Source Document: 1) Kupperman N, et al. Identification of children at very low risk of clinically-important brain injuries after head trauma: a prospective cohort study. Lancet 2009; 374: 1160–70 2) Reilly PL, et al. Patients with head injury who talk and die. Lancet, 1975; 306 (7931): 375 - 377 3) Galbraith S. Misdiagnosis and delayed diagnosis in traumatic intracranial haematoma. BMJ. 1976; 1: 1438. 4) Jennett et al. Severe head injuries in thre countries. J Neurol Neurosurg Psychiatry 1977 40: 291-298 5) Rose J, et al. Avoidable factors contributing to death after head injury. BMJ. 1977; 2 : 615 6) DaceyRG, et al Neurosurgical complications after apparently minor head injury. J Neurosurg. 1986; 65: 203-210. 7) RockswoldGL, et al. Analysis of management in thirty-three closed head iniury patients who "talked and deteriorated." Neurosurgery 1987; 21: 51-55 8) Rosckwold GL, et al. Patients who talk and deteriorate. Ann Emerg Med. 1993; 22(6): 1004-7
Narrative: Minor head injury is an extremely common problem experienced by children everywhere and seen by physicians everywhere.
A small proportion of patients who appear to be neurologically normal will go on to have intracranial bleeding that will typically be treated with neurosurgery, but reliably predicting which children will have these outcomes without over-treating and over-testing has been difficult. This review examines the highest quality data available from a huge, multi-center emergency department study examining the evaluation of children with minor head injuries.(1)
The decision aids that the study presents are excellent, and present precise risk estimates in the setting of children with specific findings. Children who are under age 2 who are acting normally, do not have signs of skull fracture or a scalp hematoma (frontal hematomas are allowed), did not lose consciousness for >5 seconds, and did not have a 'severe mechanism'* have a risk of <1 in 5000 of having a major neurosurgical injury. For those over age 2 if they are neurologically normal, and have no severe mechanism*, severe headache, signs of skull fracture, or vomiting, their risk is <1 in 2000.
This is important information but should not be taken to mean that anyone with these factors is 'high risk'. Even among those children who appear to have a decreased level of consciousness (GCS of 14) or a skull fracture, 23 out of 24 children will not have a clinically important brain injury. Parents and physicians should know this information before moving to expensive radiographic imaging that carries risks and may require sedation to complete.
As with adults, when minor head injury children do have major problems such as neurologic deterioration it typically occurs in the first few hours. Rarely, it occurs beyond this time period, and a substantial body of case series literature suggests that when delayed major problems occur they occur in approximately 20% of bad outcome cases, and they occur after 4-6 hours.(2,3,4,5,6,7,8) In many cases these occur more than 12-24 hours later, making them quite delayed and in some cases detectable neither through a full day of observation nor with an immediate CT scan. Delayed hemorrhages do occur and are not always detectable with immediate imaging or observation.
Caveats: These data are from studies of head injured children being evaluated in ED's. This will not apply to many minor head injuries for whom risk is lower. In addition, enrolled children are in a subset of minor head injuries that is at higher risk than average because of the typical selection bias that comes with study enrollment (study subjects are usually higher risk than those who do not get enrolled in such studies). Therefore these risk estimates are likely to be high when applied to a general minor head injury population of children.
Alternately, the reported cohorts here were evaluated thoroughly by physicians for signs of skull fractures, neurologic abnormalities, and other signs and symptoms of brain injury. Therefore in the absence of this thorough physician evaluation it is not possible to know whether or not a specific patient is appropriately classified as a minor head injury meeting the above criteria.
Unlike data used in our review of adult minor head injury the data from the PECARN study of children that we cite (1) have not been validated. While they have been 'statistically' validated, the decision aid has not been applied to children in the clinical environment to determine how well it performs as a tool in this setting. Therefore while it is a large and excellent dataset, and it may even work better than unstructured physician judgment, this remains untested.
Finally, the above estimates of risk of delayed deterioration are based on extrapolation from case series' of patients, some of whom are pediatric and many of whom are adult, and these data are combined with the data from this large pediatric head injury study. They are not an observed set of outcomes, but rather a statistical speculation about the likelihood of delayed events after an initial 4-6 hours of normal neurologic findings in children.
*severe mechanism: car accident with ejection or fatality or rollover, pedestrian or non-helmeted bicyclist hit by a car, falls of more than 3 feet for <2 y/o and 5 feet for >2 y/o, or head struck by a high-impact object (e.g. baseball bat)
Author: David Newman, MD
Published/Updated: September 19, 2010
If you have suggestions, requests, or questions about a particular NNT review, please send us a message and we’ll try to address it as soon as possible.