Though readily available, and the reference standard for diagnosing intraabdominal injuries (IAI) the radiation exposure from a CT is not benign – especially in children. Ultrasound in trauma (FAST) is a valuable tool – but moreso in adults as children may have physiologic free fluid and injuries that will not require surgery as compared to grown ups with similar findings. Even obtaining labs can be a stressful procedure for a small child. What then should we do when presented with a child with blunt abdominal injury whom we consider to be very low risk for intraabdominal injuries that would require a surgical intervention (laparotomy, angiographic embolization, transfusion for intraabdominal hemorrhage, IV Fluids for ≥2 nights for pancreatic/bowel injuries)? Is there any way to be able to be reassured after a history and physical exam alone that no additional studies are necessary?
The short answer in all likelihood is yes.
Well, it just so happens that The Pediatric Emergency Care Applied Research Network (PECARN) addressed this particular issue in a multicenter study designed to derive a prediction rule to help identify which children were very low risk for IAI. Holmes et al. prospectively enrolled over 12,000 children with blunt torso trauma, who had an average age of 11 years. 761/12,044 had IAI (6.3%). Of these 761 a little more than a quarter needed one or more of the acute interventions listed above – 203/761 (26.7%). Using binary recursive partitioning (a multivariable statistical method used to create decision trees/rules) the authors identified the following factors, in descending order of importance, that were related to children being low risk for IAI requiring acute intervention:
- No evidence of abdominal wall trauma or seat belt sign
- GCS >13
- No complaints of abdominal pain
- No decreased breath sounds
- No vomiting
The statistical characteristics of this rule were:
- Negative predictive value – 99.9% (95%CI 99.7% to 100%)
- Sensitivity – 97%( 95% CI 94% to 99%)
- Specificity – 42.5% (95% CI 41.6% to 43.4%)
- Negative likelihood ratio – 0.07 (95% CI 0.03 to 0.15)
Here is the prediction rule in its glorious flow-chartiness:
This essentially means, that if a child with blunt torso trauma has NONE of the above factors then they are highly unlikely to have a clinically significant IAI – even before getting labs or ultrasound. The prediction rule itself only failed to identify 5 patients – you can read more about them in the article. Though conducted on a large number of patients the authors noted that external validation is necessary. I also agree with their assertion that even if a patient has 1 or more factors they still don’t NEED to get a CT. It may be just as appropriate to screen those children with labs (AST/ALT, pancreatic enzymes, urinalysis, CBC and CXR). Like any other decision rule it is designed to supplement, not replace clinical judgment. Still, this article represents a significant step towards better stratifying the risk of clinically important IAI and is therefore a must read. Check it out at the following link – which you may need institutional accounts to access:
