Center for Pharmacoepidemiology
Please visit the new home of the Center for Pharmacoepidemiology at http://pharmacoepi.unc.edu
The Center for Pharmacoepidemiology (the “Center”) is a unique partnership between the Department of Epidemiology in the UNC Gillings School of Global Public Health, the leading public school of public health in the U.S, and notable non-academic stakeholders. As a recognized academic leader in epidemiology and pharmacoepidemiology, UNC and the Center provide a unique forum to generate innovative, evidence-based solutions to challenges related to comparative effectiveness research and evaluation of the safety of drugs and devices in real-world clinical settings.
Initiated through a grant from GlaxoSmithKline in 2002, the Center has a history of fostering collaborative research between academia and industry partners who share a commitment to developing robust methodologies and generating valid evidence using observational data on real-world patients from a variety of sources.
The mission is to advance the field of pharmacoepidemiology through innovation in methods, application of those methods to critical issues in the field, and training of future leaders.
Message from the Director:
I welcome all those interested in the Center for Pharmacoepidemiology to browse our pages. Initiated through a grant from GlaxoSmithKline in 2002, the Center is currently expanding its collaborations across the pharmaceutical industry, government, and regulatory agencies. This expansion is very timely, because the need to evaluate the intended and unintended effects of drugs in those actually taking these drugs rather than only in small, selected populations, is increasingly recognized (see, e.g., the 2007 FDA Amendment Act (FDAAA) and the 2009 Institute of Medicine report on Comparative Effectiveness Research). Pharmacoepidemiologists are at the center of generating such evidence focusing on benefit to harm evaluations for populations and defined subpopulations. Epidemiologic studies of medications are often criticized because of their nonexperimental study design. While I certainly disagree with the notion that we can only learn from randomized trials, I think we can all agree that continued developments and application of methods to address various biases in both experimental and nonexperimental studies are essential to improve our understanding of what works in current medical practice and what does not. To achieve this, we need to train researchers within and outside of academia, that is all those implementing and interpreting pharmacoepidemiologic studies, in state-of-the art methods. This is the mission of the Center and I am very proud to be its director. The Center helps the UNC pharmacoepidemiology program to attract the best faculty and students. This creates synergy between the Center and the academic program. I hope these pages provide a good overview over the Center’s vision, mission, and past and current activities and I am personally interested in your feedback.
- Til Stürmer
A number of issues in pharmacoepidemiology present challenges to the validity and interpretation of research in the field and warrant considerable scholarly attention. In addition to student research, the Center provides an opportunity to explore innovative research avenues in more conceptual areas, for example:
- Coding of health outcomes in large automated (claims) databases and clinical data (electronic medical records) for pharmacoepidemiologic research
- Adverse drug events monitoring using large automated databases
- Comparative Effectiveness Research
Databases with which we routinely work include:
- Medicare Parts A, B, and D
- Surveillance, Epidemiology and End Results (SEER)-Medicare
- Medicare Current Beneficiary Survey
- Truven Health Analytic’s MarketScan Commercial Insurance Claims, Medicare Supplementary, and Supplemental Laboratory Data
- U.S. Renal Data System linked to DaVita clinical data warehouse
- Atherosclerosis Risk in Communities Study (ARIC), Women’s Health Initiative Study (WHI), both linked with Medicare data
UNC is part of the U.S. Agency for Healthcare Research and Quality’s (AHRQ) DEcIDE (Developing Evidence to Inform Decisions about Effectiveness) network. The DEcIDE network aims to generate valid scientific evidence about outcomes, comparative clinical effectiveness, and safety and appropriateness of healthcare interventions and services. The UNC DEcIDE center includes the University’s schools of pharmacy and public health. Past projects included comparing treatment strategies in patients with depression, advanced colorectal cancer, early prostate cancer, and end-stage renal failure.
The Center serves as a unique source of scholars and practitioners in pharmacoepidemiology trained at the doctoral level. Students can gain practical experience in the use of healthcare databases and appreciate their strengths and limitations. Focused course work on issues distinct to pharmacoepidemiology supplements the excellent curriculum in epidemiologic and biostatistical methods currently provided to doctoral students in the Epidemiology department.
The Center includes a number of pre-doctoral trainee and post-doctoral fellow positions. In addition to weekly mentored research hours, doctoral students participate in the pharmacoepidemiology/healthcare stream of the epidemiology program, write the pharmacoepidemiology/healthcare qualifying exam at the end of their second year, and engage in pharmacoepidemiology-related dissertation research. Methodologic short courses and seminars are available for all trainees and affiliated industry researchers depending on the availability of instructors and interests.
Past seminars/courses have covered topics such as:
- Advanced pharmacoepidemiologic methods
- Pharmacoepidemiology for Global Health
- Longitudinal data analysis
- Introduction to Pharmacology
- Molecular epidemiology
- Strategies to Control For Confounding: Beyond Propensity Score
- Medicare Part D
Since its inception under Dr. Harry Guess, the Center has attracted world-class faculty members and doctoral students. Our current core of six faculty have unparalleled expertise in pharmacoepidemiology and comparative effectiveness research, including advanced methods, meta-analysis of healthcare data and risk assessment. Our students are selected for the combination of scientific, intellectual and management skills that will make them future leaders in all fields of healthcare, including research oriented pharmaceutical companies.
