External validation of the emergency CT head score to reduce non-trauma imaging: a multicentre retrospective study

Cerebral CT (CCT) is used in emergency departments (EDs) to assess suspected central nervous system disorders, particularly neurovascular issues. However, the increasing use of CCT raises concerns regarding cost, radiation exposure and efficiency, prompting the development of clinical scoring systems to guide informed decisions and reduce unnecessary imaging. The Emergency CT Head Score (ECHS) is a recent tool designed to streamline CCT prescriptions using four criteria: neurological deficit, acute headache, seizures with altered consciousness and transient neurological disorders. This study aimed to validate the ECHS in a different population and healthcare system in France while also providing a detailed cost analysis.

This multicentre retrospective study, conducted in two EDs in Normandy, included adults (≥18 years) presenting with non-rotatory vertigo, acute headache, altered consciousness, seizures, confusion, dizziness, syncope or presyncope without recent trauma who underwent CCT. Exclusion criteria were suspected stroke/transient ischemic attack (TIA) according to French guidelines, head trauma, brain cancer, hydrocephalus, recent stroke or minors. No follow-up was performed. The ECHS was calculated retrospectively and compared with CT findings.

Of the 600 included patients, the mean age was 62.4±18.7 years and 86 (14.3%) showed abnormalities on CT. The ECHS score demonstrated strong performance, with an area under the curve of 0.894 (95% CI 0.849 to 0.940). Using a threshold of ECHS≥1, we observed no false negatives, resulting in a sensitivity of 100% (95% CI 95.8% to 100%) and negative predictive value of 100% (95 % CI 98.2% to 100%). At this threshold, 48% of CCT scans could have been avoided without adverse outcomes.

The ECHS demonstrated high sensitivity for detecting CT abnormalities in this cohort, suggesting potential to reduce unnecessary scans in patients with a score of 0. However, prospective implementation studies are needed.