Using validated prediction rules to identify children at very low risk of clinically-important traumatic brain injuries (ciTBIs) can reduce the need for CT scans and their resultant radiation exposure. This is the conclusion of an Article published Online First and in an upcoming edition of the Lancet, written by Prof Nathan Kuppermann, Departments of Emergency Medicine and Pediatrics, University of California, Davis Medical Center, Sacramento, CA, USA and colleagues from the PECARN investigators network.
Traumatic brain injury is a leading cause of death and disability in children worldwide. In the USA, head trauma in individuals aged 18 years and younger results in about 7400 deaths, over 60 000 hospital admissions, and over 600 000 emergency department visits every year. A CT scan, while a vital diagnostic tool, elevates future risk of radiation-induced cancer. The authors aimed to identify children at very low risk of ciTBI for whom CT might be both unnecessary and undesirable.
The study analysed more than 42,000 children, of whom a quarter were aged under 2 years, with the rest aged 2- 18 years. The prediction rules were created using a 'derivation'* population, and then applied to see if they worked on a 'validation'* population. CT scans were obtained on 35% of the children, ciTBIs occurred in 376 (1%) and 60 (0.1%) underwent neurosurgery. In the validation population, the clinical characteristics used to predict that children younger than 2 years did not have a ciTBI were: normal mental status, no scalp haematoma (swelling) except frontal, no loss of consciousness or loss of consciousness of less than five seconds, non-severe injury mechanism, no palpable skull fracture, and acting normally according to the parents. This correctly predicted 100% of 1176 patients younger than 2 years who did not have a ciTBI in the validation population. 24% of the CT-imaged children younger than 2 years were in this low-risk group.
The prediction rule to identify children 2 years and older who did not have ciTBI included the characteristics: normal mental status, no loss of consciousness, no vomiting, non-severe injury mechanism, no signs of basilar skull fracture, and no severe headache. This correctly predicted all but 2 out of 3,800 patients (99.95%) who did not have a ciTBI in the validation population. 20% of the CT-scanned patients 2 years and older were in this low-risk group. Thus, the results showed that, using these prediction rules for children presenting with head trauma, 1 in 4 children younger than 2 years and 1 in 5 older than 2 years who would likely have had CT scans without prediction rules could avoid these CT scans and their accompanying radiation exposure when the rules are applied.
The authors conclude: "In this study of more than 42 000 children with minor blunt head trauma, we derived and validated highly accurate prediction rules for children at very low risk of ciTBIs for whom CT scans should be avoided. Application of these rules could limit CT use, protecting children from unnecessary radiation risks. Furthermore, these rules provide the necessary data to assist clinicians and families in CT decision making after head trauma."
In an accompanying Comment, Dr Patricia C. Parkin, Hospital for Sick Children, University of Toronto Faculty of Medicine, Canada, and Dr Jonathon L. Maguire, Hospital for Sick Children, University of Toronto Faculty of Medicine, Canada, and Dr Li Ka Shing, Knowledge Institute of St Michael's Hospital, Toronto, Canada, say: "Decision aids might provide structured presentations of options and outcomes, and many decisions, even in acute care settings, are sensitive to patients' values and preferences...Perhaps as this field moves forward to assessment of the effect of the rules on physicians' behaviour and clinical outcomes (impact analysis), clinicians and investigators might consider involving patients in the decision-making process. Then, when asking the question 'should my head-injured child have a CT scan?', parents can weigh the probability of a clinically important traumatic brain injury with the probability of harm from ionising radiation from the CT scan."
Prof Nathan Kuppermann, Departments of Emergency Medicine and Pediatrics, University of California, Davis Medical Center, Sacramento, CA, USA. T)+ 1 916-734-1535 E) nkuppermann@ucdavis.edu Alternative contact Charles Casey, Office of Public Affairs T) +1 916 734-9048 E) charles.casey@ucdmc.ucdavis.edu
Dr Patricia C. Parkin, Hospital for Sick Children, University of Toronto Faculty of Medicine, Canada. T) +1 416-813-6933 E) patricia.parkin@sickkids.ca
For full Article and Comment, see: http://press.thelancet.com/childct.pdf
Notes to editors: *When an investigator wants to create a prediction rule, s/he should do so on a large group of patients to get as precise a rule as possible. This group of patients is called a "derivation group" on which the prediction rule is "derived". However, any group of patients has certain characteristics and idiosyncrasies such that the prediction rule derived on that group may only be accurate when applied to that unique group. The investigator should therefore then test the prediction rule derived on the "derivation group" on a new group of patients, called the "validation group" to assess the validity of the rule.
Journal
The Lancet