SCEM: Australian Consortium for Classification Development (ACCD) PhD Scholarship

Abstract

There are a number of Clinical Classification & Coding and Terminology systems that are developed as international standards and adopted by various countries including Australia.

These include (but not limited to):

  • International Classification of Diseases (ICD),
  • International Classification of Functioning, Disability and Health (ICF)
  • International Classification of Health Interventions (ICHI)  developed and maintained by WHO-World Health Organisation, and
  • Systematized Nomenclature of Medicine-Clinical Terminology (SNOMED-CT) developed and maintained by IHTSDO-International Health Terminology Standards Development Organisation.

These classification and terminology systems are developed independently to each other and expected to work in harmony within clinical software. However due to lack of harmonisation between these clinical terminology and classification systems, there is much human involvement required in translating from one to another. Such human involvement creates inconsistencies and errors in coding.

With wider usage of software for management of health data, there is an urgent need for having computer-assisted harmonisation between these classification and terminology systems. Further these are evolving ontologies. Therefore, there is a greater need for keeping the ontological continuity as the core classification or terminological systems evolve.

This research is focused on involving machine learning techniques for the purpose extracting relevant ICD, ICHI and ICF codes, based on SNOMED-CT terms found in free text data that relates to a particular episode of care. Also other information such as

  • previous mapping of similar cases to ICD, ICHI or ICF codes
  • other non-clinical (age, sex) and historical data (chronic diabetes) relevant to the episode of careclinician or facility where patient was treated, etc.,  would require to be considered in improving the accuracy the algorithm.