From b0a814afb033de35a0453af9a8e13045f6991b14 Mon Sep 17 00:00:00 2001 From: raniag13 Date: Tue, 25 Mar 2025 11:01:01 -0500 Subject: [PATCH] Rania edits from Jane March 2025 --- .../Book Content/Book SOLOR Terminology.xml | 19 +- .../Book Terminology-Knowledge.xml | 386 +++++++----------- 2 files changed, 148 insertions(+), 257 deletions(-) diff --git a/statements/src/docbkx/Book Content/Book SOLOR Terminology.xml b/statements/src/docbkx/Book Content/Book SOLOR Terminology.xml index 4dd2f18..580558c 100644 --- a/statements/src/docbkx/Book Content/Book SOLOR Terminology.xml +++ b/statements/src/docbkx/Book Content/Book SOLOR Terminology.xml @@ -6,16 +6,15 @@
Solor Solor integrates terminology content (SNOMED CT®, LOINC®, RxNorm, etc.) from its - native format into a common Solor format. Once the content is in Solor, where - equivalency is determined through various methods where concepts of the same idea are - aggregated. For example, Gentamycin from SNOMED CT® is the same as Gentamycin from - LOINC®, and is also the same Gentamycin from RxNorm. The end result from this process is - the creation of a Solor concept that is devoid of any source information (but will have - traceability), and is exposed to the user to view, use, and extend. In the Gentamycin - example, a user will find a concept that is devoid of any source information and will - not need to know if this is the SNOMED CT®/LOINC®/RxNorm Gentamycin that needs to be - selected. Solor concepts are identified using a Universally Unique Identifier - (UUID). + native format into a common Solor format. Once the content is in Solor, equivalency is + determined through various methods where concepts of the same idea are aggregated. For + example, Gentamycin from SNOMED CT® is the same as Gentamycin from LOINC®, and is also + the same Gentamycin from RxNorm. The end result from this process is the creation of a + Solor concept that is devoid of any source information (but will have traceability), and + is exposed to the user to view, use, and extend. In the Gentamycin example, a user will + find a concept that is devoid of any source information and will not need to know if + this is the SNOMED CT®/LOINC®/RxNorm Gentamycin that needs to be selected. Solor + concepts are identified using a Universally Unique Identifier (UUID).
Inte<?oxy_comment_start author="tocrow" timestamp="20240624T173028-0400" comment="I think this is covered in other sections"?>roperability diff --git a/statements/src/docbkx/Book Content/Book Terminology-Knowledge.xml b/statements/src/docbkx/Book Content/Book Terminology-Knowledge.xml index 9090e82..926de46 100644 --- a/statements/src/docbkx/Book Content/Book Terminology-Knowledge.xml +++ b/statements/src/docbkx/Book Content/Book Terminology-Knowledge.xml @@ -515,6 +515,132 @@ The September 2022 edition of RxNorm includes 13856 (base) ingredients, 5065 bra </tgroup> </table> </section> + <section> + <title><?oxy_comment_start author="rgeorgacopoulos" timestamp="20250312T214316-0500" comment="moved up from section below SNOMED CT Design Criteria"?>Comparing<?oxy_comment_end?> + LOINC®, SNOMED CT® and RxNorm to Cimino’s Desiderata + Despite the guiding principles for terminology design and representation summarized + above, “LOINC® does not fulfill the definition of a computable medical terminology as + articulated by Cimino et al., that being one based in concept orientation, concept + definition, and polyhierarchy. These characteristics enable the logical inference + between concepts, parts, and part elements.” [12] LOINC® is a well-maintained + terminology – there is a deprecation process and has concept permanence where codes + aren’t reused – but they do not establish the basic parameters of hierarchy and + polyhierarchy. And, they are violating the tenants of Cimino’s Desiderata. + + Correspondence of LOINC® to Cimino’s Desiderata + + + + + + Tenant + LOINC® + + + Concept Orientation + No + + + Concept Permanence + Yes + + + Non-Semantic Concept + Identifier + Yes + + + Polyhierarchy + No + + + Formal Definitions + No + + + Classification + No – they have a miscellaneous code + + + Multiple Granularities + No + + + Multiple Consistent Views + No + + + Representation of Context + No, but they have Narrative Summary + + + +
+
+ Comparison of SNOMED CT® to Cimino’s Desiderata + SNOMED CT® is the terminology of our current awareness most closely aligns with + the Desiderata. As stated elsewhere, it uses Description Logics for computability + and facilitates the use of reasoners to determine equivalence among various + syntactic representations. The primary downsides are 1). Some proprietary concepts + and 2). Lengthy concept submissions process for consideration into either the + National or International editions. This lack of agility, efficiency and speed is + apparent in crisis situations such as COVID-19 pandemic. Otherwise, it has proven to + be robust and has a large global community to assist with its development and + vetting. + + Correspondence of SNOMED CT® to Cimino’s Desiderata + + + + + + Tenant + SNOMED CT® + + + Concept Orientation + Yes + + + Concept Permanence + Yes + + + Non-Semantic Concept + Identifier + Yes + + + Polyhierarchy + Yes + + + Formal Definitions + Yes + + + Classification + Yes + + + Multiple Granularities + Yes + + + Multiple Consistent + Views + Yes + + + Representation of + Context + Yes + + + +
