EBQ:Validation of a prediction rule for the identification of children with intra-abdominal injuries after blunt torso trauma

Complete Journal Club Article

Holmes J. et al. "Validation of a Prediction Rule for the Identification of Children With Intra-abdominal Injuries After Blunt Torso Trauma". Annals of Emergency Medicine. 2009. 54(4):528-33.
PubMed Full text PDF

Clinical Question

What is the accuracy of the prediction rule derived in the publication Holmes, JF et al. Identification of children with intra-abdominal injuries after blunt trauma. Ann Emerg Medicine. 2002;39:500-509. for detecting intra-abdominal injury in children after blunt torso trauma?

Conclusion

The clinical decision rule defined by Holmes et al. for ruling out intra-abdominal injury in children with blunt torso trauma would substantially reduce unnecessary abdominal CT in pediatric trauma patients.
Because the rule yielded a false negative in one patient, who underwent diagnostic laparotomy without surgical intervention, it may need further refinement before widespread application.

Major Points

6-variable clinical rule test characteristics:
Sensitivity 94.5%
Specificity 37.1%
33% reduction in unnecessary abdominal CT scans

Study Design

• Prospective, multicenter observational validation study
• Part of the Pediatric Emergency Care Applied Research Network (PECARN)
• 20 emergency departments across North America
• N = 2,188 children with blunt torso trauma
• Primary Outcome: validation of a previously derived clinical prediction rule for identifying children with intra-abdominal injuries requiring acute intervention

Inclusion Criteria

• <18 yr old
• Blunt torso trauma
• Underwent definitive testing for intra-abdominal injury: CT, diagnostic peritoneal lavage, diagnostic laparotomy or laparoscopy.

Exclusion Criteria

• Penetrating trauma
• Pregnant patients
• Presentation >24hr after injury
• Patients who did not undergo definitive testing because of low suspicion for intra-abdominal injury

Interventions

• No therapeutic intervention; this was a validation study of a clinical prediction rule
• Clinicians prospectively assessed 7 clinical variables before knowledge of CT results:
  • Evidence of abdominal wall trauma/seat belt sign
  • GCS score <14
  • Abdominal tenderness
  • Evidence of thoracic wall trauma
  • Complaints of abdominal pain
  • Decreased breath sounds
  • Vomiting

Outcome

Primary Outcomes

1. The presence of intra-abdominal injury
   
2. Intra-abdominal injury requiring acute intervention, defined as: blood transfusion for intra-abdominal hemorrhage, angiographic embolization of vessel or organ, or therapeutic intervention at laparotomy.

Secondary Outcomes

Reduction in abdominal CT scans if the rule were to be strictly applied to rule-out intra-abdomninal injury.

Subgroup analysis

Children who had intra-abdominal injury missed by the decision rule.

Criticisms

The clinical decision rule failed to identify 8 children with an intra-abdominal injury that was ultimately detected on definitive testing. However, none of these subjects required acute intervention for their injury.

Funding

UC Davis Children's Miracle Network Research Grant
SAEM Research Training Grant