Minor blunt head trauma in infants and young children (<2 years): Clinical features and evaluation

INTRODUCTION — This topic will discuss the evaluation of head trauma in infants and children younger than two years of age.

The clinical features and evaluation of minor blunt head trauma in children two years of age and older, the management of minor blunt head trauma in infants and children, and the diagnosis and management of concussion in children are reviewed separately:

●(See "Minor blunt head trauma in children (≥2 years): Clinical features and evaluation".)

●(See "Minor head trauma in infants and children: Management".)

●(See "Concussion in children and adolescents: Clinical manifestations and diagnosis" and "Concussion in children and adolescents: Management".)

RATIONALE — Minor blunt head trauma occurs commonly in infants and is typically not associated with brain injury or long-term sequelae. However, a small number of infants with apparently minor blunt head trauma may have a clinically important traumatic brain injury (ciTBI) that requires extended observation or acute intervention such as intensive supportive care or neurosurgery. Neuroimaging with computed tomography (CT) is highly sensitive for identifying brain injury that requires intervention; however, it exposes patients to radiation. By performing a complete history and physical and utilizing a clinical decision rule, the clinician can identify those infants at higher risk for traumatic brain injury (TBI) while minimizing overuse of CT.

DEFINITIONS

Minor blunt head trauma — Experts define minor blunt head trauma in infants and children younger than two years of age as a history of, or physical signs of, blunt trauma to the head in an infant or child who is alert or awakens to voice or light touch (ie, Glasgow Coma Scale [GCS] score of 14 to 15) [1].

Minor head trauma is generally defined separately in infants and children younger than two years of age for the following reasons [1,2]:

●Clinical assessment is more difficult

●Infants with intracranial injuries are frequently asymptomatic

●Skull fractures or clinically important traumatic brain injuries (ciTBIs) may occur despite minor trauma

●Abusive head trauma occurs more frequently in infants

The clinician should regard these patients as having apparently minor head trauma because approximately 5 percent have traumatic brain injury (TBI) on neuroimaging, and 1 percent have ciTBI that requires prolonged observation/monitoring or acute intervention such as intensive supportive care and/or neurosurgery. (See 'Epidemiology' below.)

Clinically important traumatic brain injury versus traumatic brain injury — For this topic, we use the definition of ciTBI established in large observational studies of children with minor blunt head trauma, which includes any one of the following [3-6]:

●Presence of a TBI (eg, epidural hematoma, subdural hematoma, or cerebral contusion) on CT associated with one or more of the following:

•Neurosurgical intervention (either surgery or invasive intracranial pressure monitoring)

•Endotracheal intubation for the management of head injury

•Hospitalization directly related to the head injury for at least two or more nights

•Death

●Depressed skull fracture warranting operative elevation (ie, depressed past the inner table of the skull)

●Clinical findings of a basilar skull fracture (periorbital ecchymosis, Battle sign, hemotympanum (picture 1), cerebrospinal fluid [CSF] otorrhea, or CSF rhinorrhea)

These same studies also defined TBI (as opposed to ciTBI) as any intracranial injury on CT that does not require acute intervention or prolonged hospitalization as described above [3-6]. Short-term follow-up (four weeks post-injury) indicates that the risk of morbidity or mortality in young children with TBI is very low. However, because there are no long-term studies of infants and children with TBI on CT who do not meet criteria for ciTBI, the long-term implications of these findings are unknown.

Mild traumatic brain injury — The World Health Organization and the Centers for Disease Control and Prevention define mild TBI as an acute TBI that is associated with one or more of the following signs or symptoms [7,8]:

●Confusion or disorientation

●Post-traumatic amnesia for <24 hours

●Loss of consciousness (LOC) for ≤30 minutes

●Transient neurologic abnormalities such as seizures and focal neurologic signs or symptoms

and

●GCS score 13 to 15 (table 1) at 30 minutes or longer after their injury

It is important to note that children younger than two years were not specifically addressed, and several of the signs and symptoms noted are not assessable in this age group due to their developmental stage. Additionally, this definition covers a broad range of patients with highly variable risk for TBI requiring intensive supportive care and/or neurosurgery (ciTBI). For example, the rate of abnormalities on CT in patients with a GCS score of 13 is as high as 20 percent [4,9,10]. By contrast, patients with minor symptoms and a GCS score of 15 have less than 1 percent incidence of ciTBI [9].

PATHOPHYSIOLOGY — The pathophysiology of subdural hematomas, epidural hematomas, and severe traumatic brain injury (TBI) in infants and children are discussed in detail separately. Of note, young infants with open fontanels may develop intracranial hypertension and its associated signs later in the course of a growing hematoma because the intracranial space can expand.

