Introduction and background

Pediatric head trauma is one of the commonest presentations in emergency departments accounting for 20,000 visits per year in Canadian hospitals and close to half a million emergency department (ED) visits [1] and 35,000 hospital admissions [2] in the USA. However, less than 1% of patients require neurosurgery [3].

Pediatric head trauma can be classified, according to Glasgow Coma Score (GCS), into mild/minor (GCS 14–15), moderate (GCS 9–13), and severe (GCS <9) [4], bearing in mind the differences in pediatric and adult GCS assessments [2]. Minor head trauma accounts for the overwhelming majority (70–90%) of head injury presentation to ED [5]. Only 4–7% of children with minor head trauma have a brain injury on CT [6]. Clinically important intracranial injuries occur in less than 5% of children presenting to emergency departments with minor head injuries and those requiring neurosurgery account for less than 1% [6, 7]. The challenge lies in identifying which patients with minor head injuries are likely to have clinically important brain injuries. The risk of a fatal traumatic brain injury (TBI) is 0.5 per 1000 [8]. Given most traumatic brain injuries identified on CT imaging do not need intervention studies using abnormal CT as a primary outcome measure may promote excessive CT scanning without altering management [7, 9]. Children with mild TBI account for 40–60% of TBI’s on CT yet are the least likely to need acute surgical intervention [7].

The causes of head injury in children differ to adults. Children are not drivers and are more likely to be passengers or pedestrians in motor vehicle accidents. Alcohol typically does not contribute to head injury in children. Falls and sports-related injury are the most common cause with the risk of intracranial injury being greater if the fall is from greater than 3 ft (or twice the length of the child).

Challenges in pediatric head injury

Children are unique to adults due to their relatively larger heads compared to overall body size, thinner calvarium and softer, less myelinated brain issue all resulting in a higher propensity to develop TBI after head trauma [10]. Children are more likely to suffer with diffuse axonal injury and less likely to develop space occupying hematomas requiring evacuation when compared to adults [10]. Clinical symptoms and signs of head injury, acute concussion, and raised intracranial pressure in children are less reliable and differ according to age and stage of development requiring vigilance in emergency assessment. In older children, symptoms mirror those in adults, namely headache, nausea, vomiting, seizures, lethargy, and drowsiness. In younger children, lethargy, irritability, and anorexia may be pertinent features. Amnesia is difficult to elicit in young children. In infants, hypotonia, listlessness, bulging, tense fontanelles, and sunsetting are features to be wary of.

As yet, no single symptom or sign has been identified as a reliable predictor of the severity of intracranial injury. Children under the age of 2 years and preverbal children can present particular dilemmas. Clinical and neurological assessment is more challenging, given their stage of development and communication and imaging, which requires a static patient, may necessitate sedation. Clinicians must also be particularly vigilant for cues pertaining to non-accidental injury.

Radiation exposure from CT remains a concern given the risk of lethal malignancy, estimated to be between 1 in 5000 and 1 in 1000 cranial CT scans in children [11]. This is risk in increased in younger children, particularly those less than 2 years of age. Younger children can also be challenging to assess which, in the setting of a general, non-pediatric hospital, may lead to a higher use of CT imaging in patient assessment [12]. These children arguably have the most to gain from standardization of management protocols to guide decision-making.

Initial assessment and the use of CT

There is no debate that children presenting with moderate or severe head injuries require computed tomography (CT) brain scans [7] as well as imaging of other body regions according to ATLS guidelines. Less well defined are the indications for brain imaging in those presenting with mild head injuries. The use of CT scans for head injury assessment has risen significantly in recent years [3] and now more than a third of children with minor head injury undergo CT in North America [13]. The concerns regarding sedation, radiation exposure [14], and low pick-up rate of positive findings [11] must be balanced against the unreliability of clinical signs as predictors of brain injury, particularly in preverbal children, and the potential for late deterioration due to missed pathology [15]. Historically, there has been wide variation in the use of CT, up to fourfold in some studies [16]. The trend of rising CT use is not evident in Australian emergency departments where the rate has remained static for a decade, at around 10% [3]. The considerable debate and need for management standardization has, over the past two decades, led to several clinical decision-making tools being devised to determine which patients need CT scans as part of their head injury workup. Many studies highlighting different clinical variables’ predictive values for CT findings are small and unvalidated. None of the initial studies were validated for children and early pediatric guidelines were based on adult data.

