‘A Flurry of Terms, But Not Many Users’, in the current issue of Health Data Management, describes the Systematized Nomenclature of Medicine – SNOMED. SNOMED is a standard medical vocabulary that offers consistent identification of a patients medical condition, something that cannot be done effectively through the use of free form text entries in a medical record. Medical record databases that make use of ICD-9 and CPT-4 code sets, which are used primarily for billing and reimbursement purposes, cannot match a SNOMED encoded database, with respect to aggregating clinical information for the purpose of performing clinical outcomes analysis and research. So what’s the problem? It isn’t free.
As pointed out in the article, extracting a dataset from a clinical information system that consists of “all female patients between the ages of 30 and 45 with chronic hepatitis C who had undergone liver biopsies during the preceding 30 months” is a snap with a SNOMED-encoded database. With ICD-9 and CPT-4 encoded data, the resulting dataset would be of questionable integrity (missing and/or cases included that do not fit the criteria). Medical records that consist of free-form text entries would require extensive data analysis, validation and scrubbing prior to being certified for use in a clinical research project.
Despite the benefits of SNOMED, it is not as widely used in medical information systems as one would suppose. As pointed out in the article, licensing of SNOMED can be costly with respect to annual fees and the process of negotiating reasonable terms. Additionally, the learning curve associated with SNOMED may be akin to the learning curve associated with Linux – not rocket science, but time will need to be set aside for training and experimentation, before results are seen in everyday use.
As Open Source and Free Software Foundation enthusiasts have shown, proprietary systems and software that present steep ownership costs, distribution/licensing challenges, and extreme manpower requirements can be met effectively through the creation of Open and ‘Free’ solutions. So when can we expect an Open Source and/or ‘Free’ medical vocabulary project that would offer the benefits of SNOMED as well as integrate with Open Source medical information systems currently under development? Inquiring minds want to know.