D2.384 - Biostatistics in allergy and clinical immunology: reasons for concern and editorial recommendations

In recent years, a noticeable decline in the quality of statistical reporting has been observed, leading, among other consequences, to reduced public trust in science, particularly during the COVID-19 pandemic. This study aimed to identify the main statistical errors made by authors submitting articles to EAACI journals.

The analysis included 70 articles reviewed in the capacity of editor or statistical reviewer for Allergy and Clinical and Translational Allergy during the years 2022–2024. Each submitted article was assessed for the appropriateness of the chosen methodology, the interpretation of the obtained results, and the validity of the drawn conclusions.

Among the 70 reviewed manuscripts, 96% contained at least one statistical error. The most common mistakes included failure to assess the necessary assumptions for the proposed analyses, leading to the application of incorrect statistical tests. Additional frequent errors involved the use of inappropriate descriptive statistics, overinterpretation or misinterpretation of results, omission of effect size calculations, improper execution of pairwise comparisons, inadequate reporting of results, and superficial descriptions of performed analyses. A persistent issue remains authors' tendency to provide brief responses to statistical recommendations without implementing the suggested corrections.

Greater emphasis should be placed on implementing statistical recommendations to enhance the quality of reporting in articles submitted to EAACI journals. One suggested approach is to develop and host educational videos on the EAACI website, focusing on key statistical aspects relevant to allergology and clinical immunology.