●(See "Intracranial subdural hematoma in children: Epidemiology, anatomy, and pathophysiology".)

●(See "Severe traumatic brain injury (TBI) in children: Initial evaluation and management".)

EPIDEMIOLOGY — Head trauma occurs commonly in infants and young children with approximate annual rates of 1620 emergency department visits, 45 hospitalizations, and 4 deaths per 100,000 population in the United States [11,12]. The vast majority of injuries are due to falls (followed by motor vehicle collisions and pedestrian/bicycle accidents); however, it is important to note that abusive head trauma occurs most frequently in infants younger than two years. Accidental injuries typically cause minor and isolated head trauma in the majority of patients [3]. For infants younger than two years with minor blunt head trauma and a normal neurologic examination, approximately 3 to 10 percent have a traumatic brain injury (TBI) on CT, 1 percent have clinically important traumatic brain injury (ciTBI), and 0.2 percent require neurosurgical intervention [3,4,13-16]. Some infants with TBI have no clinical symptoms of brain injury, although most of these patients have nonfrontal scalp hematomas [3,17,18]. (See "Physical child abuse: Recognition" and "Child abuse: Evaluation and diagnosis of abusive head trauma in infants and children".)

Many clinicians regard infants with ventricular shunts or bleeding disorders as having a higher risk for ciTBI after minor head trauma. Two large cohort studies suggest that this risk is not substantially different from normal infants [19-21]. However, these studies had low numbers of infants younger than two years of age. Thus, until further evidence is available, a more cautious approach to imaging in these patients may still be warranted.

CLINICAL FEATURES

Common findings — Observational studies have identified altered mental status, loss of consciousness (LOC), abnormal behavior per caregiver, scalp hematoma, and vomiting as common features associated with minor head trauma in infants evaluated in emergency departments [22].

Altered mental status — Altered mental status, defined in one large cohort study as Glasgow Coma Scale (GCS) score <15, agitation, lethargy, or not acting normally per the caregiver, has been reported in over 10 percent of infants with minor head trauma presenting for emergency department evaluation and is strongly associated with clinically important traumatic brain injury (ciTBI) [3].

Loss of consciousness — LOC following minor head trauma is reported in 4 to 5 percent of infants <2 years of age undergoing emergency evaluation [3,23]. LOC may be associated with ciTBI depending upon its duration and the presence of other clinical predictors [3,5,23].

●Isolated LOC – For this discussion, isolated LOC refers to LOC without any of the age-related Pediatric Emergency Care Applied Research Network (PECARN) clinical predictors for ciTBI such as severe mechanism of injury, altered mental status, vomiting, headache, or signs of skull fracture (table 2). In such infants, the risk of traumatic brain injury (TBI) in the setting of brief, isolated LOC is low. For example, in one multicenter, prospective cohort study of over 150 infants with isolated LOC after mild head trauma, 0.6 percent had ciTBI and 2.2 percent had TBI noted on CT (any intracranial injury or depressed skull fracture) [23]. Longer duration of LOC was not associated with an increased risk although the sample size was small.

●LOC and at least one additional predictor – The presence of one additional PECARN clinical predictor in patients with LOC markedly increases the change of ciTBI. In the study described above, of almost 350 infants with LOC and at least one additional clinical predictor, 5 percent had ciTBI and 9 percent had TBI on CT [23]. For these patients, longer duration appears to be associated with ciTBI although most patients in this analysis were two years of age or older.

Abnormal behavior — In the prospective PECARN study, which included over 10,500 infants <2 years of age with minor blunt head trauma, abnormal behavior was reported by the routine caregiver in nearly 4 percent of infants; ciTBI occurred in 0.2 percent of these patients [3,24].

Scalp hematoma — Scalp hematomas are an important indicator of potential TBI, especially when they appear in younger infants (eg, <6 months of age), are larger (eg, >3 cm), and are located in the temporal, parietal, or occipital regions [25-28]. As an example, in a multicenter observational study of nearly 3000 children younger than two years of age with an isolated scalp hematoma (no other signs or symptoms of TBI), 50 of 570 children (9 percent) who underwent CT had TBI (eg, subdural or epidural hematomas, cerebral or subarachnoid hemorrhages, or cerebral contusion). Of over 100 patients younger than three months with any scalp hematoma who underwent CT, 21 percent had TBI on CT, confirming that an isolated scalp hematoma is an important indicator of potential brain injury in neonates and young infants. However, ciTBI occurred in only 12 patients (0.4 percent of all patients), none of whom died or required neurosurgery [25]. The infant scalp score, a previously validated clinical score, incorporates age and hematoma characteristics and may help stratify the risk of TBI in young infants (table 3) [26,29] (see 'Isolated scalp hematoma' below). The mechanism of injury is also an important consideration not addressed in this score. (See 'Approach' below.)

