Take CARE of your data! FAIRly!
This repository is a new, evolving version of the CARE Semantic Model, derived from the original CARE-SM developed by Pablo Alarcón-Moreno as part of his PhD thesis, "Applying deep semantics to the representation of clinical data to improve machine usability" (Universidad Politécnica de Madrid, 2024).
Building on that foundation, this version aims to:
- Consolidate and harmonize several of the original individual data element models into a more unified, consistent representation.
- Adapt the models to handle real-world use-cases that could not be represented in the original CARE-SM. For example, the original model had no way to represent a negative observation — e.g. a phenotype test whose result is "false" — even though that absence-of-finding is itself an important, recordable fact.
This work is under active construction. Models, structures, and documentation here may change as the harmonization effort progresses. We welcome feedback, questions, and suggestions — see Communication and Feedback below.
If you're migrating existing data or tooling from the original CARE-SM, here's what changed. Full details, including why each change was made, are on the Migrating from CARE-SM v1 page on ReadTheDocs — this is a condensed summary.
The most important thing to know: the Attribute_ node (Output_ --refers to--> Attribute_) no longer means one single thing everywhere. It now has three different cardinalities depending on the model — unconditional (same as v1, most models), conditional (Phenotype, Diagnosis — only asserted when the result is true, which is also how negative findings became representable for the first time), and multi-valued/repeating (Genetic — one per sequence variant, instead of a single identifier per report). See the RTD page for the full table of which models use which behavior.
| Model | Summary of change |
|---|---|
| Phenotype | Negative-observation support, duration fix, identifier-bnode fix |
| Diagnosis | Same three fixes as Phenotype; new (currently unpopulated) evidentiary-provenance placeholder |
| Disability → Functional Assessment (renamed) | Recategorized; instrument identity split from administration protocol; optional output label |
| Genetic | Identifier moved onto Attribute_ — a report can now describe multiple variants |
| Birthplace | Identifier-bnode fix |
| Consent | New optional consent-form reference (filename + version, not a URL) |
| Hospitalization, Surgery | Added a missing output/date branch — v1 couldn't record dates for these at all |
| Clinical_trial, Cohort, Medication, Deathdate, Questionnaire, Symptoms_onset | Node renamed for consistency only — no RDF change |
None of this has propagated to the CARE-SM Toolkit, YARRRML, or CSV templates yet — see the RTD page's "Not yet migrated" section.
You can explore the complete documentation on ReadTheDocs
The documentation includes:
- Detailed descriptions of all data elements
- Implementation guidelines
- Tools
- Exemplar data
- Additional resources
Your feedback is more than welcome and will help us improve the CARE Semantic Model.
Please use GitHub Issues to provide feedback or report problems:
github.com/wilkinsonlab/CARE-Semantic-Model-Version-2/issues
Zenodo link here
If you used CARE-SM in your work, please cite the following publication:
@inproceedings{caresm2024,
author = {Pablo Alarc{\'o}n-Moreno and Mark Denis Wilkinson},
title = {{Take CARE of your patient data: Clinical And Registry Entries (CARE) Semantic Model}},
booktitle = {Proceedings of the 15th International Conference on Semantic Web Applications and Tools for Health Care and Life Sciences (SWAT4HCLS 2024)},
year = {2024},
publisher = {CEUR-WS.org},
series = {CEUR Workshop Proceedings},
volume = {3890},
url = {https://ceur-ws.org/Vol-3890/paper-11.pdf}
}Previous publication:
Semantic modeling of common data elements for rare disease registries, and a prototype workflow for their deployment over registry data
This work originated in the European Joint Programme on Rare Diseases (EJP RD), which received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 825575.
It is now continued as part of the European Rare Disease Research Alliance (ERDERA), which receives funding from the European Union’s Horizon Europe research and innovation programme under grant agreement No. 101156595.
After the end of the EJP RD project, this work has been led and maintained by researchers from the Wilkinson Lab at Universidad Politécnica de Madrid.
