Semantic Mapping Model between ICD-10 and SNOMED-CT Health Terminologies

Abstract

The International Statistical Classification of Diseases and Health Problems and the Systematized Nomenclature of Medicine are terminologies that aim for data transparency. Terminologies have differences in their compositions, and a mapping between these terms is necessary in order to obtain meaning, seeking to improve the daily life of health professionals with their patients through a model that structures the information in a comprehensive syntactic and semantic way. The goal of this research is to develop a model for semantic mapping between these health terminologies. This is exploratory research, a case study carried out at the Hospital das Clinicas da Faculdade de Medicina de Marilia, which provided the International Statistical Classification of Diseases and Health-Related Problems codes recorded in the medical records for the mapping, aiming to migrate the stored data that were in a relational database to an international structure and data sharing network. The results showed that there are four types of situations during the mapping process: semantic accuracy between terminologies, use of expressions that make the health condition generic, terms that are not exactly equivalent but are semantically close, and a variety of terms to represent a single health condition. It is concluded that it is possible to develop a replicable model that preserves the semantic layer of terms between the International Statistical Classification of Diseases and Health Problems and the Systematized Nomenclature of Medicine.