Vomiting — At least one episode of vomiting occurs in approximately 15 percent of infants undergoing emergency evaluation following minor head trauma [3,30]. Because vomiting is a potential sign of increased intracranial pressure in infants, we regard it as a potential sign of TBI that may warrant observation after a minor head injury, especially if associated with other clinical predictors. However, isolated vomiting after minor blunt head trauma in infants is not a predictor of ciTBI and TBI as discussed separately [3,4,30,31]. (See "Minor blunt head trauma in children (≥2 years): Clinical features and evaluation", section on 'Vomiting'.)

Seizures — Post-traumatic seizures appear to be associated with a significant risk of TBI and ciTBI after minor blunt head trauma. For example, in a planned secondary analysis of the PECARN head trauma study cohort, of the 332 infants and children with posttraumatic seizures who presented with a GCS score of 15 and underwent a head CT, 20 patients (6 percent) had TBI on CT (eg, any intracranial injury hemorrhage, pneumocephalus, cerebral edema, depressed skull fracture of at least the width of the skull, or traumatic skull diastasis) [32], 19 of whom had other clinical predictors of ciTBI. Of these 20 patients, 10 (3 percent) met criteria for ciTBI (hospitalized for ≥2 days or required neurosurgery); four of these patients were <2 years of age. The frequency of TBI was higher for patients with a lower GCS, longer seizure duration, and longer interval since the traumatic event. Data specifically for children younger than two years were not reported, except that none of the 21 who had a GCS of 15 and no other PECARN criteria for ciTBI had TBI on CT (0 percent, 95% CI 0-16). (See "Minor blunt head trauma in children (≥2 years): Clinical features and evaluation", section on 'Seizures'.)

Skull fractures — Skull fractures occur in up to 10 percent of infants following minor head trauma [27,33,34]. Most skull fractures in this population are linear. Among children with linear skull fractures, 15 to 30 percent have associated intracranial injuries [2,15,18,35-37], and the risk of ciTBI is >4 percent [3].

In infants younger than one year of age, large scalp hematoma size, younger age, and/or nonfrontal location (highest risk with temporal or parietal), suggest a higher incidence of skull fracture (see 'Scalp hematoma' above). By contrast, in one prospective series of 422 children, no infant with a frontal hematoma had an intracranial injury [27].

The evaluation and management of skull fractures in children are discussed in greater detail separately. (See "Skull fractures in children: Clinical manifestations, diagnosis, and management".)

EVALUATION — Features of the history and physical examination, along with selective neuroimaging, generally identify infants who have sustained a brain injury with acute management implications. The priority for the evaluation of infants with apparently minor head trauma is to identify those patients with clinically important traumatic brain injury (ciTBI) while limiting unnecessary neuroimaging.

In infants with minor blunt head trauma, clinical findings can help stratify a patient's risk of ciTBI as high, intermediate, or low, with implications for the need for neuroimaging. Focused evaluation to support these clinical decisions can be facilitated by clinical practice guidelines and implementation of computerized decision support based upon validated clinical prediction rules such as the Pediatric Emergency Care Applied Research Network (PECARN) traumatic brain injury (TBI) rules (table 2) [38-42]. (See 'Approach' below.)

Of great importance is that infants who may have sustained child abuse must be identified, and clinical prediction rules have not been shown to be adequately sensitive [43] nor were they intended to be used for identifying child abuse. Suspicion for child abuse is based upon concerning historical features and recognition of important red flag injuries that are found during careful physical examination (table 4) (see "Physical child abuse: Recognition"). Once child abuse is suspected, the provider should report the suspicion to child protective services and law enforcement and involve a child abuse specialty team, whenever available. (See "Child abuse: Social and medicolegal issues" and "Child abuse: Evaluation and diagnosis of abusive head trauma in infants and children".)

History — The history should include time since the event, mechanism of injury (height and landing surface for falls), loss of consciousness (LOC), vomiting, seizure, mental status (includes level of alertness and irritability), as well as course since the traumatic event. Historical features that may suggest an increased risk of ciTBI include the following [3,5,10,44-49]:

●Caregiver concern that their infant is not acting normally

●Seizure or LOC

●Vomiting

●High-risk mechanism of injury such as a fall from greater than 0.9 m (3 feet), head struck by a high-impact object, significant motor vehicle collision (eg, patient ejection, death of another passenger, or rollover), concern for inflicted injury, or unknown mechanism (which may represent inflicted injury)

●Pre-existing conditions that place the child at risk for intracranial hemorrhage, such as arteriovenous malformation or a bleeding disorder (see 'Epidemiology' above)

Physical examination — The clinician should perform a thorough head and neurological examination (including mental status). Vital signs and evidence of associated extracranial injury, such as neck, abdominal, or extremity tenderness, also warrant careful evaluation.

