Standardizing Biochemistry Dataset for Medical Research

Abstract

Harnessing clinical datasets from the repository of electronic health records for research and medical intelligence has become the norm of the 21st century. Clinical datasets present a great opportunity for medical researchers and data analysts to perform cohort selections and data linkages to support better informed clinical decision-making and evidence-based medicine. This paper utilized Logical Observation Identifiers Names and Codes (LOINC®) encoding methodology to encode the biochemistry tests in the anonymized biochemistry dataset obtained from the Health Informatics Centre (HIC) at the University of Dundee. Preliminary results indicated that the encoded dataset was flexible in supporting statistical analysis and data mining techniques. Moreover, the results indicated that the LOINC codes cover most of the biochemistry tests used in National Health Service (NHS) Tayside, Scotland.