gms | German Medical Science

Clinical healthcare actors spend about 30% of their working hours documenting their daily work, which is almost as much time as they spend in direct interaction with patients [1]. The routinely captured data, which is associated with a lot of work, effort and time, is so valuable that it should be used beyond its primary purpose. The non-direct care use of clinical data is essential for effective clinical research and known as secondary use [2]. A convincing use case is secondary use in clinical trials. Despite the potential, there are not many practical examples, and systematic literature reviews show that reuse in clinical trials plays an underestimated role [3], [4]. This study investigated the potential for secondary use in a local environment. The clinical data warehouse with an integrated clinical trial management system at the UKSH Campus Lübeck serves as the source system with integrated clinical and research data of consented patients. The IT infrastructure based on validated data protection and consent concepts is in operation since 2015. The infrastructure includes solutions for study registration, recruitment, CRF-based study documentation and data management. The metadata examined included case report forms (CRFs) of monocentric trials from the Department of Cardiology, Angiology and Intensive Care Medicine, originally initiated on the Lübeck campus. The hypothesis of this study states that the potential for reuse of data is greater when considering an in-house study. If a study eCRF is developed at the same site, we assume that mainly parameters are included which are already known and routinely collected anyway. For this purpose, all metadata of the data warehouse were exported and transferred to a metadata repository at regular intervals. We use the open source metadata repository Samply.MDR [5]. This facilitates the later administration of the metadata and enables similarity analysis among the metadata. Following the hypothesis, the analysis was divided into two steps. In the first step, all data elements that occurred in one of the CRFs were compared against all data that was routinely and automatically transmitted via a RESTful interface. This was also done by comparing the namespaces of the MDR, but an additional mainly manual step added the equivalence classes [6] of the found matches. The second step was to determine the similarity between the eCRFs of a study and the eCRFs of all studies. For this purpose, the data used were compared with elements from the different namespaces, and the respective coverage was calculated. The considered studies combine 9 arms with a total of 35 visits and 23 CRFs with an overall of 494 data elements. It could be covered by 18,9% out of a total of 5487 data items administered. Without considering data quality problems, such as missing values this study shows the high potential of secondary use, but there is still space for improvement. Assumed that in the future more and more information in clinical life will be collected in a standardized or structured way, a positive trend can be considered.

The authors declare that they have no competing interests.

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