The presence of the following specific findings on physical examination in an infant with an otherwise nonfocal neurologic examination raises concern for ciTBI [5,44-48]:

●Abnormal mental status (eg, persistent Glasgow Coma Scale [GCS] score ≤14) (table 1)

●Palpable skull defect (depression or irregularity) indicating a likely skull fracture

●Scalp hematoma (especially temporal, parietal, or occipital hematoma in an infant younger than one year of age)

●Bulging anterior fontanelle

●In children with open skull sutures, increased head circumference for age (figure 1 and figure 2)

●Signs of basilar skull fracture such as periorbital ecchymosis (raccoon eyes) (picture 2), posterior auricular ecchymosis (Battle sign), hemotympanum (picture 1), or cerebrospinal fluid [CSF] otorrhea or rhinorrhea

Neuroimaging

Approach — The goal of evaluating minor blunt head trauma in pediatric patients is to identify those infants with ciTBI (by neuroimaging) while limiting unnecessary radiographic imaging and the radiation exposure of CT. Most children with minor head trauma do not need CT to exclude ciTBI. The decision to obtain a head CT must balance the importance of identifying a significant but uncommon ciTBI with the estimated risks of late-onset malignancy associated with radiation exposure from CT. Although only a very small number of patients with ciTBI require neurosurgery, the risk of missing such an injury is immediate and potentially considerable. The risks of radiation-related malignancy are real; however, they are very low and over the lifetime of the child [50,51]. Although preliminary evidence suggests that magnetic resonance imaging (MRI) rather than CT may be feasible, it is not widely available in many settings. (See 'Recommended study' below.)

Evaluation for high-risk findings and the use of a clinical decision rule can provide a balanced approach that stratifies patients by risk for ciTBI (table 5) [3,9] (see 'Clinical decision rules' below):

●High risk: Perform neuroimaging

●Intermediate risk: Observe; neuroimaging in selected patients

●Low risk: No neuroimaging

This approach does not apply to infants for whom there is clinical concern for abusive head trauma. All such patients warrant neuroimaging. (See "Child abuse: Evaluation and diagnosis of abusive head trauma in infants and children".)

A clinical decision analysis of neuroimaging in children <2 years of age with minor blunt head trauma found that, when weighing the risk of missing a ciTBI with the risk of radiation exposure from head CT and subsequent malignancy, no imaging was the best strategy when the probability of ciTBI was about 0.9 percent; CT was the preferred strategy when the probability was >4.8 percent [52].

Clinical decision rules — Clinical decision rules are valuable aids that inform the clinician's approach to neuroimaging. However, they are not intended to fully replace clinical judgment and are intended to assist with rather than completely direct the clinician's approach to neuroimaging of children with minor head trauma.

For infants with minor blunt head trauma and no suspicion for child abuse, we suggest that management decisions, especially the performance of neuroimaging and observation, be guided by the use of the PECARN low-risk clinical decision rules (table 2) rather than other rules because it was derived in the largest cohort and has the best discrimination for ciTBI (table 5) [3,43,53-57].

The PECARN rules consist of six predictors in two age groups (table 2), including more than 10,000 patients younger than two years of age, and were derived and validated in large pediatric cohorts at multiple centers, including pediatric and general emergency departments [3]. The goal of these decision rules is to identify infants and children who are at low risk for ciTBI and do not need neuroimaging. In prospective validation, the sensitivity for detection of ciTBI in infants <2 years was 100 percent (95% CI 86-100 percent), and the negative predictive value was also 100 percent (95% CI 99.7-100 percent) [3]. The prevalence of ciTBI was approximately 1 percent in both the derivation and validation cohorts. In addition to deriving and validating the clinical rules to identify children at low risk for ciTBI, these results support grouping of children with minor head trauma into three risk categories to inform CT decision-making and provide guidance for children who are not at low risk (table 5). (See 'Patient populations' below.)

Subsequently, the PECARN rule has shown high accuracy for determining the risk of ciTBI and maintained consistency in both prospective and retrospective validation studies in a variety of settings [43,53-56,58,59]. For example, in a prospective, multicenter study from Australia and New Zealand among the over 4000 infants and young children <2 years old, the rate of ciTBI for each risk group was: high risk 8.5 percent (95% CI 6-11.6 percent), intermediate risk 0.2 percent (95% CI 0.0-0.6 percent), and low risk 0 percent (95% CI 0.0-0.2 percent) [59].

Of note, because there have been no long-term studies of patients with TBI on CT who did not require prolonged observation or intervention, the importance of this injury (and need for its identification) is unclear.

