Aim Cranial CT is the gold standard for the investigation of intracranial emergencies. The aim of this pilot study was to audit whether senior emergency physicians were able to report CT head scans accurately and reliably having attended structured teaching.

Materials and methods Senior emergency physicians attended a 3 h teaching session. Following this, they independently reported adult CT head scans between 22:00 and 08:00 using a pro forma. CT head examinations performed in this ‘out of hours’ period were formally reported by a consultant radiologist on the following morning. Data were collected in a blinded fashion over an 8-month period.

Results 405 adult CT head examinations were performed. 360 pro formas were available for analysis, and the rest were excluded either because a consultant radiologist had been rung to discuss the results (five patients) or because the pro forma was not completed (40 patients). Concordance between consultant radiologists and emergency physicians was found in 339 (94%) of the cases (κ coefficient 0.78). None of the discordant cases was managed inappropriately or had an adverse clinical outcome. All cases of extradural, subdural and subarachnoid haemorrhage were detected by emergency physicians.

Conclusions In conclusion, we feel that this model can be employed as a safe and long-term alternative provided that the radiology department are committed to providing ongoing teaching and that a database is maintained to highlight problem areas. Emergency physicians need to remember that the clinical status of the patient must never be ignored, irrespective of their CT head findings.