A Retrospective Analysis to Investigate Contact Sensitization in Greek Population Using Classic and Machine Learning Techniques.

Allergic contact dermatitis (ACD) is an inflammatory reaction affecting all age groups and both sexes. ACD is characterized by a delayed-type hypersensitivity reaction IV caused by skin contact with haptens. Chronic exposure typically leads to a decrease in erythema accompanied by lichenification (thickening and hardening of the skin) and persistent itching. The current study aims to investigate the patterns of contact sensitization in the Greek population using patch test data analysis. Patch test data from 240 patients (120 Males/120 Females) with allergic contact dermatitis were collected at the Laboratory for Patch Testing, National Reference Center for Occupational Dermatoses "Andreas Syggros" Hospital in Athens Greece. The contact allergic reactions were caused by ethylenediamine dihydrochloride 1%, thimerosal 0.5%, and methyldibromo-glutaronitrile 0.1% from the European baseline series of allergens; information was also collected for ICDRG evaluation, an extended MOAHLFA index and patient-reported outcomes (daily routine questionnaire). The chi-square test for independence and Spearman's rank were used to evaluate the association and correlation, respectively, between patient characteristics and ACD-related factors. Multiple correspondence analysis (MCA), which is a data analysis approach, was used to find and depict underlying structures in the data collection for nominal categorical data. Statistically significant associations were found between the following pairs of characteristics: eczema triggers and gender and eczema triggers and hand dermatitis. The results from MCA showed that there is correlation between allergic contact dermatitis onset, allergens, and demographic variables.