Data Model Documentation

[bwalsh]

From today's discussion:

Notes to bootstrap the documentation

ACED data model

Introduction

TODO: this whole section Notes:

• We define all our metadata as FHIR resources.
• Metadata composition happens on the client
• There is a workflow

FHIR graph model

Examine resource definitions here:

• Details on uploaded files are captured as DocumentReference
• DocumentReference.subject indicates who or what the document is about:
  • Can simply point to the ResearchStudy, to indicate the file is part of the study
  • Can point to Patient, or Specimen, to indicate the file is based on them
• An Observation can point to any entity
• A Task can provide provenance on how the file was created

Each resource has at least one study controlled official Identifier. Child resources have Reference fields to point to their parent.

Key resources

ResearchStudy

A scientific study of nature that sometimes includes processes involved in health and disease. More

ResearchSubject

A ResearchSubject is a participant or object which is the recipient of investigative activities in a research study. More

Patient

Demographics and other administrative information about an individual or animal receiving care or other health-related services. More

Specimen

A sample to be used for analysis. More

DocumentReference

A reference to a document of any kind for any purpose. More

Identifiers

A string, typically numeric or alphanumeric, that is associated with a single object or entity within a given system. Typically, identifiers are used to connect content in resources to external content available in other frameworks or protocols.

More

References

Many of the defined elements in a resource are references to other resources. Using these references, the resources combine to build a web of information about healthcare.

More

CodeSystems (Ontologies)

The Fast Healthcare Interoperability Resources (FHIR) standard defines a set of terminology resources that provide a mechanism to access clinical terminology in a standardised way. Many clinical terminologies are currently supported, including SNOMED CT, RxNorm and LOINC. Recently, there has been increasing interest in supporting terminologies from other domains. In genomics, for example, there is increasing use of the Human Phenotype Ontology (HPO)1 to capture information about rare diseases.

PMID: 32308861

SNOMED: While there are many possible uses of snomed in the entire FHIR model, our focus is on these fields:

• Specimen.collection.bodySite
• Condition.bodySite
• Substance.code TODO: pubchem?

MONDO:

• Condition.code

[Other: TODO]():

• TODO

Happy path

• TODO As an ACED user, in order to contribute data, I need docs, code and examples on how to populate spreadsheets, validate and submit data.

Workflow

• TODO Use synthetic data as an example

[bwalsh]

Graph above created by See http://gb2.clinfhir.com/

my test project-Example view-fsh.fsh.txt my test project-Example view.json

[matthewpeterkort]

Probably Also worth it to mention LOINC codes? Also, This seems like good bare bones, but a new data model user might need some more context on FHIR before pointing them to many FHIR docs pages. Will try to contribute to this a bit in the coming days.