Objectives and Backgrounds 500 000 children present to UK hospitals annually following minor head injury (MHI). Mortality is low (6–10 per 100 000), and most patients can be safely discharged from emergency departments. However, a small proportion of children with significant intracranial pathology are at risk for deterioration. CT is the gold-standard for investigation of MHI, but resource limitations, and risks from ionising radiation and pharmacological sedation, preclude its universal use. Clinical decision rules (CDR) were recently developed in a North American population by the PECARN research group, to guide CT use in childhood MHI. Our aim was to externally validate the rule for children over 5 years in a UK setting.

Methods The CHALICE database consists of all children (<16 years) presenting to the emergency departments of 10 northwest England hospitals with head injury between February 2000 and August 2002. A dataset of demographic, clinical, radiological and outcome variables were recorded prospectively, and clinically significant head injuries identified. CHALICE patients >5 years were categorised according to PECARN CDR predictors and outcomes (clinically important head injury [ciHI]: death from HI, neurosurgery, intubation >24 h, hospital admission >2 nights with positive CT head). We then calculated the sensitivity, specificity, and negative predictive value for identifying ciHI in the CHALICE population using the PECARN CDR.

Results The CHLAICE database consists of 22 772 children, with 65% males, and a mean age of 5.7 years. Following application of PECARN exclusion criteria a study population of 10 415 children aged >5 years were available for study. 2778 children met PECARN CT scanning criteria, and 246 had ciHI. Information for assessing severity of injury mechanism was incomplete (patient ejection and vehicle rollover variables absent). Data were present for all other PECARN predictors and outcomes. The PECARN rule demonstrated a sensitivity of 95% (95% CI 91 to 97), specificity of 75% (95% CI 74 to 76), and negative predictive value of 99.8% (95% CI 99.7 to 99.9) for the prediction of ciHI.

Conclusions The PECARN clinical prediction rule for children >5 years performed well in a large UK cohort of children. Further work will examine refinements to the PECARN rule, and validate the CDR for children <2 years.