INTRODUCTIONHigh performing clinical decision rules (CDRs) have been derived to predict which head-injured child requires a computed tomography (CT) of the brain. We set out to evaluate the performance of these rules in the Singapore population.

MATERIALS AND METHODSThis is a prospective observational cohort study of children aged less than 16 who presented to the emergency department (ED) from April 2014 to June 2014 with a history of head injury. Predictor variables used in the Canadian Assessment of Tomography for Childhood Head Injury (CATCH), Children's Head Injury Algorithm for the Prediction of Important Clinical Events (CHALICE) and Pediatric Emergency Care Applied Research Network (PECARN) CDRs were collected. Decisions on CT imaging and disposition were made at the physician's discretion. The performance of the CDRs were assessed and compared to current practices.

RESULTSA total of 1179 children were included in this study. Twelve (1%) CT scans were ordered; 6 (0.5%) of them had positive findings. The application of the CDRs would have resulted in a significant increase in the number of children being subjected to CT (as follows): CATCH 237 (20.1%), CHALICE 282 (23.9%), PECARN high- and intermediate-risk 456 (38.7%), PECARN high-risk only 45 (3.8%). The CDRs demonstrated sensitivities of: CATCH 100% (54.1 to 100), CHALICE 83.3% (35.9 to 99.6), PECARN 100% (54.1 to 100), and specificities of: CATCH 80.3% (77.9 to 82.5), CHALICE 76.4% (73.8 to 78.8), PECARN high- and intermediate-risk 61.6% (58.8 to 64.4) and PECARN high-risk only 96.7% (95.5 to 97.6).

CONCLUSIONThe CDRs demonstrated high accuracy in detecting children with positive CT findings but direct application in areas with low rates of significant traumatic brain injury (TBI) is likely to increase unnecessary CT scans ordered. Clinical observation in most cases may be a better alternative.