gms | German Medical Science

Objective: For its lack of radiation exposure and its known diagnostic sensitivity cMRI (cranial magnetic resonance imaging) holds potential for becoming a primary diagnostic tool in pediatric traumatic brain injury (TBI). This study investigated the practicability of MRI in the setting of acute pediatric mild TBI and predictive factors evaluated with regard to developing a potential trauma algorithm.

Method: Our institutional electronic database was screened for patients (aged up to 18 years) admitted for mild TBI (between 1/2009 and 8/2015). Following criteria were analyzed retrospectively: Initial GCS (Glasgow Coma Scale), GCS deterioration, loss of consciousness, vomiting, post-traumatic amnesia, presence of confusion or lethargy, neurological abnormalities (i.e. focal signs, speech disturbances), seizures, skull hematoma. Cranial MRI was obtained within 24 hours after trauma with T2, T2*-weighted gradient-echo (“heme”-sequence), susceptibility (SWI)-, diffusion/apparent diffusion coefficient (DWI/ADC) and T2 FLAIR (fluid attenuated inversion recovery)-weighted sequences, and were evaluated for feasibility and time expenditure. Furthermore we performed a univariate and multivariate regression analysis to evaluate factors associated to trauma findings.

Results: 529 patients (231 male) were included. Age groups: 0-1 years: 2.5%, 1-6 years: 21.9%, 6-12 years: 38.9%, 12-18 years 36.7%. GCS was 15 in more than 90%, GCS-deterioration appeared in 0.9%. Trauma spectrum included falls (47.6%), sport injuries (14%), pedestrian struck by vehicle (10.2%), motor vehicle crash (7.6%), riding accidents (5.9%), bicycle accidents (5.9%), physical attacks (3%), and other mechanisms of injury (5.9%). Sedation for cMRI was needed in 1.9%; general anesthesia was required in 1.5%. In 97.9% axial T2- and in 88.8% coronal T2*-sequences could be obtained. In 12.9% MRI revealed trauma-associated findings. Incidental findings appeared in 4.7%. In the logistic regression model the variables GCS deterioration, clinical signs of skull fracture and basilar skull fracture, age between 0-6 years and injuries involving horses showed significant association with traumatic brain injury on cMRI.

Conclusions: The MRI-protocol established in our clinic, proves practicable in acute mild pediatric TBI settings. It is a safe imaging alternative avoiding long-term risks associated with radiation exposure of cranial computed tomography. Furthermore this study identified potential risk factors which should prompt MRI evaluation.