorrhage with or without unilateral or bilateral diffuse brain swelling. The hemorrhage may be thin and may be located over one or both cerebral convexities and may also occur in the posterior interhemispheric space. Larger acute subdural hemorrhages that require surgical evacuation are sometimes seen, particularly in older children. Subarachnoid blood and diffuse axonal injury have also been reported. One of the striking features of this syndrome is the association of holohemispheric or bilateral supratentorial loss of gray-white differentiation with brain swelling, which is seen in the most severely affected patients. During the acute phase, CT scanning may reveal a “reversal sign”; that is, a diffuse decrease in density of cortical gray and white matter with attenuation of the gray/white interface, or reversal of the usual gray/white densities and relative increase in density of the thalami, brainstem, and cerebellum (class III data).¹³ In survivors, this radiologic picture may progress to one of profound atrophy. In other children, CT scans may show patchy areas of hypodensity that have the appearance of infarcts or other types of pathophysiologic perfusion/metabolic mismatch (class III data),^13–18 an assumption that is supported by the results of neuropathological studies.¹⁹ It should be noted that these findings are suggestive of, but not absolutely specific for, inflicted injuries.

Chronic subdural hematomas have also been associated with inflicted injuries in infancy (class III data).²⁰ However, this pattern is less typical, and its association with a mechanism of inflicted injury rests largely on the presence of additional associated findings suggestive of nonaccidental trauma (class II data).²¹

Associated Findings

Although skull fractures, scalp swelling, subgaleal hemorrhages, andother signs of head impact are reported in variable percentages of infants with nonaccidental injuries, such signs are generally present in the majority of patients. Skull fractures are seen most often in the occipital and/or parietal regions (class III data).7,11 Skeletal surveys and/or bone scans reveal that 30 to 70% of patients sustained extracranial bony injuries, including rib fractures and metaphyseal fractures (class III data).^{15,22 23} Other injuries commonly associated with physical abuse include frenulum tears and pattern bruising or pattern burns, that is, bruises or burns that occur in a pattern suggesting that they may have been caused by identifiable objects.

Retinal hemorrhages have been reported in an average of 75% of patients with inflicted injuries. Their severity may range from severe and bilateral to relatively sparse. Folds, detachments, and other retinal pathology may be seen (class III data).^24,25 Retinal hemorrhages have been reported in accidental trauma, but except in the setting of severe head injury, these tend to be less profuse and are often unilateral (class III data).²⁶