The Biomedical and Health Data Science Collaborative (BHDSC), a cross-disciplinary group formed by the Tufts Clinical and Translational Science Institute (CTSI) and Institute for Clinical Research and Health Policy Studies (ICRHPS) at Tufts Medical Center, invites you to attend a virtual seminar on Wednesday, January 25, from 2:00PM-3:00PM. Jeremy P. Brown, PhD will give a talk titled “The Elephant in the Room: Unmeasured Confounding in Pharmacoepidemiology”.
Unmeasured confounding is a major challenge to causal inference in pharmacoepidemiological research, as it can introduce bias and lead to incorrect conclusions. In this presentation, I will discuss methods for diagnosing, reducing, and quantifying unmeasured confounding in pharmacoepidemiological studies. These methods include negative controls, the high-dimensional propensity score, self-controlled study designs, and quantitative bias analysis. I will provide examples of these methods in action, as applied to two non-interventional studies conducted using UK electronic health records: one investigating the effect of proton pump inhibitors on mortality, and the other examining the effect of fluoroquinolones on aortic aneurysm or dissection. By using methods such as these, we can improve the accuracy and reliability of our findings and make more informed decisions about the safety and effectiveness of medications.
Jeremy Brown is a Postdoctoral Research Fellow in the CAUSALab at Harvard T.H. Chan School of Public Health. His research focuses on methods, and in particular causal inference methods, in pharmacoepidemiology and their application to understanding the safety and effectiveness of medications using insurance claims and electronic health records data. Prior to Harvard T.H. Chan he conducted his PhD in pharmacoepidemiology at the London School of Hygiene and Tropical Medicine.
Wednesday, January 25, 2023
Zoom Link: https://wellforce.zoom.us/j/91467788400?pwd=U2FXdTZRZENkdzdnQTYxczlWVCtFdz09&from=addon
Feel free to pass on to others who may be interested.