Medicaid Billing Data Used to Study the Effects of Marketed Drugs

Over the past 12 years, the Food and Drug Administration (FDA) has provided funding for the development of The Computerized Online Medicaid Pharmaceutical Analysis and Surveillance System (COMPASS^R) as a research tool for pharmacoepidemiology studies. Since 1985, these resources have been supplemented by a cooperative agreement to the University of Pennsylvania to perform COMPASS^R studies in support of FDA's regulatory mission. In the process, the system itself has developed considerably as a research tool, and we have learned a good deal about the strengths and limitations of this type of database for pharmacoepidemiology studies. A description of the dataset is presented, followed by a description of the logistics of how we perform such studies, a very abbreviated example of one study performed — a study of drug-induced neutropenia, and then a brief discussion of selected methodologic caveats in using Medicaid data for pharmacoepidemiology research is provided. We look forward in the future to continue to expand our knowledge of the appropriate use of this important pharmacoepidemiology resource, while using it to address questions of clinical importance