Successful implementation of the PECARN rules have been described using a variety of methods, including clinician education, clinical pathways, and clinical decision support embedded in the electronic health record [38,40,60-63]. Implementation has generally resulted in high clinician satisfaction, high adherence to the rule, and either a decrease or no significant change to the rate of neuroimaging (primarily head CT), depending upon the baseline imaging rate [9].

In one large prospective study that compared the PECARN rules with two other decision rules (the children's head injury algorithm for the prediction of important clinical events [CHALICE] and Canadian assessment of tomography for childhood head injury [CATCH], which are described below), the PECARN rules had the highest sensitivity and identified all patients with ciTBI while the other rules misclassified some patients with ciTBI as low risk, including one patient that required neurosurgery [64]. Depending upon the setting and how they are implemented, the PECARN rules are also associated with safe but variable decreases in the use of neuroimaging (ie, CT) [9,38]. In addition, cost-effectiveness analysis has found that the use of the PECARN rules, when compared with usual care in settings with a 33 percent baseline rate of CT utilization for minor blunt head trauma, was more beneficial in terms of quality-adjusted life-years lost and cost savings [65]. Comparison of the PECARN decision rules with physician clinical judgment is provided below. (See 'Clinical judgment' below.)

Several other clinical decision rules for children with minor head trauma have been derived [22,66]. Of these, CHALICE (table 6) and CATCH (table 7) were derived in large, heterogeneous cohorts. Of note is that neither clinical decision rule addresses children younger than two years separately:

●CHALICE rule – Unlike PECARN, which sought to identify children at low risk who would not require imaging, the CHALICE rule, derived from a prospective study of 22,772 patients, identified 14 high-risk criteria for ciTBI (table 6) that serve as indicators for head CT. During derivation, the rule had a sensitivity of 98 percent (95% CI 96-100 percent) and a specificity of 87 percent [5]. However, evidence suggests that the CHALICE rule may increase the rate of head CT without improved detection of ciTBI as well as misclassify patients with ciTBI as being low risk [64,67].

The CHALICE rule has been implemented in the United Kingdom [68], and other investigators have suggested use of the CHALICE rule instead of the PECARN rules. For example, in two separate analyses, investigators determined, based upon theoretical estimates, that implementation of the PECARN rule would increase the amount of neuroimaging to a greater extent than the CHALICE rule [22,66]. However, as noted above, actual implementation of the PECARN rule in a variety of different emergency departments resulted in decreases or no significant change in rates of neuroimaging [9,38].

In a decision analysis that compared the CHALICE and PECARN rules and assigned costs based upon the United Kingdom National Health Service experience, CHALICE was the optimal strategy for evaluating children with minor head trauma [69]. However, all decision rules evaluated in this study had similar costs and outcomes. Specifically, the PECARN rule achieved quality-of-life estimates similar to the CHALICE rule while costing about ₤75 (approximately $125 USD) more per patient. This difference in cost is likely within the margin of error for the model.

●CATCH rule – The CATCH rule consists of four high-risk factors and three additional medium-risk factors (table 7) [4]. During derivation in almost 4000 children, including those with a GCS score of 13, it achieved high sensitivity (98 percent for intermediate-risk factors and 100 percent for high-risk factors). However, it would require that 52 percent of patients undergo CT. Subsequently, CATCH-2, which adds vomiting ≥4 episodes to the seven original risk factors validated for CATCH, has achieved higher sensitivity for detecting patients with brain injury and patients requiring neurosurgical intervention but still requires high rates of neuroimaging (>50 percent) and has not been validated [70]. Furthermore, evidence suggests that the PECARN rule has better discrimination for ciTBI in infants with minor blunt head trauma than the CATCH rule while not increasing the baseline rate of neuroimaging [64].

Clinical judgment — Evidence suggests that clinical decision rules improve the detection of ciTBI in infants with minor blunt head trauma compared with clinical judgment alone. For example, in a planned secondary analysis of almost 8500 children from the validation group of the original PECARN study [3], the PECARN rules had higher sensitivity than clinician judgment (100 versus 60 percent, respectively) [71]. The PECARN rules had lower specificity than clinical judgment (54 versus 92 percent), which has caused some concern that use of these rules could theoretically cause an increase in rates of neuroimaging. However, implementation of the PECARN rules in a variety of settings, including those with a low baseline rate of neuroimaging, has not increased subsequent utilization of neuroimaging for infants with minor blunt head trauma [9,38,40,60]. (See 'Clinical decision rules' above.)

