Second DICA oncological clinical registry (inter)nationally reusable
Clinical registries improve the quality of care by providing insights through research on the collected data. The Netherlands has quality registries for various diseases, which have already led to significant improvements in healthcare. However, to date the reusability of these quality registries has remained mostly limited to the Netherlands, whilst these gained insights could potentially make an impact internationally.
In a joint initiative with the NVRO (the Dutch Society for Radiotherapy and Oncology), DICA (Dutch Institute for Clinical Auditing) and MRDM have, following the Colorectal Radiotherapy registry (DCRA-R), now also made the Breast cancer Radiotherapy registry (NBCA-R) (inter)nationally reusable by making it adhere to the FAIR principles.
FAIR is an acronym for Findable (via a unique identifier and clear descriptive metadata), Accessible (via a protocol that allows for authorization/authentication), Interoperable (due to a data model and commonly used controlled vocabularies) and Reusable (due to metadata describing the context under which data was generated and a license for data (re)use). The objective of FAIR data is stimulating and optimising the reuse of data, by both humans and computers. Therefore, the FAIR principles promote the importance of machine readability, due to the increasing need on the computing power in processing data, the rapid increase in volume, complexity and creation rate of data.
By making the DCRA-R and NBCA-R adhere to the FAIR principles, the first large-scale FAIR database in the field of clinical radiotherapy variables in the Netherlands has been launched. This means that these registries can now be (re)used (inter)nationally for scientific research and to improve patient care. Besides (inter)national reusability, adhering to the FAIR principles is also a first step towards application of Personal Health Train (PHT) principles. The PHT facilitates analysis of (privacy-sensitive) data sources at the sources themselves, without human access to the (privacy-sensitive) data, and only returns aggregated statistics.
Implementation of FAIR principles
The FAIR principles were drafted generically, due to the required broad applicability, with the result that these can be met in a variety of ways. We took the following specific steps to make the NBCA-R adhere to the principles of FAIR principles:
A FAIR sharing page and GitHub repository were created. In here, various metadata are made available, such as background information and the data dictionary of the registry. Further, the FAIRsharing page provides a globally unique and permanent identifier.
A protocol has been made available (NL/EN) for gaining access to the privacy-sensitive data. If the data supplying parties decide to release the privacy-sensitive data, it is made available through the secure service Zivver. The overall metadata is publicly and freely accessible and permanently available.
The data of the registry is converted to RDF format, in which the variables are linked to concepts and their definitions from existing ontologies. Also, relationships between the variables and corresponding ontology concepts are established, mainly using the Radiation Oncology Ontology (ROO).
The DICA licence for data (re)use and information about data collection, pre-processing and validation have been published as metadata.
A possible next step is to make other DICA clinical registries adhere to the FAIR principles and subsequently (inter)nationally reusable. In this way, the collected datasets and gained insights can have an even greater impact on improving the quality of healthcare.
Want to know more?
Would you like to know more about this joint initiative and what it could mean for you? Please reach out to the MRDM Service Desk (via firstname.lastname@example.org with the subject FAIR).