Objective
To develop a clinical score to predict appendicitis among older, male children who present to the emergency department (ED) with suspected appendicitis.
Methods
Patients with suspected appendicitis were prospectively recruited at 9 pediatric EDs. A total 2,625 patients enrolled; a subset of 961 males, age 8-18 were analyzed in this secondary analysis. Outcomes determined by pathology, operative reports and follow-up calls. Clinical and laboratory predictors with < 10% missing data and Kappa > 0.4 were entered into a multivariable model. Resultant beta-coefficients were used to develop a clinical score. Test performance was assessed by calculating the sensitivity, specificity, positive predictive value [PPV], negative predictive value [NPV], and likelihood ratios [LRs].
Results
The mean age was 12.2 years; 49.9% (480) had appendicitis, 22.3% (107) had perforation, and the negative appendectomy rate was 3%. In patients with and without appendicitis, overall imaging rates were 68.6% (329) and 84.4% (406), respectively. Variables retained in the model included maximum tenderness in RLQ, pain with walking/coughing or hopping, and the absolute neutrophil count. A score ≥ 8.1 had a sensitivity of 25% [95% CI 20-29%], specificity of 98% [96-99%] and PPV of 93% [86-97%] for ruling in appendicitis.
Conclusion
We developed an accurate scoring system for predicting appendicitis in older boys. If validated, the score may allow clinicians to manage a proportion of these male patients without diagnostic imaging.