In settings where clinician judgment is already documented to be highly accurate (eg, pediatric emergency medicine attendings practicing in high-volume tertiary care pediatric facilities), implementation of the PECARN decision rules may have limited impact on improving detection of ciTBI. For example, in a prospective, single-center observational study that compared physician judgment with clinical decision rules, both clinical judgment and the PECARN rules identified all ciTBIs, including four patients requiring neurosurgical intervention (sensitivity 100 percent, 95% CI 84-100 percent for both) [64]. However, in this study, all physicians were aware of the published PECARN rules and had to document the PECARN predictors for all included patients. Thus, it is likely that the PECARN rules indirectly impacted clinical judgment [9]. Furthermore, since this was a single-center study, it may not be generalizable to other settings, especially those with a low volume of pediatric patients or those not staffed by pediatric emergency medicine specialists.

Patient populations

High risk — We recommend that infants and children younger than two years of age with a high risk for ciTBI according to the PECARN or other clinical decision rules undergo neuroimaging. Based upon observational studies, high-risk findings include any one of the following (table 5) [3-5,9]:

●GCS score of 14 or less (table 1)

●Altered mental status (irritability or somnolence)

●Palpable skull fracture or signs of basilar skull fracture

●Posttraumatic seizure (see 'Seizures' above)

●Definite LOC if longer than a few seconds and associated with other clinical predictors of ciTBI (table 2) (see 'Loss of consciousness' above)

●Bulging fontanelle

Intermediate risk — We suggest that intermediate-risk infants and children younger than two years of age undergo close observation for four to six hours after the injury, with neuroimaging obtained in individuals who do not improve or who develop worsening signs or symptoms during this period (table 5). Some intermediate-risk patients may undergo immediate CT based on the clinical scenario; factors influencing the decision may include age of patient, severity of symptoms, presence of multiple findings, clinical course, physician experience, and shared decision-making.

Intermediate-risk findings for ciTBI include any one of the following:

●LOC that is isolated and brief (less than a few seconds) including unwitnessed fall in adjacent room with brief or uncertain LOC (see 'Loss of consciousness' above and "Minor blunt head trauma in children (≥2 years): Clinical features and evaluation", section on 'Loss of consciousness')

●History of lethargy or irritability, now resolved

●Behavioral change reported by caregiver

●Injury caused by high-risk mechanism (eg, fall >0.9 m [3 feet]; head struck by high-impact object; motor vehicle collision with patient ejection, death of another passenger, or rollover)

●Nonfrontal scalp hematoma

●Age <3 months with nontrivial trauma

●Vomiting

The clinician should have a lower threshold for neuroimaging if symptoms are multiple, worsening, or severe; if vomiting is persistent; and for infants younger than three months of age with nontrivial mechanism of injury [3,44,72]. Physicians performing observation of these patients should have appropriate pediatric experience.

Evidence suggests that observation of selected children younger than two years of age at intermediate risk by PECARN criteria reduces the number of head CTs without missing ciTBI:

●In a planned secondary analysis of a multicenter, prospective, observational study of over 18,000 children with minor blunt head trauma (about 25 percent younger than two years of age) who presented to emergency departments in New Zealand and Australia, planned observation was associated with lower CT use than no planned observation (unadjusted CT rate 4 versus 10 percent, respectively; adjusted odds ratio [OR] 0.2) [73]. Most patients undergoing planned observation were at intermediate risk of ciTBI by PECARN criteria. Among intermediate-risk patients, there was no difference in ciTBI between patients observed (0.5 percent) and those with no planned observation (0.6 percent). Two intermediate-risk patients who underwent planned observation had ciTBI requiring neurosurgery. Among high-risk patients, ciTBI was noted in 1.2 percent of the observed group compared with 9.4 percent of patients with no planned observation. Since decision to observe was the treating clinician's choice, and both intermediate- and high-risk groups are heterogeneous, it is possible that the observed patients may have been less symptomatic.

●In an observational study of 1318 children evaluated for minor head trauma at a single pediatric emergency department, including 509 children younger than two years of age, the rate of head CT among children who were chosen for observation was 5 percent [74,75]. The adjusted time-dependent decrease in CT rate for children at intermediate risk according to the PECARN decision rule was 72 percent, and their median time of observation was three hours. None of the intermediate-risk children were subsequently found to have a ciTBI. A decision aid can increase caregiver knowledge and encourage their involvement in the decision regarding observation versus neuroimaging in such patients [76].

Low risk — For infants <2 years of age at low risk for brain injury, we recommend no neuroimaging. These patients should have no suspicion for child abuse, a normal neurologic examination (including no bulging fontanelle), no history of seizure, and no persistent vomiting. In addition, low-risk patients should meet all of the following criteria (table 2) [3]:

●Normal behavior according to the routine caregiver

●No LOC >5 seconds by history

●No high-risk mechanism of injury (eg, fall >0.9 m [3 feet]; head struck by high-impact object; motor vehicle collision with patient ejection, death of another passenger, or rollover)

●Normal mental status

●No parietal, occipital, or temporal scalp hematoma

●No evidence of skull fracture

●Age ≥3 months

Caretakers of all discharged patients should be given appropriate discharge instructions with reasons to return reviewed. (See "Minor head trauma in infants and children: Management", section on 'Discharge instructions'.)

