MBUFNI 8 F3OBU J#SLFOGME
that we suppose will lead to a better FAIRification score. A higher FAIRification score is the
benchmark for success of this work. At the time of conducting this work the CDS was only
available as an MS Excel file on the DZD website. A substantial part of it was not available in
any structured form nor did it have an accompanying data dictionary. To improve its FAIRness
score, we converted the CDS into the recommended format for spreadsheets, annotated the
parameters therein with LOINC [
5
], licensed the dataset, indexed the metadata in a searchable
resource and enriched the dataset with metadata. These steps resulted in an overall increase in
the FAIRness score by 47%.
The DZD strives to set standards for good scientific management of clinical studies. The
development of a common DZD CDS increased the interoperability of clinical studies under the
DZD umbrella. The next step towards reusability and comparibility of studies is the adherence to
the FAIR principles. The data owners therefore agreed to direct their efforts into structuring the
data and mapping the local codes to standard terminologies in order to provide understandable,
valuable and fit-for-purpose data. The mapping task is not trivial and requires domain knowledge
as well as understanding of the used standard terminology terms to make sure that the semantic
meaning is correctly translated from the local codes to the applied standards. Data validation
remains a critical step to ensure that the data generated and codes used are valid. According to
the SATIFYD, a dataset containing an open license is FAIRer than one containing a restricted,
embargo or any other license. The FAIR Cookbook however indicates that the type of license
chosen does not cause differences in the FAIRness. This is a common misconception about
licensing in matters FAIR [
6
]. The implementation of a machine actionable format will come
with the cost of time and immense effort because the DZD CDS is not encoded in any structured
format. The return on this investment is increased certainty of the future data readability.
We postulate that the results of this evaluation will help formulate the first steps towards a
FAIRification workflow for related DZD datasets, facilitate the planning for the resources that
the FAIRification journey will require and motivate the stakeholders to engage in this journey.