Identifying Children at Very Low Risk for Blunt Intra-Abdominal Injury in Whom CT of the Abdomen Can Be Avoided Safely

J Am Coll Surg. 2017 Apr;224(4):449-458.e3. doi: 10.1016/j.jamcollsurg.2016.12.041. Epub 2017 Jan 24.

Abstract

Background: Computed tomography is commonly used to rule out intra-abdominal injury (IAI) in children, despite associated cost and radiation exposure. Our purpose was to derive a prediction rule to identify children at very low risk for IAI after blunt abdominal trauma (BAT) for whom a CT scan of the abdomen would be unnecessary.

Study design: We prospectively enrolled children younger than 16 years of age who presented after BAT at 14 Level I pediatric trauma centers during 1 year. We excluded patients who presented more than 6 hours after injury or underwent abdominal CT before transfer. We used binary recursive partitioning to derive a prediction rule identifying children at very low risk of IAI and IAI requiring acute intervention (IAI-I) using clinical information available in the trauma bay.

Results: We included 2,188 children with a median age of 8 years. There were 261 patients with IAI (11.9%) and 62 patients with IAI-I (2.8%). The prediction rule consisted of (in descending order of significance): aspartate aminotransferase >200 U/L, abnormal abdominal examination, abnormal chest x-ray, report of abdominal pain, and abnormal pancreatic enzymes. The rule had a negative predictive value of 99.4% for IAI and 100.0% for IAI-I in patients with none of the prediction rule variables present. The very-low-risk population consisted of 34% of the patients and 23% received a CT scan. Computed tomography frequency ranged from 4% to 96% by center.

Conclusions: A prediction rule using history and physical examination, chest x-ray, and laboratory evaluation at the time of presentation after BAT identifies children at very low risk for IAI for whom CT can be avoided.

Publication types

  • Observational Study

MeSH terms

  • Abdominal Injuries / diagnostic imaging*
  • Adolescent
  • Child
  • Child, Preschool
  • Decision Support Techniques*
  • Female
  • Follow-Up Studies
  • Humans
  • Infant
  • Infant, Newborn
  • Logistic Models
  • Male
  • Prospective Studies
  • Risk Assessment
  • Sensitivity and Specificity
  • Tomography, X-Ray Computed*
  • Unnecessary Procedures*
  • Wounds, Nonpenetrating / diagnostic imaging*