RESUMEN
Aims: Several scoring systems have been developed for diagnosis of appendicitis. This study aims to systematically explore how those scores were derived and validated, and to compare their performance. Study Design: Systematic review. Place and Duration of Study: We searched Medline from 1949 and EMBASE from 1974 to March 2012 to identify relevant articles published in English. Methodology: Information about model development and performance was extracted. The “risk of bias” assessment tool was developed based on a critical appraisal guide for clinical prediction rules. Calibration (O/E ratio) and discrimination (C-statistic) coefficients were estimated. A meta-analysis was applied to pool calibration coefficients and Cstatistics. Results: Forty-four out of 468 studies were eligible. Of these, 14 developed or modified diagnostic scoring systems and 30 validated existing models. Four scores had been most frequently validated, i.e., Alvarado, modified Alvarado, Fenyo, and Eskelinen. Among them, only the Eskelinen model was derived based on a multivariate regression whereas the rest used univariate or non-statistical methodology. All studies reported very good but imprecise calibration. For discrimination, the pooled C-statistics for these corresponding scores were 0.77, 0.86, 0.81, and 0.84 respectively. In the external validation, the discriminative performance decreased about 25.3% and 10.1% for the Alvarado and Fenyo scores respectively. Conclusion: The research methods for scoring systems of appendicitis were inconsistent. More efficient scoring systems which have been internally and externally validated are required.