External validation of the Haemorrhage Estimate Risk in Oral anticoagulation for Mild head trauma nomogram in Hong Kong

Abstract

Objective: Currently, there is no consensus among international guidelines on the management of patients with mild traumatic brain injuries (MTBIs) who are on direct oral anti-coagulants (DOACs). Overuse of computed tomography (CT) scan of the brain results in unnecessary radiation exposure, longer waits and higher costs. The Haemorrhage Estimate Risk in Oral anticoagulation for Mild head trauma (HERO-M) nomogram has been derived to assess the risk of post-traumatic intracranial haemorrhage (ICH) in this group of patients in the emergency department (ED). This study aimed to externally validate this nomogram in Hong Kong and to identify an actionable cut-off point, which was not defined in the original paper. Methods: This was a single-centre retrospective cohort study that included adult ED patients presenting with MTBI and were on DOACs concurrently from 1 January 2022 to 31 December 2023. MTBI was defined as any closed trauma over the craniofacial region with a Glasgow Coma Scale of 14–15. The discriminatory performance of the HERO-M nomogram was assessed using the area under the receiver operating characteristic curve (AUROC). Diagnostic metrics, including sensitivity and specificity, were evaluated at different cut-off points. Results: Of the 180 patients included in the study, 23 (12.8%) had post-traumatic ICH. The nomogram demonstrated good discriminatory performance in predicting post-traumatic ICH with an AUROC of 0.87 (95% confidence interval [CI], 0.79–0.95). At the cut-off point of ≥5, the sensitivity and specificity for ICH were 0.91 (95% CI 0.7–0.98) and 0.75 (95% CI 0.68–0.82), respectively. Conclusion: The HERO-M nomogram has a high discriminatory performance, good sensitivity and reasonable specificity at the cut-off point of ≥5 in identifying ED patients on DOACs who are at risk of ICH after MTBIs. It may help to supplement the current ED head injury guidelines on CT use.