The de novo FAIRification process of a registry for vascular anomalies

Abstract

BackgroundPatient data registries that are FAIR - Findable, Accessible, Interoperable, and Reusable for humans and computers - facilitate research across multiple resources. This is particularly relevant to rare diseases, where data often are scarce and scattered. Specific research questions can be asked across FAIR rare disease registries and other FAIR resources without physically combining the data.ResultsWe successfully developed and implemented a process of making a rare disease registry for vascular anomalies FAIR from its conception - de novo. Here, we describe the five phases of this process in detail: i) pre-FAIRification, ii) facilitating FAIRification, iii) data collection, iv) generating FAIR data in real-time, and v) using FAIR data. This includes the creation of an electronic case report form and a semantic data model of the “Set of Common Data Elements for Rare Disease Registration” released by the European Commission, and the technical implementation in the Electronic Data Capture (EDC) system for real-time, automatic data FAIRification. Further, we describe how we contribute to the four facets of FAIR, and how our FAIRification process can be reused by other registries.ConclusionsIn conclusion, a detailed de novo FAIRification process of a registry for vascular anomalies is described. To a large extent, the process may be reused by other rare disease registries, and we envision this work to be a substantial contribution to an ecosystem of FAIR rare disease resources.

De-novo FAIRification via an Electronic Data Capture system by automated transformation of filled electronic Case Report Forms into machine-readable data

“…Post-hoc workflows can be partially reused for other research projects, but typically still require a FAIRification team. In [13] we described a workflow for de-novo FAIRification, where the FAIRification steps are incorporated in the process of setting up and collecting data for a registry or research project (Figure 2.2). Moreover, if a large part of the de-novo FAIRification workflow can take place in an EDC system, the reusability and, thus, scalability of FAIRification across research projects can be significantly improved.…”

“…The transformed data can be exposed and queried 5. The data and metadata are exposed in a manner following the FAIR Data Principles Figure 2: Post-hoc (1) and de-novo FAIRification (2) with an EDC system (adapted from [12] and [13], respectively). The EDC system takes care of most de-novo FAIRification steps.…”

“…Figure 5 shows a filled eCRF in the EDC system. The CDE for Disability was not added for logistic reasons, as described in [13].…”

“…However, creating a semantic data model is a complex task. We recommend that users of our FAIRification method gather expertise from domain and semantic data modeling experts, which is the first step of the de-novo FAIRification workflow [13].…”

“…The method is evaluated by assessing its impact on the Findability, Accessibility, Interoperability, and Reusability of the entered data using the FAIR Evaluation Services [19]. This article reports on the technical implementation and evaluation of the denovo FAIRification workflow, the entire workflow is described in detail in Groenen and Jacobsen et al [13].…”