M. Alan Brookhart, PhD, is associate professor of epidemiology. His methods-oriented healthcare epidemiologic research focuses primarily upon the development and application of statistical methods and study designs for comparative effectiveness research using large healthcare utilization databases. He has made significant contributions to the development of instrumental variable approaches that can be used to estimate causal effects in the presence of unmeasured or poorly recorded confounding variables. He has also been involved with the development of propensity score and marginal structural model methodology and new epidemiologic approaches for studying adherence and use of healthcare services. He is interested in comparative effectiveness of medications in the elderly, and in patients with inflammatory conditions and end-stage renal disease.
Michele Jonsson-Funk, PhD, is research assistant professor of epidemiology. She is a pharmacoepidemiologist with a particular interest in epidemiology methods. Her work focuses on evaluating doubly robust estimators and methods for detecting treatment effect heterogeneity. In addition to her portfolio of methods-focused research, she directs ongoing research studies using large administrative claims data. These projects address the safety and effectiveness of treatments for a range of conditions including gestational diabetes, medical devices for stress urinary incontinence and pelvic organ prolapse, prevention of unintended pregnancy, hypertension, and hyperlipidemia. Dr Jonsson Funk is also director of the Methods Core at the Center for Women’s Health Research.
Jennifer Lund, PhD, is assistant professor of epidemiology. Dr. Lund is a cancer health services researcher and pharmacoepidemiologist specializing in the use of large administrative healthcare databases to evaluate the uses, benefits and harms of medical interventions, specifically in older adults. She has amassed experience in validating claims-based algorithms for identifying cancer therapies using data sources from the United States and Denmark and is currently working to improve claims-based measures of comorbidity and frailty among older adults diagnosed with cancer. She is particularly interested in applying these novel measures to improve the validity of comparative effectiveness research in the areas of gastrointestinal cancers, hematologic malignancies, and inflammatory bowel disease.
Charles Poole, ScD, is associate professor of epidemiology. Dr. Poole has more than 30 years’ experience as an epidemiologic researcher and methodologist. He has published, taught and lectured widely on a broad range of applied and theoretical topics on research methods in nearly every substantive area of epidemiologic research. His areas of special expertise include study design, data collection, statistical analysis, risk assessment, systematic review and meta-analysis.
Til Stürmer, MD, MPH, PhD, is director of the Center for Pharmacoepidemiology and professor/head of the pharmacoepidemiology program in the Department of Epidemiology. Dr. Stürmer has a dual focus in epidemiologic methods and clinical epidemiology. His research in epidemiologic methods includes the development of more efficient matching strategies in genetic epidemiology; measurement error correction methods in case-control studies; the value of propensity scores for pharmacoepidemiologic studies; and the novel integration of ideas from measurement error correction and propensity score analysis to improve control for confounding by medication-use choices using data from validation studies. His research in clinical epidemiology has covered many topics, including nephro- and cardiovascular toxicity of paracetamol and non-steroidal anti-inflammatory drugs; chemoprevention of colorectal cancer and dementia with non-steroidal anti-inflammatory drugs; determinants of antibiotic resistance; and hormone therapy, pain and sub-clinical inflammation in osteoarthritis. More recently he has been assessing relative benefits and harms of antidiabetic treatments.
Alice White MSPH, PhD, is visiting professor of pharmacoepidemiology in the Epidemiology Department. From 2001 to 2011, she was vice president of the Worldwide Epidemiology Department at GlaxoSmithKline, where she headed a global department of 75 epidemiologists, analysts and support staff, with headquarters in North Carolina and offices in Philadelphia, London, Singapore, Tokyo and Rio de Janeiro. She was responsible for research and development epidemiologic strategy and use of observational data across the development continuum to support disease understanding and decision-making, and for the design and management of large post-marketing safety studies, including GSK’s prospective pregnancy registries. She was a member of the Orange County, N.C., Board of Health from 2000 to 2006, and its chair from 2003 to 2004. She has authored numerous publications and scientific presentations, primarily in the cardiovascular and HIV epidemiology areas.
Each year, the Center’s core faculty selects a few outstanding doctoral students from the broader pharmacoepidemiology program to support directly. We are proud of these students and list their names and research interests here.
Currently funded students:
- 1st year PhD student
- MPH, Epidemiology,
- Hometown: Boston, Massachusetts
- Research Interests: Antidiabetics, TBD
- 4th year PhD student
- MSPH epidemiology
- Hometown: Greensboro, North Carolina
- Research interests: pregnancy, perinatal, and pediatric pharmacoepidemiology
Recent research topics funded by the Center:
- Antiretrovirals and cardiovascular outcomes in HIV patients
- Validation of codes for myocardial infarction in Medicaid
- Estimation of adverse events associated with the use of methotrexate in the treatment of ectopic pregnancy
- Development of methods to estimate gestational age in claims databases
- Incretin and pancreatic cancer incidence
- Effects of code aggregation on automated propensity score variable selection algorithms
- Disease Risk Scores
Since 2008, the UNC Student Chapter of the International Society for Pharmacoepidemiology has sponsored the Harry Guess Memorial Lecture. This lecture series was conceived of by UNC students to honor Dr. Guess, a man who built the pharmacoepidemiology program at UNC, and to remember his outstanding dedication to teaching and mentoring. We hope that this lecture series continues to be a moment where we can pay tribute to our roots as a program and Dr. Guess’ unique and lasting contribution.
Upcoming speaker: Watch for the 2015 program
Department of Epidemiology – Healthcare Epidemiology and Pharmacoepidemiology Program area
International Society for Pharmacoepidemiology (ISPE) – The professional association that promotes pharmacoepidemiology and hosts an annual conference
Pharmacoepidemiology and Drug Safety – a journal focused on pharmacoepidemiology research