Background National Institute for Health and Care Excellence guidelines used to triage patients with head injury to CT imaging are based on research conducted in populations presenting within 24 h of injury.

We aim to compare guideline use, and outcomes, in patients with head injury that undergo CT imaging presenting within, and after 24 h of injury.

Methods ED trauma CT head scan requests over a period of 6 months were matched to ED records. Case note review of adult patients with head injury that had undergone CT imaging was completed. Logistic regression was used to assess whether presentation after 24 h affected the guideline's ability to predict significant injuries.

Results 650 case records were available for analysis. 8.6% (56/650) showed a traumatic abnormality, 1.5% (10/650) required neurosurgery or died. 15.5% (101/650) of CT scans were for patients presenting after 24 h. 8.4% (46/549) of those presenting within, and 9.9% (10/101) of those presenting after 24 h had traumatic CT abnormalities.

The sensitivity of the guidelines for intracranial injuries was 98% (95% CI 87.0% to 99.9%) in those presenting within 24 h and 70% (95% CI 35.4% to 91.9%) in those presenting after 24 h of injury. The presence of a guideline indication statistically predicted significant injury, and this was unaffected by time of presentation.

Conclusions Patients with head injury presenting after 24 h of injury are a clinically significant population. Existing guidelines appear to predict traumatic CT abnormalities irrespective of timing of presentation. However, their sole use in patients presenting after 24 h may result in significant injuries not being identified.