This page collects some initial thoughts about the FHIR4FAIR FHIR IG
To be discussed
- Structure of the guide
- Extension of the "informative" part
- Proposed development approach
- Organization of a FAIR track ?
Structure of the guide
Proposed development approach
FAIR data object is a quite wide concept that could cover very different things of real life: from a collection of data (data set) to a single atomic information (e.g. a diagnosis).
Depending of what is a "FAIR object" the boundary between data and metadata may change; including how the "object" could be represented in FHIR and instances assessed.
The suggestion to be discussed with the team is the therefore the following:
- identify relevant types of FAIR objects for health data in real life (to be determined how many and which one)
- the guide is restricted to social and health data, other kinds of health-related artefacts, as clinical guidelines, algorithms, software, models are out of scope.
- select one of them as reference case and focus on that kind of object
- identify a minimal set of information that are relevant in real life for the components of that FAIR object.
- proceed incrementally starting from few essential information, in order to excursive also the development process (i.e. not to try to cover everything from the beginning)
The following picture shows one proposed conceptual structure of a FAIR data object , (another representation includes the provenance in the metadata and add as a separate component "standards and code")
hereafter a preliminary draft of types of objects we may be interested to consider as starting point, suggesting to start with data collections.