A Retrospective Analysis of Risk Factors for Atopic Dermatitis Severity

Atopic dermatitis (AD) has the highest burden of any skin disease; however, the severity-associated factors remain unclear. To evaluate potential severity-associated factors of AD and to design and validate a severity prediction model to inform the management of AD patients. A cross-sectional study of 900 AD patients was conducted from December 2021 to October 2022 at our hospital. The primary outcome was disease severity, categorized as mild, moderate, or severe using the scoring atopic dermatitis index. Ordinal logistic regression and bootstrapped validation were used to derive and internally validate the model. Increasing age, elevated eosinophil level, higher economic status, and urban residence were associated with severe AD. Breastfeeding, disinfectants and topical emollients use, and short duration of bathing were associated with mild AD. In the prediction model, predictors included age, eosinophil and economic status, residence, feeding, disinfectants and emollients use, and duration of bathing. Prediction models demonstrated good discrimination (bias-corrected concordance index [c-index] = 0.72) and good calibration. Risk factors for the severity of AD were identified that could aid the early prediction of AD progression. The predictive model included variables that are easily evaluated and could inform personalized prevention and therapy.