By Jeff Jianfei Guo, BPharm, PhD, Assistant Professor of Pharmacoepidemiology, University of Cincinnati Medical Center, Cincinnati, OH and Fadia T. Shaya, PhD, MPH, Assistant Professor, University of Maryland Schools of Pharmacy, and Medicine, Dept of Epidemiology and Preventive Medicine, Baltimore, MD

One more time, the "Introduction to Pharmacoepidemiology" short course drew a large number of participants. Over 270 attendees registered for this course at this year's ICPE in France. This course was well received by the audience, especially first time attendees, as reported in the evaluation forms. The course was structured so as to provide a good foundation in principles of epidemiological investigation. Four different lectures, somewhat complementary, covered unique research designs and methodological approaches in pharmacoepidemiology. Dr. Jacques LeLorier presented "Confounding by Indication: a Clinician's Point of View"; Dr. Sean Hennessy presented "Case-Control Studies: Theory and Practice"; Dr. Judith Jones presented "Cohort Study Design for Pharmacoepidemiology"; and Dr. Nicholas Moore presented "Spontaneous Reporting Systems."

The topics were selected so as to provide insights and tips on a wide range of tools commonly used in pharmacoepidemiology. In consideration of the increasing number of retrospective investigations, researchers need to have a good handle on methods that adjust for channeling bias. Confounding is a risk factor for the outcome and correlates with the exposure. It occurs when the estimate of an outcome measurement of association between drug exposure and health status is distorted by the effect of one or several other risk factors. Confounding by indication is likely when patients with certain risk factors are selectively assigned to different treatment groups. Dr. LeLorier discussed several important confounders in pharmacoepidemiology that are often encountered in clinical studies and he used several examples to demonstrate possible confounding by indication from a clinical point of view.

Attendees also learned about the case-control design, a research design that is at the basis of a large portion of studies. Dr. Hennessy used the 2x2 table and design flowchart to explain the case-control study design, and compared case-control studies to cohort studies and randomized clinical trials, using some user-friendly examples. He expanded on the importance of matching on the correct factors, and the pitfalls of over-matching; he compared retrospective and prospective designs, and defined the different approaches based on conditioning on disease versus conditioning on exposure, respectively. Dr Hennessy ended with the bigger perspective. He provided an overview of the major analytic observational study methods used in pharmacoepidemiologic research, and he elaborated on the advantages and limitations of these approaches.

The third section of this introductory course covered the design of cohort studies. Dr. Jones explained the theory supporting cohort studies, and expanded to concrete examples of cohort studies. She provided several sophisticated research examples to demonstrate how to conduct a cohort study from theory to practice. She also emphasized procedures and advantages of using medical databases for pharmacoepidemiologic research.

From the different designs available, to confounding issues, to cohort studies, the course then moved to address research in pharmacovigilance and regulatory issues. After talking about the importance of surveillance and pharmacovigilance, Dr. Moore introduced the European spontaneous reporting systems (SRS) and outlined their strengths and weaknesses. He talked about the latest in pharmacovigilance research and how it relates to regulatory issues. He also discussed the characteristics of adverse reaction reporting systems, factors related for reporting patterns, as well as the objectives of using SRS databases for identifying new problems with newly marketed drugs.

Participants reported that these sessions were very helpful. Since its inception, the "Introduction to Pharmacoepidemiology" course, always offered one day ahead of the meeting, has been widely attended by a diverse audience. It is intended for early to mid-career researchers, as well as those who are interested in getting an overview of methods. Seasoned researchers also like to attend in order to refine some of their own research or teaching techniques. Other attendees report that they benefit from attending this course ahead of the meeting, as a way to prepare for the rest of the program in the meeting. It is precisely this diversity of the audience that makes for a very dynamic course, much enhanced by active participation by attendees. This course has developed as the mainstay of ISPE pre-meetings, and the Education Committee expects to keep it as a service to attendees. The topics introduced are chosen so as to provide a solid basis in an overview of principles, but also touch upon specific topics and techniques. If you have any suggestions or comments, please make sure you forward them to us (fshaya@rx.umaryland.edu).

Return to table of contents