¹Department of Medical Informatics, University Medical Center Göttingen (UMG), Göttingen, Germany
²German Center for Child and Adolescent Health (DZKJ), Partner Site Göttingen, Göttingen, Germany
³Cluster of Excellence “Multiscale Bioimaging: from Molecular Machines to Networks of Excitable Cells” (MBExC), University of Göttingen, Göttingen, Germany
⁴Göttingen State and University Library, University of Göttingen, Göttingen, Germany
⁵Campus-Institute Data Science (CIDAS), Göttingen, Germany

Text

Introduction: The development of data management plans (DMPs) and data policies (DPs) is becoming an increasingly prevalent requirement for researchers engaged in biomedical projects [1], [2], [3]. However, these concepts are often conflated or inconsistently interpreted, resulting in inefficiencies in data management practices and leading to uncertainty among researchers and institutions. We therefore felt that a clearer understanding of these concepts is needed to move beyond their perceived bureaucratic nature and fully utilize their capacity for ensuring long-term research data quality, accessibility, and reusability.

Methods: We conducted a cross-disciplinary review of DMP and DP implementations at Göttingen Campus, analyzing their application and requirements across diverse research contexts. This included the examination of DMPs utilized by PhD students at biomedical Göttingen graduate schools, the data management team of the DZKJ consortium, and clinical trial management teams. We identified commonalities, terminological variations, and specific expectations, enabling us to assess the scope, complexity, and overall effectiveness of DMPs in various settings.

Results: Our research highlighted significant differences in terminology and application of DMPs across different domains. For example, PhD students are introduced to DMPs as part of their individual research projects primarily for educational purposes, while the DZKJ data management team develops DMPs at a consortium-wide level. Clinical trial management teams primarily use DMPs to comply with specific regulatory requirements. This range from simple documentation of current workflows to comprehensive best-practice frameworks underscores the need to clearly define the purpose and scope of DMPs for each case. Without proper instruction, DMPs are also at risk to merely reflect the current state of data management without offering pathways for improvement, limiting their utility. As an already existing countermeasure, the Göttingen eResearch Alliance provides first-level-support for DMP/DP development and connects activities at Göttingen campus to broader approaches like the RDMO Community [4] and DMP4NFDI [5].

Discussion: Our findings emphasize the importance of balancing the time and effort required to complete DMPs and DPs with the practical benefits they provide. Therefore, it is crucial to clarify the intended objectives and scope of each document before its development, whether it is focused on individual project needs, institutional guidelines, or compliance with funding or regulatory requirements. Providing well-informed, customized guidance is a key determinant at this step. By clearly defining their goals, DMPs and DPs can serve as practical and efficient tools for improving data management, rather than simply fulfilling bureaucratic obligations. Based on our findings, we recommend to implement a flexible, modular approach to DMP design allowing researchers to select relevant components based on their specific demands and contexts, thereby optimizing both the time investment and the overall impact of DMPs.

Conclusion: By adopting a modular design and providing precise guidance, we can effectively use DMPs and DPs as dynamic instruments that improve data management and research outcomes while maintaining a favorable cost-benefit-ratio.

Acknowledgements: Human resources were mainly funded by DFG project funding [CRC1565 (Project 469281184, Z), Germany’s Excellence Strategy – EXC 2067/1–390729940, NMDR3 – DFG Project 315072261, NFDI4Health – DFG Project 442326535] and by BMBF funding through the DZKJ (01GL2402C).

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.

References

[1] Michener WK. Ten simple rules for creating a good data management plan. PLoS Comput Biol. 2015 Oct 22;11(10):e1004525. DOI: 10.1371/journal.pcbi.1004525
[2] Miksa T, Simms S, Mietchen D, Jones S. Ten principles for machine-actionable data management plans. PLoS Comput Biol. 2019 Mar 28;15(3):e1006750. DOI: 10.1371/journal.pcbi.1006750
[3] Mittal D, Mease R, Kuner T, Flor H, Kuner R, Andoh J. Data management strategy for a collaborative research center. GigaScience. 2022 Dec 28;12:giad049. DOI: 10.1093/gigascience/giad049
[4] Anders I, Enke H, Hausen DA, Henzen C, Jagusch G, Lanza G et al. The Research Data Management Organiser (RDMO) - a strong community behind an established software for DMPs and much more. Data Science Journal. 2024 May 8;23(1). DOI: 10.5334/dsj-2024-028
[5] Diederichs K, Förstner K, Hausen D, Jagusch G, Johannsen J, Lindstädt B, et al. DMP4NFDI - NFDI Basic Service for Data Management Plans. Zenodo; 2024 Aug 29. DOI: 10.5281/zenodo.13445355