The Canadian CT rule group developed a sensitive clinical decision rule for CT in minor head injury, but for adults only [17]. The Children’s Head Injury Algorithm for the Prediction of Important Clinical Events (CHALICE) group in the UK were among the first to derive such a rule specifically for children in order to answer question who should have a CT scan and who can go home [18]. This was based on 23,000 children of all ages and was the first large-scale study of this type. They supported the discontinuation of skull X-rays in children with acute head injury (other than for NAI).

In 2009, the Pediatric Emergency Care Applied Research Network (PECARN) group took this one step further and derived validated rules separately for children younger than and older than 2 years of age on the premise that children under two have a different brain injury risk profile and are more sensitive to the effects of radiation from CT [7]. They concluded that using their six clinical variables 20–25% of CT scans could be avoided.

The Canadian Assessment of Tomography for Childhood Head Injury (CATCH) group prospectively devised a clinical decision rule to identify two levels of risk in children with minor head injury—the risk of requiring neurosurgical intervention and the risk of sustaining a brain injury on CT scan [6].

There are subtle differences between these three studies. CHALICE derived its rules from children of all ages presenting at any time point after the injury and, therefore, more compatible with real-world practice. CATCH and PECARN included patients presenting within 24 h of injury. In Australia and New Zealand, no such clinical rules have been devised [19] and this may contribute to the significantly lower scan rate compared to North America [3].

The three are not directly comparable as they addressed different clinical questions (and hence used different outcomes) assessing different ages of patients and injury severities [19]. Nonetheless, in the first prospective comparison of the three protocols [20], the PECARN guidelines demonstrated 100% sensitivity in identifying patients with clinically important brain injuries. CATCH and CHALICE missed one clinically important brain injury each and five and 14 clinically non-important injuries, respectively. CHALICE provided the greatest specificity of the rules followed by physician judgment. The same study also compared the guidelines to two measures of physician practice and suggested that a combination of PECARN rules and physician judgment provides an adequate combination of sensitivity and specificity though may depend on the experience of the physician. A later comparison with over 20,000 patients [21], a multicenter validation study, showed all three rules to perform well in identifying children with clinically important brain injury. However, it confirmed the findings of Easter et al. [20] that PECARN showed the highest sensitivity in identifying important brain injury and CHALICE demonstrated the highest specificity. Applying PECARN rules strictly would significantly increase the number of CT scans performed to unacceptable levels [22, 23]. The ideal rules must have high sensitivities, i.e., must be able to detect injury, and high negative predictive values, i.e., patients deemed to be low risk should not have significant brain injury. However, given the morbidity associated with missing a significant intracranial lesion, it is difficult to justify a high specificity at the cost of lower sensitivity [19].

Skull X-ray

Fractures on skull X-rays have previously been thought to predict an intracranial injury [24, 25] and, therefore, the need for CT imaging. However, this is less predictable in children, in whom severe intracranial injury can occur in the absence of a fracture, notwithstanding the propensity for inexperienced doctors to miss fractures on X-ray [26]. In 1998, the Society of British Neurological Surgeons recommended skull X-rays be performed in patients with mild or moderate, but not severe, head injuries [27] and this was restricted further in 2003, in the UK by the National Institute of Clinical Excellence, to infants at risk of non-accidental injury and to children in remote areas where access to was problematic [28]. The use of X-rays has continued to decline as the indications for CT have broadened [28]. The more restricted use of X-rays has not been detrimental to children, has not increased admission rates and, although it has increased the use of CT, it has slightly reduced the radiation dose per head injury [29]. Nonetheless, there may remain a role for the use of skull X-rays in the setting of minor head injury in children under 2 years of age [30, 31]. In keeping with this, the Canadian Paediatric Society supports the use of skull X-rays in children in this group with large boggy scalp hematomas [2]. Confirmed or suspected abusive head trauma is a further indication for skull X-ray in children.

