Predictive Model for Operative Intervention after Blunt Abdominal Trauma in Children with Equivocal CT Findings: A Pilot Study

Background: To study the clinical and radiologic factors predicting the need for surgical intervention after blunt abdominal trauma (BAT) in children with equivocal computed tomography (CT) scan findings. Methods: We performed a retrospective review of the trauma database at our level I pediatric trauma center between 2011 and 2019. We selected patients with BAT and equivocal findings for surgical intervention on CT scan. We studied five factors: abdominal wall bruising (AWB), abdominal pain/tenderness (APT), thoracolumbar fracture, the presence of free fluid (FF), and the presence of solid organ injury, all previously reported in the literature to predict the need for operative intervention. We used t-test, the Kruskal-Wallis test and logistic regression to study the association of these factors with the need for operation in our pediatric cohort. Results: Of 3044 blunt trauma patients, 288 had abdominal CT scans with 61 patients demonstrating equivocal findings. Operation was performed for 12 patients (19.7%) confirming surgically correctable traumatic injuries. The need for surgical intervention was significantly associated with the age of the patients (P = 0.03), the presence of APT (P = 0.001), AWB (P = 0.01), and FF (P = 0.04). The presence of thoracolumbar fracture and solid organ injury were not significantly associated with the need for operation. For the subset of 37 patients who were injured in a motor vehicle crash, five (13.5%) required surgical intervention, which was significantly associated with the presence of AWB (P = 0.04), APT (P = 0.01), and FF (P = 0.03). A predictive model that used these factors produced a receiver operating characteristic curve of 0.86. Conclusions: In cases of equivocal abdominal CT scan findings to evaluate BAT in children, the presence of abdominal wall tenderness, AWB, or FF may be significant factors predicting more accurately the need for operative intervention. A predictive model using the combination of clinical and image findings might determine with more certainty, the need for surgical intervention in children with BAT and equivocal CT findings. Validation on a larger multi-institutional data set should be done.