Background/purpose: Appendicectomy is an operation that is often performed without certainty of diagnosis. This study aimed to construct and to validate a prognostic score for the diagnosis of acute appendicitis in children.
Methods: Data for 35 symptoms and signs were prospectively recorded for 131 consecutive children with suspected appendicitis. Logistic regression analysis of the variables yielded a diagnostic score: gender (male 2 points, female 0) + intensity of abdominal pain (severe 2, mild or moderate 0) + relocation of pain (yes 4, no 0) + vomiting (yes 2, no 0) + pain in the right lower abdominal quadrant (yes 4, no 0) + fever (yes 3, no 0) + guarding (yes 4, no 0) + bowel sounds (abnormal 4, normal 0) + rebound tenderness (yes 7, no 0). The cut-off level for recommendation of appendicectomy was > or =21, and the cut-off level for non-appendicitis was < or =15. The score was prospectively validated on 109 children.
Results: In the validation sample, based on clinical judgment, unnecessary appendicectomy was performed in ten (27%) children, and one (4%) child was misdiagnosed as not having appendicitis. By application of the score, unnecessary appendicectomies would have been reduced to four (13%), and three children (11%) with appendicitis would have been discharged.
Conclusion: The use of a predictive mathematical model may facilitate the diagnosis of appendicitis to avoid unnecessary operations.