Intracranial injury is the leading cause of death and disability in children. It can arise after severe, moderate, or minor head injury. Children with minor head injury present the greatest diagnostic dilemma for emergency physicians, as they appear well but a small number will develop intracranial injuries. The question that often arises in the ED is:
To CT or not to CT?
What constitutes a minor head injury?
Children with minor head injury have Glasgow Coma Scale (GCS) scores ≥13 and demonstrate minimal alterations to their mental status. Essentially, they are alert and interactive.
What are the benefits of CT?
CT clearly and accurately identifies intracranial injuries. Early identification of injuries can reduce the morbidity and mortality associated with intracranial injury. CT can also allay parental and provider concerns about missed injuries.
What are the disadvantages of CT?
CT also exposes children to harmful radiation. A single head CT provides between 10 and 20 mGy of radiation to the brain, with higher absorbed doses at younger ages. Extrapolation of data from radiation exposure associated with atomic bombs detonated in World War II suggests that one child develops fatal cancer for every 1,500 one-year-old children undergoing single head CTs.1 Meanwhile, retrospective data from radiation exposure associated with CTs in the 1980s reveals a smaller risk where 2-3 brain CTs triple a child’s risk of brain cancer and 5-10 brain CTs triple a child’s risk of leukemia.2 There is a dose dependent effect with greater amounts of radiation exposure associated with increased cancer risk. In addition, the risk of cancer is greater at younger ages and for females.3
The average charge for a head CT is $996, and the average reimbursement provided to the hospital is $231. These costs are small compared to the costs of other testing or observing patients in the ED. However in aggregate, these costs can become substantial.
Identifying LOW RISK patients (do NOT need a CT)
Clinical decision rules identify patients, who are at extremely low risk of intracranial injuries that require prompt intervention. For these patients the disadvantages of CT outweigh the benefits, and CTs should not be obtained. Three recently published rules show promise for improving clinical decision making following minor head injury:
- Pediatric Emergency Care Applied Research Network (PECARN)4
- Canadian Assessment of Tomography for Childhood Head Injury (CATCH)5
- Children’s Head Injury Algorithm for the Prediction of Important Clinical Events (CHALICE)6
Each rule was rigorously derived in large cohorts; however, it is difficult to compare the rules as they have different inclusion criteria and definitions of their outcomes.7
PECARN is the largest study and is the only rule that has undergone external validation. Therefore, it is the closest to being ready for utilization. PECARN divided patients with minor head injury into high, medium, and low risk criteria, with separate criteria for children <2 and ≥2 years of age. Children with none of the high or medium risk criteria of the rule have <0.1% risk of intracranial injury requiring prompt intervention, as summarized in the PV card below. Low risk patients can be discharged safely with good return precautions.
At first glance, these criteria seem onerous to remember and therefore difficult to apply in everyday practice. However, several of the criteria are likely features that most EM providers already incorporate into their practice; many EM providers routinely obtain CTs on children with abnormal mental status or apparent skull fractures. There are then only 4 other criteria to remember, and 2 of these criteria (LOC and severe mechanism) are present for both children <2 and ≥2 years. Then the only additional criteria that need to be remembered to apply the rule are abnormal behavior to parents or non-frontal scalp hematoma for children <2 years, and vomiting or severe headache for children ≥2 years.
The primary limitation of the PECARN low risk criteria is the small numbers of intracranial injuries identified in the study. This results in wide confidence intervals for the sensitivity of the rule. While the PECARN study was quite large, including over 42,000 children, the incidence of intracranial injury requiring intervention was low, and this limited the precision of their estimates. The reported sensitivity of the rule for intracranial injury requiring prompt intervention was 100% for children <2 years and 97% for children ≥2 years. However, the lower bounds of the 95% confidence intervals for the sensitivities of these criteria were 86% and 89%, respectively. These lower bounds of the 95% confidence interval are much lower than the bounds of sensitivities for other commonly employed clinical decision rules, such as the New Orleans Criteria for head CT (95%), Canadian CT Head Rule (96%), NEXUS C-spine rule (99%), and Canadian Cervical Spine Rule (98%). Providers may consider the sensitivity of PECARN potentially too low to utilize the rule.
Be aware that, unlike other commonly employed rules, the performance of PECARN has not been validated outside of the original derivation hospitals. The impact of the rules on CT acquisition has also not been assessed.
Bottom Line: Nearly all children with minor head injury do not have intracranial injuries, and most children do not require CT. The combination of this low pretest probability coupled with the absence of any of the risk factors from PECARN indicate a child has an extremely low risk of intracranial injury requiring intervention and likely should not undergo CT.
Identifying HIGH RISK patients (DO need a CT)
PECARN found that patients with any of the following were at high risk (4% had intracranial injury requiring prompt intervention):
- GCS score <15
- Altered mental status
- Skull fracture apparent on exam
These patients should undergo CT emergently.
Bottom Line: Similar to adults, children with depressed GCS scores, altered mental status, and skull fractures should undergo emergent CT.
What to do with everyone else, the MEDIUM RISK patients?
The management of patients classified as medium risk criteria by PECARN is more controversial. Most patients presenting to the ED with minor head injury have one of the medium risk criteria, and PECARN does not provide clear guidance for these patients. The rule recommends these patients undergo CT versus ED observation, based on:
- Provider experience
- Number of criteria present
- Worsening of symptoms in ED
- Age <3 months, or
- Parenteral preference
Unfortunately, the medium risk criteria have poor positive predictive values, and CT acquisition on all patients with medium risk criteria might actually increase the frequency of acquisition of negative CTs. In order to avoid this, in our practice we typically obtain CTs on medium risk patients with ≥2 medium risk criteria present or children <3 months with 1 medium risk criteria. We observe all other medium risk patients. If patients worsen during observation, we obtain CTs.
Bottom Line: Observation should be considered for any patients that are not classified as high or low risk by PECARN.
Are there any downsides to observation?
Observation is associated with reduced frequency of CT acquisition.8 The primary potential downside of observation is that it could lead to a delay in the diagnosis of an intracranial injury. It is unclear if such a delay would be clinically meaningful. In addition, children undergoing observation have longer lengths of stay in the ED compared to other children.9 These disadvantages of observation are likely less than the disadvantages of CT (radiation, cost) in medium or low risk patients.
How long should a child with minor head injury be observed in the ED?
Again there are no clear guidelines about the duration of observation for children with minor head injury. A retrospective review of minor head injury in Canada found only 0.03% of children who were awake and alert with normal neurologic examinations 6 hours after minor head injury had intracranial injuries.10 We determine the duration of ED observation based on the time elapsed since the initial injury, the time of day, and our confidence in the reliability of the family to monitor the child’s condition at home. Observation following a negative head CT is extremely low yield and should not be performed routinely.11