+
+
Unified Medical Language System (UMLS) The NLM integrates terms and codes from over 150 source vocabularies by concept, @@ -543,32 +669,10 @@ The September 2022 edition of RxNorm includes 13856 (base) ingredients, 5065 bra language processing tools - The Semantic Network and Lexical Tools are used to produce the Metathesaurus. - The - steps to produce the Metathesaurus involve: - - - Processing the terms and codes from source terminologies using the Lexical - Tools - - - Grouping synonymous terms into UMLS concepts - - - Categorizing concepts by semantic types using the Semantic Network - - - Incorporating relationships and attributes provided by source - terminologies - - - Releasing the data in a common format called the Rich Release Format - (RRF) - - - These tools, while helpful, have gaps. Raje et al. highlighted issues with - completeness, correctness, and redundancy when they found gaps in the UMLS - Metathesaurus’ coverage of disease concepts. + The Semantic Network and Lexical Tools are used to produce the Metathesaurus. These + tools, while helpful, have gaps. Raje et al. highlighted issues with completeness, + correctness, and redundancy when they found gaps in the UMLS Metathesaurus’ coverage of + disease concepts. Raje The UMLS does not necessarily look at semantic equivalence of concepts across or within terminologies, rather focuses on lexical equivalence. A good example of this is how the @@ -863,8 +967,8 @@ OHDSI's Standard Vocabulary relies heavily on mapping non-standard concepts place to ensure only correct terms are associated with an object in Solor. Having a consistent naming convention defined will assist with textual queries to identify duplicates when concepts are primitive and not able to be fully defined using relationships within Solor. Having a consistent way of representing Fully Specified Names will alleviate the issue of users creating duplicate concepts like "Disorder of immune function" and "Immune function disorder". Consistent naming is also important to support effective retrieval. For example, - the SNOMED CT® concept 386560004 |Glasgow coma score finding (finding)| has 13 - children all with the string Glasgow coma scale instead of Glasgow coma score. + the SNOMED CT® concept 386557006 Glasgow coma score finding (finding)| has 13 + children all with the string Glasgow coma scale instead of Glasgow coma score. Another common issue is to add a synonym to a concept that is more specific than the concept itself. A concept should only have synonyms that accurately represent a concept and not any of its children. If a synonym has a more specific meaning, a new @@ -1059,218 +1163,6 @@ OHDSI's Standard Vocabulary relies heavily on mapping non-standard concepts LOINC® parts.
-
- Comparing LOINC®, SNOMED CT® and RxNorm to Cimino’s Desiderata - Despite the guiding principles for terminology design and representation summarized - above, “LOINC® does not fulfill the definition of a computable medical terminology as - articulated by Cimino et al., that being one based in concept orientation, concept - definition, and polyhierarchy. These characteristics enable the logical inference - between concepts, parts, and part elements.” [12] LOINC® is a well-maintained - terminology – there is a deprecation process and has concept permanence where codes - aren’t reused – but they do not establish the basic parameters of hierarchy and - polyhierarchy. And, they are violating the tenants of Cimino’s Desiderata. - - Correspondence of LOINC® to Cimino’s Desiderata - - - - - - Tenant - LOINC® - - - Concept Orientation - No - - - Concept Permanence - Yes - - - Non-Semantic Concept - Identifier - Yes - - - Polyhierarchy - No - - - Formal Definitions - No - - - Classification - No – they have a miscellaneous code - - - Multiple Granularities - No - - - Multiple Consistent - Views - No - - - Representation of - Context - No, but they have Narrative Summary - - - -
-
- Comparison of SNOMED CT® to Cimino’s Desiderata - SNOMED CT® is the terminology of our current awareness most closely aligns with - the Desiderata. As stated elsewhere, it uses Description Logics for computability - and facilitates the use of reasoners to determine equivalence among various - syntactic representations. The primary downsides are 1). Some proprietary concepts - and 2). Lengthy concept submissions process for consideration into either the - National or International editions. This lack of agility, efficiency and speed is - apparent in crisis situations such as COVID-19 pandemic. Otherwise, it has proven to - be robust and has a large global community to assist with its development and - vetting. - - Correspondence of SNOMED CT® to Cimino’s Desiderata - - - - - - Tenant - SNOMED CT® - - - Concept Orientation - Yes - - - Concept Permanence - Yes - - - Non-Semantic Concept - Identifier - Yes - - - Polyhierarchy - Yes - - - Formal Definitions - Yes - - - Classification - Yes - - - Multiple - Granularities - Yes - - - Multiple Consistent - Views - Yes - - - Representation of - Context - Yes - - - -
