Library and Information Science

Medical records are different from journal articles and books in the sense that the former consists of raw as opposed to processed data, while the latter are the product of ideas processed by authors. This difference effects the characteristics of retrieval. The purpose of this paper is to make several unique characteristics of medical records retrieval clear through comparison with the retrieval methods of journal articles and books on medicine.

Research was carried out on two levels, one, a study of classificatory materials used for retrieving medical records, and two, user behavior in terms of searching for them. ICD-9 (International Classification of Diseases, 1975 Revision), SNOP (Systematized Nomenclature of Pathology), MeSH (Medical Subject Headings), and the NLM classification were examined to check their usage as source material through the classification of the names of ten diseases. Then, doctors and medical record administrators in several university hospitals were interviewed to get information about the methods they used in retrieving the desired information. This research revealed that the medical records are more specific than another kinds of materials in expressing medical concepts.

This paper analyzes the trends of the present retrieval system used for medical records and forecasts future trends in the retrieval of both medical records and other publications, as well as stressing their gradual computerization. It concludes that the biggest problem in the computerization of retrieval systems at the moment is one concerning the coding and inputting of medical terms.