Building the knowledge base for non-combinable codes according to the Korean Standard Classification of Diseases

 
 
 
  • Abstract
  • Keywords
  • References
  • PDF
  • Abstract


    The purpose of this study is to develop a knowledge base for non-combinable combinatorial codes to improve the accuracy of disease classification. We defined the rules related to non-combinable codes according to the list of code pairs proposed by the HIRA and the KCD-7 classification rules. A knowledge base was created according to defined rules and verified. To validate the knowledge base, inpatients who were billed for diabetes mellitus in December 2016 were selected as the subject of the study. As a result, the number of combinatorial codes proposed by the HIRA was 1,195, but the number of code pairs generated in the knowledge base was 25,439. Non-combinable codes by confirming with an indication of the HIRA have discovered 1,391 cases. As a result of verification with the code pair of the proposed knowledge base, 100 combinations were found. Non-combinable codes by confirming with an indication of the HIRA have discovered 1,391 cases. As a result of verification with the code pair of the proposed knowledge base, 3,525 combinations were found. It is meaningful that a convenient authoring tool that can automatically catch combinatorial codes was developed to build a knowledge base.

     

     


  • Keywords


    Coding rules; Insurance claims code; Knowledgebase

  • References


      [1] Lee JH, Shim MS, “The Accuracy of the ICD-10 Code for Trauma Patients Visiting on Emergency Department and the Error in the ICISS”, Journal of Trauma and Injury, Vol. 22, No. 1, (2009), pp.108-115.

      [2] Larkey LS, Croft WB, “Technical Report - Automatic assignment of icd9 codes to discharge summaries”, the University of Massachusetts at Amherst, Amherst, MA., Ph.D. thesis, 1995

      [3] Pakhomov SV, Buntrock JD, Chute CG., “Automating the assignment of diagnosis codes to patient encounters using example-based and machine learning techniques”, Journal of the American Medical Informatics Association, Vol. 13, No. 5, pp.516-525, 2006

      [4] Bae SO, Kang KW, Boo YK, Lee Y, Cheo HS, Choi HY, “A Study on the Difference in Disease Coding of Doctors, Medical Insurance Review Nurses and Medical Record Administrators based on Coding Simulation”, Journal of Health Informatics and Statistics, Vol. 40, No. 3, (2015), pp.161-174.

      [5] Kim MJ, “A Knowledge-based Model for Efficient Classification of Diseases”, Jeju National University, Korea, Doctoral thesis, 2017

      [6] Kim MJ, Kim HC, “Building Rule-based support system for disease code classification”, The e-Business Studies, Vol. 15, No. 4, (2014), pp.61-81.

      [7] Serguei V.S. Pakhomov, JAMES D. Buntrock, Christopher G. Chute, “Automating the Assignment of Diagnosis Codes to Patient Encounters Using Example-based and Machine Learning Techniques”, Journal of the American Medical Informatics Association, Vol. 13, No. 5, (2006), pp.516-525

      [8] Statistics Korea Korean standard classification of disease 7th revision, Korean medical record association, 2015

      [9] Statistics Korea, Classification of disease manual, Statistics Korea,(2015), pp: 3-35


 

View

Download

Article ID: 21017
 
DOI: 10.14419/ijet.v7i3.33.21017




Copyright © 2012-2015 Science Publishing Corporation Inc. All rights reserved.