Prospective validation of two systems of classification for the diagnosis of acute appendicitis

The objectives of this study were (1) to validate the Alvarado score and pediatric appendicitis score (PAS) in a prospectively identified pediatric cohort and (2) to assess abdominal ultrasonography (AUS) as a tool to increase the diagnostic reliability of both scores.

Prospective study conducted from January 10, 2008, to January 1, 2009. All patients attended at the emergency department with suspected acute appendicitis (AA) who had a blood sample collected were included. Items from both scores were recorded. The performance of an AUS, the decision to admit the patient, and the therapeutics were decided by the physician, disregarding the scores values. Nonadmitted patients were contacted by telephone.

Ninety-nine patients were included. Mean age was 11 years, and 62.6% were males. Appendectomy was performed in 44.4% patients. The area under the receiver operating characteristic curve for the Alvarado score was 0.96 and that for PAS was 0.97. Not a single patient with an Alvarado score less than 5 or PAS less than 4 had AA. All patients with an Alvarado score greater than 8 or PAS greater than 7 had AA. For both scores, the optimum cutoff point was 6 (sensibility of 90.4% and specificity of 91.2% for the Alvarado score and sensibility of 88.1% and specificity of 98.2% for PAS). Abdominal ultrasonography was performed on 31 patients (sensibility of 84.6% and specificity of 94.4%). We studied the value of scores and AUS together. Assuming an Alvarado score from 1 to 4 and PAS from 1 to 3 as no AA, an Alvarado score from 9 to 10 and PAS from 8 to 9 as AA, and proceeding according to the AUS for intermediate values, a sensibility of 93.3% and 97.2% and a specificity of 100% and 97.6%, respectively, were obtained.

Both scores are a useful tool in the evaluation of children with possible AA. For extreme values of scores, the results really ensure their use in the emergency department. The AUS can help on decision making for intermediate values.