Even after refinement, the rule is not sufficient for excluding this common diagnosis.

A previously published prediction rule developed at a single center classifies children as having low risk for appendicitis if they meet the following criteria: absolute neutrophil count 6.75 x 10³/µL, absence of nausea, and absence of maximal tenderness in the right lower quadrant (RLQ). The rule yielded a sensitivity of 98% and specificity of 32%. Now, the investigators validated and refined the rule in a prospective cohort of 2625 children (age range, 3 to 18 years; mean, 11 years) who presented to 9 pediatric emergency departments during 1 year with abdominal pain of <96 hours duration and suspected appendicitis.

Overall, 1018 patients (39%) were diagnosed with appendicitis, and of these, 275 (27%) had a perforated appendix. Computed tomography was performed in 55% of patients, ultrasound in 37%, and both in 12%. The negative appendectomy rate was 9% (95 patients). Had the rule been applied, 22 unnecessary operations would have been prevented, but 42 patients with appendicitis would have been missed (sensitivity, 96%; specificity, 36%). A refined rule, consisting of absolute neutrophil count 6.75 x 10³/µL and either no maximal tenderness in the RLQ or maximal tenderness in the RLQ and no abdominal pain with walking, jumping, or coughing, identified 400 patients as having low risk for appendicitis. The refined rule would have missed 19 patients with appendicitis, yielding a sensitivity of 98% and specificity of 24%.

Comment: The investigators failed to compare performance between these rules and clinical suspicion for appendicitis based on physical examination alone. Overall clinical gestalt should still be the basis for identifying children at low risk for appendicitis. When the diagnosis is uncertain, investigation should begin with ultrasound, followed by computed tomography if ultrasound is nondiagnostic.

Source: Journal Watch Emergency Medicine