DEVELOPMENT OF CONCEPTS NOMENCLATURE FOR CLINICAL DECISION SUPPORT SYSTEM IN DIAGNOSTICS OF ANGINA PECTORIS

Introduction. Since 2000^th CHD morbidity in Russia is growing constantly, especially, angina pectoris morbidity. Increasing workload on physicians due to high number of patients makes us find the ways to optimize human resource applying. One of the possible ways is development of clinical decision support systems, particularly, ontology-based systems. These ontologies must adopt semantic interoperability for the concepts included, to be clearly interpreted by physicians of different medical schools.

Research target. Develop the concepts nomenclature for building of angina pectoris ontology. Methods. For primary concepts extraction we used state clinical recommendations “Stable coronary heart disease” issued in 2016. Development and filling of the nomenclature were performed in MS Excel and MS Excel Online. The work through the primarily extracted concepts was performed with expert-cardiologists by cross-validation method and common voting on final inclusion of the concept in the nomenclature. The final nomenclature validation was performed by semi-automatic method on anonymized 610 electronic health records of the patients.

Results. After primary extraction of the concepts without experts, we could extract 336 basic concepts and 144 synonyms, keeping in mind the aspects of this nomenclature development. The shared work through this nomenclature with the expert-cardiologists let us increase the number of the basic concepts up to 409 and the number of the synonyms up to 619. The nomenclature validation on electronic health records showed compliance of the concepts in aspects of diagnostics methods and medicine prescriptions and 92% compliance in diagnostic signs, which doctors evaluate in diagnostics results. Percentages of primarily extracted concepts in draft nomenclature and after being worked through by the experts and also validation results based on electronic health records, let us make a conclusion that clinical recommendations can be used as a basic data source for concepts extraction, but the draft result must be evaluated by the experts and also filled up with the usage of additional knowledge sources about angina pectoris, and after that the nomenclature can be used for angina pectoris ontology development.

Conclusion. The usage of clinical recommendations is proved to be the basic data source for concepts of subject area extraction. However, the draft nomenclature must be evaluated and filled up with additional synonyms and validated on electronic health records. After that, the extracted concepts and their synonyms can be used for ontology development.

1. Federal statistics service. Russian statistical yearbook [Internet]. 2017 [cited 2018 Mar 31]. Available from: http://www.gks.ru/bgd/regl/b17_13/IssWWW.exe/Stg/08-12-23.doc. [In Russian]

2. Gusev A.V., Zarubina T.V. Clinical decision support in medical information systems. Physician and information technologies. 2017;2:60–72. [In Russian]

3. Shen Y, Yuan K, Chen D, Colloc J, Yang M, Li Y, Lei K. An ontology-driven clinical decision support system (IDDAP) for infectious disease diagnosis and antibiotic prescription. Artif Intell Med. 2018;86:20–32.

4. Ministry of health of Russian Federation/ Stable coronary heart disease: clinical guidelines [Internet]. 2016 [cited 2018 Apr 3]. Available from: http://cr.rosminzdrav.ru/#!/recomend/133.

5. [In Russian]

6. Stozhok EV. Concept, semantics and meaning. ONV [Internet]. 2011 [cited 2018 Apr 3];(95). Available from: https://cyberleninka.ru/article/n/termin-ponyatie-i-znachenie. [In Russian]