Small traumatic intracranial hemorrhages identified in routine radiology reports are associated with a low risk of adverse events: a retrospective cohort study

PURPOSE: To determine whether routinely available radiology reports, together with basic clinical data, can identify patients with traumatic intracranial hemorrhage (TICH) who are at low risk of adverse events. METHODS: This retrospective cohort study of adults with TICH in Region Jönköping, Sweden (2019-2021). Clinical data, findings from radiology reports and outcomes were extracted from medical records. Hemorrhage size was classified as small (≤ 4 mm or described as minimal/very small/discrete) or larger. Adverse events were defined as neurosurgical intervention or death directly attributable to the TICH. Risk difference (RD), relative risk (RR), and Firth's penalized logistic regression were used to assess associations with adverse events. RESULTS: Among 527 included patients, 195 (37%) had small TICHs. None of these patients experienced adverse events, compared with 13.6% neurosurgical interventions and 13.0% trauma-related deaths in the group with larger TICHs (RD 24.5% points, 95% CI 18.4-29.6; RR 97.1, 95% CI 6.1-1557.1; p < 0.001). Small TICH size had the strongest association with absence of adverse events. Normal neurological status and GCS 14-15 were also associated with a low risk of adverse events. Anticoagulant or antiplatelet therapy showed no significant association with adverse events. CONCLUSION: Routinely available radiology reports, combined with basic clinical data, can identify a low-risk subgroup of patients with small TICHs. Hemorrhage size appears to be a useful factor for risk stratification, but the findings require internal and external prospective validation before implementation in clinical practice.