-
-
- Comparison of RxNorm to Cimino’s Desiderata - While RxNorm meets most of the Desiderata’s requirements, there are a few that are - not met. Of particular importance is the lack of Description Logics. Use of a - Description Logics reasoner can help in overcoming important drawbacks of - multi-axial systems: (1) the capability of detecting semantic equivalence of - syntactically different expressions, and (2) the automated classification of - concepts in a hierarchy. RxNorm has another issue of not being a global standard as - its development and main use is in the U.S. where it is a nationally recognized - standard, i.e. Meaningful Use. - However, recent work by Bona et al. have been developing a Drug Ontology product - (DrOn), based on RxNorm and Chemical Entities of Biological Interest Ontology, which - is a modular and extensible ontology of drug products, their ingredients and - biological activity to enable comparative effectiveness and allow researchers to - query National Drug Codes (NDCs) in multiple ways. Bone et al. go on to say the - following: - We have implemented a full accounting of national drug - codes and RxNorm unique concept identifiers as information content entities, and - of the processes involved in managing their creation and changes. This includes - an OWL file that implements and defines the classes necessary to model these - entities. A separate file contains an instance-level prototype in OWL that - demonstrates the feasibility of this approach to representing NDCs and the - RxCUIs and the processes of managing them by retrieving and representing several - individual NDCs, both active and inactive, and the RxCUIs to which they are - connected. We also demonstrate how historic information about these identifiers - in DrOn can be easily retrieved using a simple SPARQL Protocol and RDF Query - Language (SPARQL).[13] - This development may have a significant impact on improving knowledge management - as well as the ability to utilize Description Logics reasoners with its attendant - benefits. - - Correspondence of RxNorm to Cimino’s Desiderata - - - - - - Tenant - RxNorm - - - Concept Orientation - Yes - - - Concept Permanence - No – remapping, changing atoms, splitting, etc. - - - Non-Semantic Concept - Identifier - Yes - - - Polyhierarchy - Yes - - - Formal Definitions - No – [Ingredients] [Strength] [Dose] enough? - - - Classification - Yes - - - Multiple - Granularities - Yes - - - Multiple Consistent - Views - Yes - - - Representation of - Context - Yes - - - -
-
-
Current Opportunities for SNOMED CT®, LOINC® and RxNorm to Integrate, Enhance and Extend @@ -1566,10 +1458,10 @@ OHDSI's Standard Vocabulary relies heavily on mapping non-standard concepts connectivity by an additional 9000 LOINC® codes - Overall, these results suggest decent quality maintenance for LOINC® as these - counts are relatively small considering the 450k+ concepts in LOINC® as part of - this study. However, the key takeaway from this body of work is that LOINC® - could benefit from the application of Description Logic by allowing for + Overall, these results suggest adequate quality maintenance for LOINC® as + these counts are relatively small considering the 450k+ concepts in LOINC® as + part of this study. However, the key takeaway from this body of work is that + LOINC® could benefit from the application of Description Logic by allowing for automated classification and equivalence detection. [19]
@@ -1818,9 +1710,9 @@ OHDSI's Standard Vocabulary relies heavily on mapping non-standard concepts The objective of this work is to document data collection methods and statement model standards to demonstrate the importance of an interoperable data system and downstream analytical scenarios through a use case for Long COVID as an example of - an emerging acute chronic infectious disease. We aim to highlight how Analysis - Normal Form (ANF) can be used to standardize the representation of equivalent - concepts associated with data capture, storage and aggregation within the Long COVID + an emerging chronic infectious disease. We aim to highlight how Analysis Normal Form + (ANF) can be used to standardize the representation of equivalent concepts + associated with data capture, storage and aggregation within the Long COVID workflow. This work utilized four main aims to demonstrate how ANF-thinking could improve use case outcomes and reduce the clinical burden on data consumers during data collection and aggregation: