Risk factors for neurosurgical intervention within 48 hours of admission for patients with mild traumatic brain injury and isolated subdural hematoma
Alessandro Orlando,Ripul R Panchal,Lane Mellor,Laxmi Dhakal,David Hamilton,Glenda Quan,Timbre Backen,Jeffrey Gordon,Carlos H Palacio,Justin Kerby,Gina M Berg,Andrew Stewart Levy,Benjamin Rubin,Josef Coresh,David Bar-Or
DOI: https://doi.org/10.3171/2024.5.JNS232476
2024-08-30
Abstract:Objective: The objective was to identify demographic, clinical, and radiographic risk factors for neurosurgical intervention within 48 hours of admission in patients with mild traumatic brain injury and isolated subdural hematoma. Methods: The authors conducted a multicenter retrospective cohort study of all trauma patients admitted to 6 level I/II trauma centers who met the following criteria: admitted between January 1, 2016, and December 31, 2020, age ≥ 18 years, ICD-10 diagnosis code for isolated subdural hematoma, available initial head imaging, initial Glasgow Coma Scale score of 13-15, and arrival at the hospital within 48 hours of injury. Patients were excluded for skull fracture, non-subdural hematoma, and absence of neurosurgical consultation. The study outcome was neurosurgical intervention within 48 hours of hospital admission. Multivariable logistic regression with backward selection examined 30 demographic, clinical, and radiographic risk factors for neurosurgery. Results: In total, 1333 patients were included, of whom 117 (8.8%) received a neurosurgical intervention. When only demographic and clinical variables were considered, sex, mechanism of injury, and hours from injury to initial head imaging were significant covariates (area under the receiver operating characteristic curve [AUROC] [95% CI] 0.70 [0.65-0.75]). When only radiographic risk factors were considered, only maximum hemorrhage thickness (in mm) and midline shift (in mm) were independent risk factors for the outcome (AUROC 0.95 [0.92-0.97]). When all demographic, clinical, and radiographic variables were considered together, advanced directive, Injury Severity Score, midline shift, and maximum hemorrhage thickness were identified as significant risk factors for neurosurgical intervention within 48 hours of hospital admission (AUROC 0.95 [0.94-0.97]). Conclusions: In the setting of mild traumatic brain injury with isolated subdural hematoma, radiographic risk factors were shown to be stronger than demographic and clinical variables in understanding future risk of neurosurgical intervention. These final radiographic risk factors should be considered in the creation of future prediction models and used to increase the efficiency of existing management guidelines.