Is it possible to convert OMIM and HPO directly to ORPHAcode?

[Orphabe]

Dear all,

Here are two questions received from a Belgian university hospital (UZ Brussel) that currently codes RD patients with a combination of OMIM and HPO terms:

• If hospitals are asked to convert to SNOMED-CT coding, why would they also learn how to work with ORPHAcode? I’m thinking of the time and effort it will take for physicians to learn how to apply SNOMED-CT and to have them learn ORPHAcodes as well.
• Is it possible to convert OMIM and HPO directly to ORPHAcode?

We already replied to this person, but I post it here for reasons of traceability, and also to ask you if the white paper regarding codification with ORPHAcodes VS SNOMED CT is ready to be shared?

Thank for the support, kind regards, Annabelle (Orphanet Belgium)

[JTorphanet]

Dear @Orphabe,

Following our discussion by e-mail, please find below our answers:

1. SNOMED CT, as for “Systematized Nomenclature of Medecine Clinical Terms”, is a comprehensive, multilingual clinical healthcare terminology that enables the representation of clinical content in electronic health records and is currently in use in more than eighty countries. SNOMED CT is not a terminology dedicated to code rare diseases and it does not include any mention to specify if a given disease is rare or not. Moreover, rare diseases included in SNOMED CT do not have a dedicated classification of rare diseases for data aggregation but are mixed together with non-rare diseases. Therefore, using only SNOMED CT to code rare patients will not allow to retrieve any specific rare disease data nor to make any specific statistical analysis on rare diseases. It is also important to mention that interoperability with other systems that don't use SNOMED CT and registries of rare diseases (that use ORPHAcodes) is not granted just by using SNOMED CT codification. ORPHANET is the only existing medical terminology dedicated to rare diseases. Orphanet produces the Nomenclature of rare diseases following the evolution of knowledge in the research and clinical literature and by consulting experts of the domain. Rare disease inclusion and modification in the Orphanet Nomenclature is evaluated by Orphanet according to standardized procedures, working with international rare disease expert networks. In a second time the information produced by Orphanet is transmitted to SNOMED CT as part of an ongoing collaboration. This proposed updated information is then evaluated by SNOMED CT according to their internal rules for inclusion. Because of the different inclusion criteria, a 100% equivalence between the two nomenclatures is not possible to achieve. Therefore, despite the high level of alignment between the two terminologies that consents an optimal level of interoperability, the entirety of rare disease is not represented in SNOMED CT. Orphanet is working on a document (to be released in the following weeks) aimed at explaining in detail the importance of using ORPHAcodes to complement the SNOMED CT codification.

2. Orphanet database of rare diseases is annotated with the Human Phenotype Ontology (HPO) terms, a standardized and controlled terminology covering phenotypic abnormalities in human diseases (https://hpo.jax.org/app/) . The description of rare diseases by HPO terms can be found on Orphadata (https://www.orphadata.com/phenotypes/) by downloading the product number 4. The annotation production steps of a rare disease are as follows:

   • Phenotypic abnormalities are clearly identified
     • The frequency of the phenotypic abnormality is assessed (Obligate 100%, Very frequent 99- 80%, Frequent 79-30%, Occasional 29-5%, Very rare 4-1%, Excluded 0%)
     • Further possible qualification of the phenotypic abnormalities is added: a major diagnostic criterion or a pathognomonic sign of the rare disease, consensual criteria for diagnosing the disorder.

To avoid any ambiguity, Orphanet newly add the term ‘isolated’ to a disease presenting as a single clinical sign (see below an example), if the clinical sign exists independently as a disease. This is done to avoid confusion while coding patients. On the contrary, if a clinical sign cannot appear in isolation (but only as part of a syndrome) and it has been created in Orphanet in the past (probably as an error, or for lack of better knowledge), upon discovery of the incoherence the "incorrect" disease is removed from the nomenclature. (example: ORPHAcode 2238, Familial isolated hypoparathyroidism and HP:0000829, Hypoparathyroidism) Moreover, and for your information, Phenomizer, a software provided by HPO (https://hpo.jax.org/app/tools/phenomizer) propose a tool where you can list HPO terms and obtain diagnostic possibilities. The ouput is based on OMIM and Orphanet clinical entities. The mapping between OMIM and ORPHANET is not fully possible. Indeed, alignment is qualified following 3 different relationships: exact, NTBT (narrower term to broader term) or BTNT (broader term to narrower term). The direct conversion is only possible for the exact mapping. On the Orphanet website, you can use the search tool using an OMIM number. The output will give you all the ORPHAcodes link to this OMIM number but to go further and confirm the conversion, all the alignments and their natures are available on ORPHADATA (https://www.orphadata.com/alignments/) by downloading the product number 1. To your purpose, you have to select only the exact mapping and create a new list of direct conversion.

About the white paper, I am sorry for the delay, the final validation is still on going.

Thank you for posting this message on the Github, great interest for the Orphanet community !

Best,

Julie