These two concepts have FSNs that use exactly the same words in different orders. The only difference is that the second one doesn’t have the word ‘disorder’ in the main description, but only in the semantic tag, which is a part of the description and is searchable as such.
However, SI says “The two concepts do not share duplicate descriptions.” (CRS Ticket 977864). The reasoning is that the first one represents the ORPHANET code 633099 for PAICS deficiency and is required for mapping purposes, whereas the second one is just the enzyme deficiency.
While I understand this theoretical reasoning, I don’t see how to effectively convey these differences in translation. To begin with, I maintain that the two FSNs are actually duplicates, considering how SNOMED CT is applied in EHR software systems. If a clinician looks up these words and is confronted with both, the choice will be likely made at random, since we can’t expect all clinicians to be familiar with all the theoretical intricacies of SNOMED.
Having to keep these two formally identical concepts encourages us translators to come up with artificial ways to distinguish the two. This would mean forcing the word ‘disorder’ somewhere in the translation of the first one, something that would result in an artificial description that no clinician would use.
In fact, the official ORPHANET French translation of the rare disease is “Déficience en phosphoribosylaminoimidazole carboxylase”, which is a literal translation of the second one.
What do you think is the best way to proceed in these cases?
I get even more confused when I look at the modelling and se that the Phosphoribosylaminoimidazole carboxylase deficiency disorder (disorder) is due to Deficiency of phosphoribosylaminoimidazole carboxylase (disorder). Can a deficiency cause a disorder? Yes. Can that disorder cause the same disorder?
Or is it really about congenital or acquired deficiency?
I would have understood the wish to have two concepts if Deficiency of phosphoribosylaminoimidazole carboxylase was a finding. It could then represent the lab finding of a deficiency of this enzyme, regardless of the actual diagnosis of the orphan disease. I don’t know this specific disorder but usually in nature, you can have isolated enzymatic low levels, that may qualify as “deficiencies” when set against lab normal values, but that you have no explanation for, are a lower end of the gaussian, or just transitory. And then you have the proven deficiency syndromes or disorders.
If you follow the idea of the lab anomaly, I would translate this one as “carence” not “déficience”. Carence will convey the lower level of the enzyme noted clinically while not implying it is necessarily inborn or even expected to be long lasting.
But one should go do the litterature to see if there is any report of this enzyme being low outside of the orphan disease context. If the deficiency exists only with this genetic anomaly, then sorry but for me they are duplicates. There is no reason to have two concepts one for the mapping and one representing the same but not mapped, If one concept exists to make a 1-1 mapping you use that one.
My logical request here would be to change the semantic tag of 124597000 |Deficiency of phosphoribosylaminoimidazole carboxylase (disorder)| to (finding). But in depht litterature review should be performed before to answer the the above mentioned questions.
Thanks, Marie-Alexandra! Yes, I also wondered by it’s not a finding. But it seems that a majority (though not all) ‘deficiency’ concepts are disorders, to be distinguished from ‘[substance] below reference range’, which are findings (sometimes translated as ‘carence’ indeed).
The SNOMED reasoning is that a substance below reference range is a finding because it does not imply a permanently abnormal situation, it’s just an observation. Hence the distinction between 238115004 |Sodium deficiency (disorder)| and 89627008 |Hyponatremia (finding)|.
I’m not a clinician, but I have my doubts such a distinction is useful. “Sodium deficiency” had initially been translated as ‘hyponatrémie’ but had to be changed before publication.
@mbraithwaite has pointed out there is a tracker on this topic. Not sure whether there’s any planned work within SI at the moment since the last update in June 2025 (Thank you, Maria!).