Abstract
Purpose :
We conducted a systematic analysis of the Ryan’s Retina atlas, a well-recognized information resource in ophthalmology, to identify gaps in representation of retinal diagnoses in the Systematized Nomenclature of Medicine (SNOMED) International terminology.
Methods :
Retinal diagnoses were identified from each chapter outline, free text, and index terms in Ryan’s Retina 7th edition. These were mapped against a standardized terminology web browser to identify gaps in representation in SNOMED. The mappings were classified according to the adequacy of representation in SNOMED: Complete match, Partial match, or No match. Retinal diagnoses with no matching concepts in SNOMED were submitted to a nationwide group of retinal specialists within the Retina Subgroup of the OHDSI Workgroup in Eye Care and Vision Research for review and prioritization based on clinical relevance. A structured survey allowed each specialist to rank the importance of developing standardized representation for each diagnosis without an existing match in SNOMED. This was followed by synchronous discussion to achieve consensus. The diagnoses that achieved consensus for high priority (at least 3 out of the 5 retinal specialists) for standardized representation were then discussed among the SNOMED International Eye Care Clinical Reference Group (CRG) and submitted to SNOMED International for terminology development. (Figure 1)
Results :
A total of 327 diagnoses were identified. Among existing SNOMED terms as of October 21, 2022, 231 (70.6%) had complete match, 63 (19.2%) partial match, and 33 (10.1%) no match. Of the 33 diagnoses with no matching concepts, consensus was achieved regarding 11 diagnoses to have a definite term or condition code defined in SNOMED. (Table 1)
Proposed condition codes were discussed at an Eye CRG meeting. Terminologists at SNOMED International requested cross-mappings with available International Classification of Disease version 10 codes, definitions for each diagnosis, and examples of peer-reviewed literature discussing the diagnosis. After addressing clarifying questions regarding mappings and hierarchies from SNOMED terminologists, new codes were added to SNOMED to represent the diagnoses that were previously not represented.
Conclusions :
This international, multi-institutional collaborative process enabled identification of gaps, prioritization, and development of data standards to address these gaps among retinal diagnoses.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.