The approach to neuroimaging in children with suspected abusive head trauma is discussed separately. (See "Child abuse: Evaluation and diagnosis of abusive head trauma in infants and children", section on 'Imaging'.)

Age <3 months — Data for infants <3 months old are limited but suggest that ciTBI and TBI are significantly higher than for older infants, but that the PECARN criteria are helpful in identifying young infants at low risk of ciTBI and TBI [77,78]. For example, among nearly 3000 infants and young children with mild head trauma and isolated scalp hematomas and no other findings, ciTBI occurred in 2.2 percent of those <3 months compared with 0.3 percent of those 3 to 24 months old [25]. In a separate analysis of over 1000 infants <3 months old (59 percent who underwent CT), of over 500 patients who did not meet low-risk PECARN criteria, 4.2 percent had ciTBI and 21 percent had TBI on CT [77]. Among 514 (47.5 percent) infants who met the PECARN low-risk criteria, 1 (0.2 percent, 95% CI 0.005-1.1 percent) had ciTBI and 10/197 (5.1 percent, 95% CI 2.1-9.1 percent) had traumatic brain injuries on CT [77].

Isolated scalp hematoma — For otherwise well-appearing infants <1 year of age whose only criterion for ciTBI or TBI after minor blunt head trauma is an isolated scalp hematoma, we suggest that clinicians use the infant scalp score (table 3) in addition to the PECARN head injury decision rules (table 2) to help determine the need for neuroimaging [26,29].

The infant scalp score is a validated score that stratifies the risk for ciTBI and TBI in infants with isolated scalp hematoma using patient age, hematoma size, and hematoma location. In a secondary analysis of the PECARN TBI dataset, the score was validated in a cohort of almost 1300 such infants younger than one year of age and had high sensitivity for identifying those with ciTBI or TBI [29]. In this cohort, an infant scalp score of ≥4 was 100 percent sensitive for ciTBI and TBI on CT; a score >5 was 100 percent sensitive for ciTBI but missed three infants with less severe TBI on CT.

Delayed presentation (>24 hours after injury) — Infants and children may present more than 24 hours from the time of injury or may return more symptomatic after initial emergency department evaluation and discharge. Data are limited for these patients with delayed presentations because many studies (including those used to derive and validate commonly used decision rules such as PECARN and the Canadian Assessment of Tomography for Childhood Head injury [CATCH]) specifically exclude children with head injury who present more than 24 hours after injury.

However, observational studies suggest that the risk of traumatic brain injury may be significant, and that neuroimaging is often indicated, especially in otherwise well-appearing infants with significant nonfrontal hematomas or signs of skull fracture such as scalp swelling or palpable bony defects [79-81]. For example, in a prospective, observational study of almost 20,000 children with minor head trauma, of the 1000 children who presented >24 hours after minor head injury and had a GCS score of 14 to 15, nearly 4 percent had an abnormal head CT, 0.8 percent had ciTBI, and 0.2 percent required neurosurgery [79]. In this study, nonfrontal scalp hematoma and clinical suspicion for a depressed skull fracture were strongly associated with ciTBI.

In a retrospective study of 76 children <24 months of age who presented with scalp swelling greater than 24 hours following head trauma, 42 percent had isolated linear/nondisplaced skull fracture, and 41 percent had linear/nondisplaced skull fracture with associated extra-axial hematoma (epidural or subdural hematoma) [80]. No patient had neurologic deficits, and none required surgical intervention nor experienced neurologic decline.

Recommended study — Cranial CT can rapidly identify TBIs (enabling appropriate management) and is readily available at most hospitals. MRI of the brain can identify ciTBI and has been shown to be feasible for the evaluation of infants with minor head trauma in some settings [82]. However, it is currently not as available as CT, may lead to delays in diagnosis, and may require use of sedation. If abuse is not a concern, skull radiography is of no added value if a head CT is performed, especially if three-dimensional reconstruction of the skull is part of the CT study.

Other markers of brain injury — Other potential markers for TBI include blood S100B protein and skull radiographs:

●Biomarker detection (S100B protein) — S100B protein is a calcium-binding protein located in glial cells that is released into the bloodstream after trauma. Based upon a meta-analysis of eight studies that evaluated 601 pediatric patients with minor head trauma who also underwent CT of the head, S100B measurement had a high sensitivity (100 percent, 95% CI 98-100 percent) and negative predictive value (100 percent; prevalence of TBI on CT 22 percent) for intracranial lesions on CT [83]. However, S100B protein testing is not widely available, there is no large multicenter study for this population, and further study is needed to establish reference ranges in children.

