Using secondary data sources for pharmacoepidemiology and outcomes research.

The increased availability of electronic medical records and administrative health care databases is adding to the popularity of pharmacoepidemiology and outcomes research studies. Despite their availability, practitioners may be reluctant to use these databases because they lack familiarity with database research in general. The basic principles of research are the same regardless of the data source, but there are a few special considerations. When using secondary data sources for research purposes, special care must be taken to select an appropriate source to ensure that relevant information is available to answer the research question at hand. Special attention must also be paid to selecting the appropriate codes to represent the outcomes and exposures of interest; therefore, a general understanding of coding schemes is necessary. Although time may be saved by not prospectively collecting data, the process of manipulating the data for analysis in secondary databases can be complex. Analysis of data from secondary sources may require special procedures to overcome the lack of randomization. By familiarizing themselves with these special issues, practitioners can use secondary sources to conduct studies that make valuable contributions to the improvement of patient care.