A standard fragment of EHR relational data model for diabetes mellitus diagnosis

Clinical Decision Support System (CDSS) helps in early detection and diagnosis of diabetes. Case Based Reasoning (CBR)-based CDSS is the applicable technique for these ill formed problems. Building CBR case-base is the most critical challenge. Electronic health record (EHR) is the complete source of cases for CBR. However, mapping EHR into case-base is challenge. This paper proposes a standard relational data model for diabetes diagnosis based on HL7 REVI, EHR and SNOMED CT. This data model will work as an operational data store to extract diabetes specific data elements from EHR, standardize its structure using REVI, standardize its data fields using diabetes diagnosis common data elements, standardize its contents using a standard terminology (SNOMED CT), and organize data in the form of problem-solution form to be suitable for case-base format. This case-base will be used as a knowledge base for a diabetes diagnosis CBR system.