Table 2 Models created to represent the CDEs. Models are created in YARRRML and made available on the CDE Project GitHub, accompanied by markdown documentation explaining the structure of an appropriate CSV file. Note that not all EU RD CDEs appear 1-to-1 with a CDE model. This is because, for example, the consent CDE can be reused for diverse types of consent (e.g., consent for contact, consent for data reuse), and the Pseudonym CDE is a part of every other model, and therefore has not been modelled as an independent element

Disease Progression [26]

A “container” node to group together all other CDEs that refer to the same diagnosis. For example, the “age of diagnosis” CDE is related to a specific rare disease via traversal into the “disease progression” container, and then traversal into the “diagnosis” CDE that is also connected to “disease progression”

Care Pathway [27]

Captures the date of first contact with the specialist healthcare system; is connected to “disease progression”

Diagnosis [28]

Captures the final disease diagnosis using ORPHA codes; is connected to “disease progression”

Disease History [29]

Captures age at first symptoms and age at diagnosis; is connected to “disease progression”

Genetic Diagnosis [30]

Captures the sequence variant(s) found in this patient, using a variety of different coding systems; is connected to “disease progression”

Patient Consent [31]

Captures the consent of the patient over several axes (e.g., consent for contact, consent for data reuse). Provides a reference to the signed consent form, as well as an input reference to the (blank) consent template.

Patient Status [32]

Captures the current status of the patient, and their date of death if the patient is deceased

Personal Information [33]

Captures (superficial) personal information such as birth date and sex (there are ongoing debates in the EJP modelling group as to whether this should be converted to an age, or an age-range, for improved privacy)

Phenotyping [34]

Captures the phenotypes of the patient, using Human Phenotype Ontology terms

Disability [35]

Captures the score for a disability test. The specific test administered is indicated as one of the child nodes of obo: NCIT_C20993 (Clinical or Research Assessment Tool), and thus this CDE model is broadly useful for many disorders.

Undiagnosed [36]

Captures the case where a patient has phenotypic anomalies, and an identified sequence variant, but for some reason has not been definitively diagnosed.