Some of these top-level domains also have additional sub-domains. The fully-specified names of SNOMED CT concepts include the top-level domain or the sub-domain in parentheses. The domain, concept’s position in the hierarchy (e.g., parents, children, siblings) and its relationships (e.g., causative agent, finding site, etc.) help clarify the meaning of a concept. In other words, SNOMED CT is not merely a huge list of terms – it has a comprehensive terminology model that should be taken into account to disambiguate concepts.
Here are some examples of seemingly duplicate concepts but can be disambiguated by the domain:
- Morphological Abnormality vs Disorder: ‘abscess (morphologic abnormality)|44132006’ and ‘abscess (disorder)|128477000’.
- Product vs Substance: ‘Aspirin (substance)|387458008’, and ‘Aspirin (product)|7947003’.
- Procedure vs Qualifier: ‘Wright stain method (procedure)|104214004’, and ‘Wright stain method (qualifier value)|703441005’.
- Specimen vs Body Structure: ‘Products of conception tissue sample (specimen)| 258428005’, and ‘Structure of product of conception (body structure)| 26864007’.
- Situation vs Procedure: ‘Hand tendon reattached (situation)|288066005’, and ‘Reattachment of tendon of hand (procedure)|45810006’.
- Specimen vs Substance: ‘Amniotic fluid specimen (specimen)|119373006’, and ‘Amniotic fluid (substance)|77012006’.
Remember to use these domains as a tool to select the appropriate concept for your use case. Unfortunately, other systems or users may send the incorrect concept if merely searching on text without looking into the meaning of each possible choice. As a result, there could be a negative impact on semantic interoperability, decision support and data analytics. Therefore, it is necessary to design the system to deal with this situation, such as grouping ‘possible’ concepts for each use case.