Discharge and advice

Canadian Paediatric Society guidelines suggest asymptomatic patients can be discharged home under the supervision of reliable parents or guardians with clearly written instructions describing yellow and red flag signs which warrant further medical attention [2].

Failure to improve warrants admission and a brain CT is recommended after 18–24 h of symptoms. Those less than 2 years old require greater caution with longer observation periods and more frequent clinical assessments.

A further challenge in minor pediatric head trauma is the discharge advice to offer parents, in particular relating to participation in contact sports. The key concern centers on the prevention of second-impact syndrome. Controversial in its existence [32], it occurs when a symptomatic head injury is followed by a second impact while symptoms from the first injury persist [33]. There is debate surrounding the required severity of the second impact (and, indeed, whether or not a direct blow to the head is even necessary), the required time lag between the two impacts and the mechanism(s) involved. Although rare, it can result in catastrophic cerebral edema, disability, and death [33]. One proposed mechanism is a hyperemia leading to vascular engorgement on a background of dysregulation of the parasympathetic system from the first injury. This leads to failure of autoregulation [34], vasodilation, malignant cerebral edema, and critically raised intracranial pressure [35]. Alternative mechanisms include metabolic disturbances secondary to multiple blows, resulting in free radical formation, neuronal damage, and susceptibility to further injury [36]. A recent review identified 17 cases in the literature [33] though up to 36 cases were identified with less strict definitions of the syndrome. Most commonly reported in adolescents and young adults between the ages of 16 and 19 years of age, it is most likely to occur within the first couple of weeks after the first injury. Nonetheless, patients may be at risk as long they are experiencing post-concussive symptoms which guide expert recommendations to avoid contact sports until symptoms have cleared [23]. Risk factors for second-impact syndrome appear to be male gender and younger age and playing American football, but further work is required to clarify this rare, but catastrophic phenomenon [33].

The American Academy of Neurology has published return to play guidelines [37] incorporating these risk factors into their guidance strategies. Similarly, the Canadian Paediatric Society has developed its own guidelines on the management of sports-related concussion including advice on returning to school and sports participation in children after minor head injury [38]. A five stage Graduated Return to Learn protocol is commenced with cognitive rest and absence from school. Gradually increasing cognitive tasks and school attendance ends with the fifth and final stage of the protocol marking the start of the Graduated Return to Play protocol. Thomas et al. argue, in a randomized study, that extended rest (5 days) offers no benefit over shorter (1 to 2 days) rest periods [39]. However, their follow-up was only 10 days and excluded children under 11 years old, those who did not speak English and those who lived several hours away from the study center [39]. Its applicability, therefore, to real-world practice is limited. The Return to Play Protocol is a six-stage process starting with no activity and incremental increases until full, normal play is resumed [38]. The principles include that full academic activity should be resumed prior to commencing exercise and although cognitive exertion can commence before symptom-resolution, it is recommended that sporting activity begins only after all signs and symptoms of concussion have been resolved for 7 to 10 days [38]. Each stage should last a minimum of 24 h and should symptoms recur, the child should rest and recommence the same stage once symptoms have resolved again. These guidelines are based on expert opinion rather than prospective, randomized studies [40] and advocate caution and conservatism in its approach.

Conclusions

Head trauma is a frequent occurrence in childhood and adolescence, with most injuries being minor and without sequelae. A systematic approach to the clinical assessment of the patient presenting with head trauma is recommended, with investigations being conducted according to the symptoms, signs, and estimated likelihood of intracranial pathology. Most patients can be either discharged or observed in hospital for a short period. Prospectively validated clinical decision rules are a useful adjunct in the physician’s armory and should be used to form standardized protocols for head injury management. Given initial management and investigation of children with minor head injuries and the fact that most never need intervention, it raises the question why do we actually admit children after minor injury? Is it mechanism of injury, is it symptom management, is it imaging findings or the likelihood of deterioration and requiring neurosurgery? Only once we really answer this question will we be able to design robust decision trees for all aspects of managing minor head injury in children.