LAUSR

Editor We read with interest the publication by Bhangu and the RIFT Study Group1 comparing different appendicitis scoring systems in identifying patients with a low risk of appendicitis. The study found that the Adult Appendicitis Score (AAS) performed best with a cut-off score of 8 or less. In men, the Appendicitis Inflammatory Response Score (AIRS) performed best with a cut-off score of 2. The cut-offs of both scores were lower than in original publications: 10 or less for AAS2 and 4 or less for AIRS3. This raises the concern whether the calculation of these scores was done properly. In both scores, muscular defence or guarding in the right inferior fossa (RIF) is categorized into three (none, mild and moderate or severe) or four (none, light, medium and strong) categories. Although the publication does not provide detailed information on data collection, the study protocol (version 4.2, dated 8 March 2017) with case report forms was available from the study website (http://wmresearch .org.uk). It seems that primary data collection did not include categorization of RIF guarding and, thus, calculation of both AAS and AIRS were biased. This would affect the optimal cut-off values and could reflect on the performance of these scores. The publication refers to a web application (http://appy-risk.org), where either AAS or AIRS is calculated according to gender. Furthermore, this calculator does not provide data entry for categorized RIF guarding and, thus, does not give a proper score when compared to the online AAS calculator (http://www.appendicitisscore.com). Because both AAS and AIRS are derived from logistic regression models, elimination of one significant factor in the calculation will worsen their performance. If adequate data collection had taken place, we believe that these scores could have performed even better. P. Mentula , H. Sammalkorpi and A. Leppäniemi Department of Gastrointestinal Surgery, Helsinki University Hospital, Helsinki, Finland