A Novel Decision-Support Tool (IniCT Score) for Repeat Head Computed Tomography in Pediatric Mild Traumatic Brain Injury

Abstract

The necessity of computed tomography (CT) has been questioned in pediatric mild traumatic brain injury (mTBI) because of concerns related to radiation exposure. Distinguishing patients with lower and higher risk of clinically important TBI (ciTBI) is paramount to the optimal management of these patients. This study aimed to analyze the imaging predictors of ciTBI and develop an algorithm to identify patients at low and high risk for ciTBI to inform clinical decision making using a large single-center cohort of pediatric patients with mTBI. We retrospectively identified pediatric patients with mTBI with repeat CT within 48 hours of injury using an institutional database. Among 3867 pediatric patients, 219 patients with mTBI with repeat CT were included. Thirty-eight had ciTBI (17%), 16 (7%) required intensive care unit admission, and 6 (3%) underwent surgery. Median time interval between initial and repeat CT was 7 hours (range, 4-10). Clinical worsening and radiologic progression were evident in 36 (16%) and 24 (11%) patients, respectively. Multivariate analysis showed that 5 pathologic findings (depressed skull fracture, pneumocephalus, epidural hematoma, subdural hematoma, and contusion) on initial CT and radiologic progression on repeat CT were independent predictors of ciTBI. A new scoring system based on these 5 factors on initial CT (IniCT [Initial CT scoring system] score) had excellent discrimination for ciTBI, need for intensive care unit admission, and neurosurgery (area under the curve >0.8). The IniCT scoring system can successfully differentiate low-risk and high-risk patients based on initial CT scan. Zero score can eliminate the need for a routine repeat CT, whereas scores ≥2 should prompt serial neurologic examinations and/or repeat CT depending on the clinical situation.