“To date, the dental profession has failed to develop a commonly accepted standardized terminology to describe oral diagnoses and has lagged far behind medicine in its codification of diagnoses.” (Kalendrian et al., 2010)
The lack of standardized clinical terminology for dental care has many negative effects, but this post focuses on a key problem: evaluating dental EHRs. How do you evaluate dental EHRs without a standardized way to track and understand clinical diagnoses? What clinical features do you need in the electronic record, and how do you know the software will store them?
The COHRI (Consortium for Oral Health-Related Informatics) members decided to undertake the task of developing a shared, clinical terminology for dental diagnoses. Understanding how a team of researchers developed a clinical taxonomy (the EZ Codes) for dental diagnoses may assist in understanding how to evaluate and consider the clinical diagnosis features in available dental EHRs. First, a practitioner needs to understand the difference between existing types of clinical classification features:
“Where ICD is a terminology, SNODENT is an ontology. The distinction between the two is that a terminology is a set of terms representing the concepts within a particular field, while an ontology represents the relationships between these concepts. SNODENT is comprised of diagnoses, signs, symptoms, and complaints and currently includes over 6,000 terms.” (Kalendrian et al., 2010)
A terminology creates standard definitions for terms; an ontology uses standardized terminology to define the relationship between various terms–for instance, anesthetic and diuretics, or diabetes and periodontal care. The researchers decided to create a terminology so that they could have a baseline for further improvement in later years. They also chose to have a set of guiding principles, the first of which was to incorporate existing terminologies and ontologies whenever possible (i.e., ICD-9, ICD-10, SNOMED and Z-codes). They also investigated and applied the best practices from terminology development research, the field of information science, including designing the terminology with information retrieval in mind at the outset. The use of hierarchies was critical to this retrieval process.
Perhaps most importantly for dentists considering purchasing an EHR, the researchers chose to link diagnostic and procedural codes. That linking provides key conceptual information. When selecting an EHR, ask the vendor if and how the diagnostic and procedural information connects. As the authors state:
“Dental Procedure Code, which is owned and biannually updated by the ADA. This code is designated by the federal government under the Health Insurance Portability and Accountability Act (HIPAA) as the national terminology for reporting dental services and is recognized by third-party payers nationwide.” (Kalendrian et al., 2010)
The EZ Codes created by these researchers are a good benchmark for asking questions to EHR vendors. Do they know of the existence of standardized terminologies? Are they able to integrate and map clinical data to existing vocabularies? Expect the same standard of EHR service that medical professionals have been receiving for years.
- A terminology provides standardized definitions, while an ontology provides relationship mapping between standardized terms.
- Different terminologies and ontologies exist for dental clinical concepts, and your EHR vendor should be able to articulate and provide clear explanations of how their clinical features map to existing standards.
- Diagnostic and procedural information should be linked to enable ease of workflow across the practice and eventually evidence-based decisions.
- The EZ Codes provide mapping between the ICD-9, the ICD-10, SNOMED and the Z codes.
Elsbeth Kalenderian, Rachel L. Ramoni, Joel M. White, Meta E. Schoonheim-Klein, Paul C. Stark, Nicole S. Kimmes, Gregory G. Zeller, George P. Willis, Muhammad F. Walji. (2011) The Development of a Dental Diagnostic Terminology. Journal of Dental Education. January 1, 75 (1): 68-76.