●Skull fractures on plain radiographs — Skull radiographs can demonstrate fractures, a predictor of TBI, but give no direct information about intracranial injury. Therefore, they have extremely limited indications as the initial imaging study in infants and children younger than two years of age. Rare instances in which skull radiographs may assist the clinician in deciding whether or not to perform CT of the head include the following:

•When the history of trauma is uncertain (eg, skeletal survey in the evaluation of suspected abuse)

•When CT imaging is not available to screen for fractures in selected asymptomatic patients 3 to 24 months of age with concerning scalp hematomas

However, skull radiographs should only be performed if a radiologist with pediatric expertise is available to provide an interpretation because physicians with pediatric emergency expertise may have limited accuracy in correctly identifying skull fractures in young children [84].

If a screening skull radiograph shows a fracture, then a head CT should be performed. If a screening skull radiograph shows no fracture, the risk of a ciTBI may be lower, but the clinician should understand that it may still be present. (See "Skull fractures in children: Clinical manifestations, diagnosis, and management", section on 'Diagnosis and radiologic evaluation' and 'Epidemiology' above.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Increased intracranial pressure and moderate-to-severe traumatic brain injury" and "Society guideline links: Minor head trauma and concussion".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Head injury in babies and children under 2 years (The Basics)")

●Beyond the Basics topic (see "Patient education: Head injury in children and adolescents (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Rationale – Minor blunt head trauma occurs commonly in infants and children younger than two years and is typically not associated with brain injury or long-term sequelae. However, a small number of infants with apparently minor injuries may have a clinically important traumatic brain injury (ciTBI) that requires extended observation or acute intervention such as intensive supportive care or neurosurgery. (See 'Rationale' above and 'Definitions' above.)

●Epidemiology – For infants <2 years of age with minor head trauma and a normal neurologic examination, approximately 3 to 10 percent have a traumatic brain injury (TBI) on CT of the head, 1 percent have ciTBI, and 0.2 percent require neurosurgical intervention. Many patients with ciTBI appear well and may only have nonfrontal hematomas. Infants younger than three months are of particular concern. (See 'Epidemiology' above.)

●Evaluation – The clinician should perform a thorough history and physical examination in infants with minor blunt head trauma with specific attention to clinical findings that should raise concern for ciTBI. In this age group, the clinician should be attentive to any concerning historical or exam features indicative of abuse (table 4). (See 'Evaluation' above.)

•PECARN clinical decision rules – Infants <2 years of age with minor blunt head trauma and no suspicion for child abuse can be stratified as high, intermediate, or low risk for ciTBI based on clinical features. We suggest that evaluation decisions, especially the performance of neuroimaging and observation, be guided by the Pediatric Emergency Care Applied Research Network (PECARN) low-risk clinical decision rules (table 2) rather than other rules. Clinical decision rules are intended to assist with rather than completely direct approach to management. (See 'Clinical decision rules' above and 'Clinical judgment' above.)

•Isolated scalp hematoma – For otherwise well-appearing infants <1 year of age whose only criterion for ciTBI or TBI after minor blunt head trauma is an isolated scalp hematoma, we suggest that clinicians use the infant scalp score (table 3) in addition to the PECARN head injury decision rules to stratify risk and determine the need for neuroimaging. (See 'Isolated scalp hematoma' above.)

•Approach to neuroimaging – The approach to neuroimaging for infants and children younger than two years of age with minor blunt head trauma based upon clinical findings is provided in the table (table 5). Patients at high risk for ciTBI should undergo prompt neuroimaging. Those at intermediate risk may undergo neuroimaging or observation based on the specific clinical scenario, with performance of imaging if persistent, worsening, or new symptoms occur during observation. Individuals at low risk for ciTBI should not undergo neuroimaging. (See 'Approach' above.)

This approach does not apply to infants and children for whom there is clinical concern for abusive head trauma. All such patients warrant neuroimaging. (See "Child abuse: Evaluation and diagnosis of abusive head trauma in infants and children".)

•Recommended study – Both unenhanced CT of the head and MRI of the brain have the necessary sensitivity to identify ciTBI. Head CT can be rapidly obtained in most hospitals. Although MRI has no associated radiation, it is less available in most settings, which can result in delays in care and may require use of sedation due to longer scanning times. (See 'Recommended study' above.)

●Management – Specific management recommendations based on results of the evaluation are addressed separately. (See "Minor head trauma in infants and children: Management".)