Barriers and challenges of using medical coding systems

The subject of this study is the generation and utilisation of coded medical data. There are essentially two uses of health data: one is the clinical use within the primary healthcare process. The other use, which is the focus of my thesis, is the central collection of data aggregated from clinical records. These aggregate data can be used for public health research and for policy decisions. A great portion of these data is coded. This thesis concentrates mainly on disease coding, more specifically on the use of the worldwide used International Classification of Diseases (ICD). The coding procedure takes as input some free text extract of the medical record, such as a discharge letter or just the stated clinical diagnoses. This procedure is error prone; therefore the use of coded data in reimbursement, research and policy making requires some precautions. My aim in this work was to better understand the generation and utilisation of coded medical data, and find ways for computers to improve both.