Dental concepts in the Unified Medical Language System

Objective: Currently, no comprehensive, controlled vocabulary for dentistry is available. The objective of this study was to determine how well the Unified Medical Language System, the largest repository of concepts and terms in biomedicine, represents dental concepts. Method and materials: The dental subset of concepts was extracted from Unified Medical Language System using the software program APEX (APplication for the EXtraction of domain-specific concepts). The relationships contained in the Unified Medical Language System Metathesaurus were used to locate the concepts related to 12 seed terms. The Encyclopedie Medico-Chirurgicale (513 dental terms) and the Diagnostic Codes developed by Leake et al (124 terms) were compared to the dental subset. Terms were classified as exact match, related term, or no match. The resultant matching characteristics were compared to those determined by the National Library of Medicine/Agency for Health Care Policy and Research Large Scale Vocabulary Test. Results: The dental subset of Unified Medical Language System contained 948 concepts. The Encyclopedie Medico-Chirurgicale and the Diagnostic Codes exhibited similar matching characteristics for exact match (61.6% and 58.9%, respectively) and related term (38.0% and 32.2%, respectively). For no match, the matching frequencies were significantly different (P < .001). Conclusion: The Unified Medical Language System may be a comprehensive source of terms suitable for various representation requirements in dentistry.