Drug data coding and analysis in epidemiologic studies.

In epidemiologic studies that collect comprehensive information on medication use, the complexity of dealing with a large number of trade and generic names may limit the utilization of these data bases. This paper shows the specific advantage of using two coding systems, one to maximize efficiency of data entry, and the other to facilitate analysis by organizing the drug ingredients into hierarchical categories. The approach used by two large surveys, one in the USA and one in Italy, is described: the Established Populations for Epidemiologic Studies of the Elderly (EPESE) and the 'Gruppo Italiano di Farmacovigilanza nell' Anziano' (GIFA). To enter the medications into a computerized database, codes matching the drug product names are needed. In the EPESE the prescription and over the counter drug products are coded with the Drug Products Information Coding System (DPICS) and the Iowa Nonprescription Drug Products Information Coding System (INDPICS), respectively. The GIFA study uses the coding system of the Italian Ministry of Health (MINSAN), with a unique numeric code for each drug product available in Italy. To simplify the analytical process the drug entry codes are converted into hierarchical coding systems with unique codes for specific drug ingredients, chemical and therapeutic categories. The EPESE and GIFA drug data are coded with the Iowa Drug Information System (IDIS) ingredient codes, and the Anatomical Therapeutic and chemical (ATC) codes, respectively. Examples are provided that show coding of diuretics in these two studies and demonstrate the analytic advantages of these systems.