Abstract
BACKGROUND AND PURPOSE: The use of head CT in trauma settings has increased significantly, driven by the need to detect and monitor intracranial hemorrhages. Among intracranial hemorrhage subtypes, epidural hematomas (EDHs) are relatively uncommon but require careful evaluation due to their potential for expansion and the need for surgical intervention. This study aimed to identify risk factors for initial EDH size, subsequent enlargement, and the need for surgical intervention to guide imaging and treatment strategies. MATERIALS AND METHODS: We conducted a retrospective review of 32,401 noncontrast head CT reports from 2 trauma centers (The Johns Hopkins Hospital and Johns Hopkins Bayview Medical Center) between 2018 and 2024. Patients with EDHs were identified using a structured search of radiology reports. Clinical, demographic, and imaging characteristics were analyzed to assess the predictors of EDH enlargement and the need for surgery. Statistical analyses included the χ(2) or Fisher exact test, Mann-Whitney U test, Kruskal-Wallis H test, and logistic regression analysis. RESULTS: Among 91 cases of EDH, a larger initial EDH size was associated with arterial bleeding sources, mixed attenuation, and the spot sign. These same factors, plus a midline shift, predicted the need for initial surgery. No clinical features or comorbidities predicted a larger EDH. Follow-up imaging revealed EDH enlargement in 25/89 cases (28.1%), with SAH as the only significant predictor (OR = 2.60; 95% CI, 1.00-6.77; P = .05). The scans that demonstrated EDH enlargement were performed after a mean of 6.6 (SD 3.3) hours. Ultimately, 25/91 (27.5%) EDHs required surgical intervention; only EDH enlargement was predictive of the need for follow-up surgery after initial observation. CONCLUSIONS: The presence of concurrent SAH was the strongest predictor of EDH enlargement, and radiologists should recommend short-term monitoring of patients with EDH and SAH. Repeat CT at 6-13 hours will detect nearly all cases of EDH enlargement, which may lead to subsequent surgery. Initial large size, midline shift, arterial sources of bleeds, and active bleeding imaging findings correlated with an early surgical intervention. Future multicenter studies are needed to refine risk stratification and optimize imaging follow-up to balance patient safety and health care resource use.