Unlocking multinational real-world research in atopic dermatitis

An open-source platform enables multinational analysis of dermatology registries for atopic dermatitis while securely safeguarding patient data, a pan-European study finds.

Led by the international TREAT Registry Taskforce, this study showed that the DataSHIELD platform can bypass General Data Protection Regulation, firewall and data-transfer restrictions, enabling secure real-world registry research on atopic dermatitis) without transferring or sharing patient-level data.

Published in the Journal of Investigative Dermatology, the study utilised data five national atopic dermatitis registries within the DREAM TO TREAT AD initiative: A-STAR UK, A-STAR Ireland, SCRATCH Denmark, TREATgermany and TREAT NL/BE. These highlight real-world use, treatment patterns and outcomes for abrocitinib compared with conventional systemic therapies for moderate and severe atopic dermatitis in routine clinical practice.

Enabling harmonised research at scale

The secure, open-source bioscience platform DataSHIELD enabled federated analyses across all five registries simultaneously, validated through parallel non-DataSHIELD analysis.

The researchers reported that clinical dermatology registries can be coordinated across countries while maintaining decentralised governance. They also found that analysts can explore data and perform subgroup analyses without compromising patient privacy. Importantly, the approach proved to be both reproducible and adaptable to support future areas of dermatology.

Crucially, the intellectual property, firewalls, and national data protections of each registry were fully maintained, and personal data never left its host country.

The researchers argued that this addresses one of the major obstacles to expanding real-world evidence in atopic dermatitis, namely, inadequate sample sizes within a single country during the early post-launch phase for new systemic agents.

Atopic dermatitis and other diseases

Study limitations were noted, as it examined only the methodology and proof-of-delivery stage, without providing clinical outcomes. Furthermore, the implementation used dummy data, leaving real-world clinical safety and effectiveness unexplored.

Federated analysis demands substantial harmonisation and IT skills to establish and sustain, and it cannot provide individual-level detail for specific statistical tasks such as outlier detection.

Nevertheless, the researchers stated that this methodology is directly scalable to other biologics and small-molecule advanced therapies, as well as being transferable to other immune-mediated inflammatory diseases.

Overall, these findings indicate that federated data analysis could strengthen clinical decision-making, improve guideline adherence in routine practice, and enable more precise subgroup identification in atopic dermatitis – particularly distinguishing responders from non-responders – by harnessing large, international real-world datasets.

Reference
Krogh NS et al. Implementing Federated Analysis Using DataSHIELD in the DREAM TO TREAT Atopic Dermatitis Registries Collaboration. J Invest Dermatol;S0022-202X(25)02407-8.