Clinical Prediction Scores for Pediatric Appendicitis

Abstract

Abstract Introduction The rate of misdiagnosis of appendicitis in children is a challenge and clinical prediction scores could be part of the solution. However, the pediatric appendicitis score (PAS) and the Alvarado score have shown disappointing diagnostic accuracy in pediatric validation studies, while the appendicitis inflammatory response (AIR) score and the novel pediatric appendicitis risk calculator (pARC) have not yet been validated thoroughly. Therefore, the aim of the present study was to evaluate these four prediction scores prospectively in children with suspected appendicitis. Materials and Methods A prospective study was conducted over a 2-year period. All patients <15 years with suspected appendicitis were eligible for inclusion. The four prediction scores were compared regarding predictive values, receiver operating characteristics (ROC) curves, decision curve analysis, and clinical outcome. Results Of the 318 patients included, 151 (47 %) patients had appendicitis. The AIR score and the pARC had substantially higher specificity and positive predictive value, and lower rate of false positives (7% and 2%), than the PAS and Alvarado score (36 and 28%, p < 0.001). Across the different gender and age groups, the AIR score and the pARC generally had fewer false positives than the PAS and Alvarado score. There were no significant differences in sensitivity, negative predictive values, rates of missed appendicitis, or ROC curve analysis. In decision curve analysis, the AIR score and the pARC outperformed the PAS and Alvarado score at most threshold probabilities. Conclusion The AIR score and the pARC are superior to the PAS and Alvarado score in diagnosing children with suspected appendicitis.