Aims To examine evidence comparing the use of computerised tomography and ultrasound in detecting paediatric skull fractures when children present with a head injury.

Methods Medline library searched using the Pubmed interface – search terms [(head OR skull) AND (fracture OR injury) AND ultrasound AND CT AND (infant or child*)]. All relevant publications were then hand searched. 5 relevant articles were identified.

Results Skull fractures in children raise the relative risk of intracranial injury. Ultrasound has a well-established evidence base in the diagnosis of fractures, including those of the nose and zygomatic arch. It allows rapid, bedside assessment, with appropriate escalation as necessary. It can be performed in the emergency department, when skilled staff are available, allowing observation and monitoring to continue in a safe environment. It does not involve exposure to ionising radiation and removes the need for sedation during the high risk post-injury period.

The cohort studies reported here, from the USA and Italy, do not outline the extent of the local expertise and practice in the use of paediatric emergency ultrasound. None of the studies attempt to look at the skills of individual emergency physicians in terms of their training and experience, or their skill in both acquiring and interpreting the images. Inter-operator agreement was not examined, and as single centre studies it was not possible to look at ultrasound use and detection rate, between centres. These studies reported that there were cases when skull fracture had been diagnosed by CT scan prior to performing the ultrasound investigation, with clear implications for potential bias.

The studies conclude that ultrasound is both a specific, and to a lesser extent, sensitive tool in the identification of skull fractures in children presenting with scalp haematoma following head injury.

Conclusion Bedside emergency ultrasound scanning can accurately diagnose cranial fractures in children presenting with head injury with visible scalp injuries, when compared to computerised tomography. There is a need for further research to examine operator dependent factors, in terms of both acquiring and interpreting the images, and in how identifying a skull fracture in asymptomatic children should influence further investigation and management.