A generic solution for web-based management of pseudonymized data

BMC Medical Informatics and Decision Making

Technical Advance

Florian Kohlmayer¹  Fabian Prasser¹  Klaus A. Kuhn¹  Ronald Lautenschläger¹ 

[1] Chair for Biomedical Informatics, Department of Medicine, Technical University of Munich (TUM), Grillparzerstraße 18, 81675, Munich, Germany;

关键词: Electronic data capture;    Security;    Privacy;    Confidentiality;    Pseudonymization;    Web-based application;    Seamless integration;    Mashup;    Cross-domain communication;   

DOI  :  10.1186/s12911-015-0222-y

 received in 2015-05-21, accepted in 2015-11-25,  发布年份 2015

来源: Springer

【 摘 要 】

BackgroundCollaborative collection and sharing of data have become a core element of biomedical research. Typical applications are multi-site registries which collect sensitive person-related data prospectively, often together with biospecimens. To secure these sensitive data, national and international data protection laws and regulations demand the separation of identifying data from biomedical data and to introduce pseudonyms. Neither the formulation in laws and regulations nor existing pseudonymization concepts, however, are precise enough to directly provide an implementation guideline. We therefore describe core requirements as well as implementation options for registries and study databases with sensitive biomedical data.MethodsWe first analyze existing concepts and compile a set of fundamental requirements for pseudonymized data management. Then we derive a system architecture that fulfills these requirements. Next, we provide a comprehensive overview and a comparison of different technical options for an implementation. Finally, we develop a generic software solution for managing pseudonymized data and show its feasibility by describing how we have used it to realize two research networks.ResultsWe have found that pseudonymization models are highly heterogeneous, already on a conceptual level. We have compiled a set of requirements from different pseudonymization schemes. We propose an architecture and present an overview of technical options. Based on a selection of technical elements, we suggest a generic solution. It supports the multi-site collection and management of biomedical data. Security measures are multi-tier pseudonymity and physical separation of data over independent backend servers. Integrated views are provided by a web-based user interface. Our approach has been successfully used to implement a national and an international rare disease network.ConclusionsWe were able to identify a set of core requirements out of several pseudonymization models. Considering various implementation options, we realized a generic solution which was implemented and deployed in research networks. Still, further conceptual work on pseudonymity is needed. Specifically, it remains unclear how exactly data is to be separated into distributed subsets. Moreover, a thorough risk and threat analysis is needed.

【 授权许可】

CC BY   
© Lautenschläger et al. 2015