Derivation and validation of a novel clinical decision aid to distinguish between uncomplicated and complicated appendicitis in children.
Abstract Aim of Study:Non-operative treatment of acute uncomplicated appendicitis (UA) in children might be equally effective to surgery but requires accurate discrimination from those with complicated appendicitis (CA) to ensure safety and maximise efficacy. We aimed to identify specific clinical and laboratory parameters that would aid distinction between UA and CA in children.Methods:Retrospective review of consecutive children with a clinical +/- radiological diagnosis of acute appendicitis that underwent appendicectomy in three specialist paediatric surgical centres between March 2017 and February 2018. Demographic, clinical and laboratory data were retrieved and analysed in relation to intra-operative and histopathological findings. CA was defined as gangrene and/or perforation seen intra-operatively and/or in histopathological analysis. Multiple logistic regression analysis was used to derive a novel prediction model that could accurately distinguish UA and CA. A priori we set analytical parameters so as to ensure the score had a positive predictive value (PPV) for UA of >95%. The resulting scoring system was validated in an independent cohort of children.Main Results: The prediction model was derived from 130 children (UA: 71; CA: 59) with median age (range) 10 (2-15) years. Initial univariate analysis identified six factors significantly associated (p<0.01) with CA (duration of abdominal pain, presence of rebound tenderness, temperature, , white cell count, , neutrophil count and C-reactive protein [CRP]). These variables were entered in the regression model, and points awarded based on the adjusted odds ratios. Receiver operating characteristic analysis demonstrated a threshold of ≥4 points for prediction of CA. The scoring system was validated in an independent cohort of 112 children (UA: 51; CA: 61); it was found to have a sensitivity of 98% and specificity 78%. A score of <4 points had a PPV for UA of 98%.Conclusions: Our novel scoring system can discriminate between UA and CA in children with high accuracy. Children with a score <4 could be eligible for non-operative treatment.