A systematic review of statistical approaches in clinical studies for atopic dermatitis treatment

Abstract

Biostatistics plays a pivotal role in clinical research by supporting accurate data analysis and interpretation, which are essential for drawing valid and evidence-based conclusions. In studies of atopic dermatitis (AD), a chronic, relapsing inflammatory skin condition, statistical methods are critical for evaluating clinical outcomes such as SCORAD scores, skin hydration, and patient-reported symptoms. As the number of trials investigating the efficacy and safety of skin care products for AD continues to grow, the quality and appropriateness of the statistical techniques used in these studies warrant close examination. This systematic review assessed statistical methods employed in clinical trials of skin care interventions for AD published between January 2020 and April 2025. Following PRISMA guidelines, a comprehensive search of PubMed was conducted, and 16 studies meeting the inclusion criteria were analysed in detail. A wide range of statistical methods was reported, including t-tests, ANOVA, ANCOVA, chi-square tests, and non-parametric alternatives such as the Wilcoxon signed-rank and Mann–Whitney U tests. Some studies also applied advanced techniques like mixed-effects models and ROC analysis. While most methods were generally appropriate for the study designs, frequent shortcomings were observed in the testing of assumptions and consistency of statistical reporting. These methodological gaps may limit the interpretability and reproducibility of trial outcomes. The findings highlight the need for improved statistical rigor and greater transparency in the analysis and reporting of clinical trials assessing skin care products for AD, ultimately to support more robust and reliable conclusions in dermatological research.