MOSAIC – A Modular Approach to Data Management in Epidemiological Studies

T. Bahls, C. Havemann, J. Piegsa, K. Weitmann, T. Wegner, W. Hoffmann, M. Bialke
2015 Methods of Information in Medicine  
Introduction: In the context of an increasing number of multi-centric studies providing data from different sites and sources the necessity for central data management (CDM) becomes undeniable. This is exacerbated by a multiplicity of featured data types, formats and interfaces. In relation to methodological medical research the definition of central data management needs to be broadened beyond the simple storage and archiving of research data. Objectives: This paper highlights typical
more » ... nts of CDM for cohort studies and registries and illustrates how orientation for CDM can be provided by addressing selected data management challenges. Methods: Therefore in the first part of this paper a short review summarises technical, organisational and legal challenges for CDM in cohort studies and registries. A deduced set of typical requirements of CDM in epidemiological research follows. Results: In the second part the MOSAIC project is introduced (a modular systematic approach to implement CDM). The modular nature of MOSAIC contributes to manage both technical and organisational challenges efficiently by providing practical tools. A short presentation of a first set of tools, aiming for selected CDM requirements in cohort studies and registries, comprises a template for comprehensive documentation of data protection measures, an interactive reference portal for gaining insights and sharing experiences, supplemented by modular software tools for generation and management of generic pseudonyms, for participant management and for sophisticated consent management. Conclusions: Altogether, work within MOSAIC addresses existing challenges in epidemiological research in the context of CDM and facilitates the standardized collection of data with pre-programmed modules and provided document templates. The necessary effort for in-house programming is reduced, which accelerates the start of data collection.
doi:10.3414/me14-01-0133 pmid:26196494 fatcat:vhaok6l4rjctzgxnxl2tbemoju