Format: Lectures, group work, and exercises.
Prerequisites: Basic training in epidemiology and biostatistics is assumed.
Equipment: Laptop. All participants will get a short-course STATA license, which needs to be installed on the laptop.
DAY 1
Topics in randomized trials (real ones, not emulated)
This course is designed to provide an overview of the design of clinical trials and will include early phase trials, feasibility/pilot studies, and phase 3 confirmatory trials. We will discuss concurrent vs historical controls, allocation to groups, blinding, choice of primary and secondary outcomes, controlled vs real world/pragmatic settings, hypothesis testing including multiple testing, and the calculation of target sample sizes for trials. The principles introduced will be illustrated with published studies, many from Janet’s own collaborative research. Participants will have the opportunity to engage in practical work in small groups, critiquing extracts from published trials, repeating published sample size calculations in Stata, and presenting conclusions to their peers. Foundational understanding of research methods and statistics will be assumed, but the emphasis will be on understanding principles and not on derivation of formulae.
Instructor
Janet Peacock, MSc, PhD
Department of Clinical Epidemiology, Aarhus University, Denmark
Department of Epidemiology, Geisel School of Medicine, Dartmouth College, USA
DAY 2
Vaccine safety and pharmacovigilance
This course in intended to be an overview of the guidance, practices, and tools involved to ensure the safety of vaccines. It will cover safety surveillance throughout all stages of clinical development and post-authorization use. The characteristics of an ideal vaccine monitoring system will be reviewed, and the differences in systems that exist between countries, largely upon economic lines, will be described.
Weaknesses in the current infrastructure and challenges to the generation of robust evidence in real world data sources will also be explored. Finally, the exciting opportunities afforded by genomics, multi-omics and artificial intelligence to shift the paradigm in vaccine safety science will be introduced.
This course is designed to provide an overview of how the safety of vaccines are monitored through clinical development to implementation in immunization programs.
Topics/elements/study designs covered include AEFI surveillance, vaccine pharmacovigilance, and vaccine pharmacoepidemiology.
Participants will have an opportunity to learn about a topic that is much more complex and nuanced than most believe. No programming or use of software, this day will be more clinically based.
Prerequisites: curiosity, enthusiasm, and an open mind.
Instructor
Rebecca Chandler, MD, Senior Clinical Development Vaccine Safety Lead
Coalition for Epidemic Preparedness Innovations (CEPI), Uppsala, Sweden
DAY 3
Self-controlled study designs
Amongst many ways to account for confounding in observational studies, self-controlled methods have been gaining popularity over recent years. Designs such as the self-controlled case series and the case cross-over make comparisons within individuals, and remove the need to account for between person differences.
This course is designed to introduce the self-controlled case series and case cross-over designs to participants. We will explore both methods, their underlying assumptions, some examples of their application, and more recent methodological developments.
Participants will have an opportunity to gain practical experience in conducting analyses for both study designs and will gain experience in thinking through self-controlled design features in small group discussions. The analytical part of the course will involve the use of Stata and will form ~ a quarter of the course.
A small amount of pre-course reading material will be circulated to participants, and it is assumed that participants are already familiar with basic epidemiological study design (e.g. cohort and case control designs) and the statistics underpinning them.
Instructor
Ian Douglas, MSc, PhD
London School of Hygiene and Tropical Medicine, London, UK
DAY 4
Topics in perinatal pharmacoepidemiology
The course will cover main methodological challenges of studying medication safety in pregnancy, offering solutions through design and analytic approaches. Topics will include manifestations of selection bias, confounding, and information bias in pregnant populations. We will discuss challenges interpreting results when these biases are present and explore ways to circumvent them. The course will combine lectures, small-group discussions, and a group project. Suggested pre-course reading includes 2-4 scientific articles.
Instructors
Irene Petersen, MSc, PhD
Department of Clinical Epidemiology, Aarhus University, Denmark
Research Department of Primary Care and Population Health, University College London, UK
Vera Ehrenstein, MPH, DSc
Department of Clinical Epidemiology, Aarhus University, Denmark
DAY 5
Artificial intelligence in pharmacoepidemiology: exploring current applications with a focus on causal machine learning and generative AI
Participants will explore the importance of AI literacy, understand stakeholder perspectives on the use of AI in pharmacoepidemiology, and delve into specific applications. Key topics covered include the academic and regulatory perspectives on AI, causal machine learning (with both theoretical and practical tutorials), and generative AI, illustrated through research examples. The course will involve programming and the use of the statistical software R. R programming experience is helpful but not mandatory. Pre-course preparation will include a brief reading list to familiarize participants with foundational concepts.
Instructor
Maurizio Sessa, MPharm, PhD
Department of Drug Design and Pharmacology, Pharmaceutical Informatics, University of Copenhagen, Denmark
Format: Lectures, group work, and exercises.
Prerequisites: Basic training in epidemiology and biostatistics is assumed.
Equipment: Laptop. All participants will get a short-course STATA license, which needs to be installed on the laptop.
DAY 1 & DAY 2
Pharmacoepidemiology and comparative effectiveness research
The aims of this class are to understand the unique research challenges and opportunities presented when conducting non-experimental studies of medical interventions and identify the specific tools required to design, analyze, and evaluate the validity of these studies. This course will cover a variety of topics central to the conduct of pharmacoepidemiology (PE) and comparative effectiveness research (CER), including (but not limited to): a review of commonly used healthcare databases for PE/CER, an overview of the major methodological challenges in evaluating the benefit and harm of medical interventions using non-experimental methods, an introduction to the study design and analytic solutions to address these challenges (including the target trial emulation framework), and other special topics. The participants will work in groups through a series of structured breakout sessions that will address the full spectrum of issues relevant to the conduct of a PE or CER study from research question development/feasibility assessment through the analytic approach and presentation of results.
Instructors
Til Stürmer, MD, MPH, PhD
University of North Carolina at Chapel Hill
Jennifer Lund, PhD
University of North Carolina at Chapel Hill
DAY 2 & DAY 3
Meta-analysis
This class will give a comprehensive overview of topics in meta-analysis. Most standard topics in meta-analysis will be covered, such as risk of bias analysis, fixed versus random effect models, heterogeneity, meta-regression, and differences between meta-analysis of randomized trials and meta-analysis of observational studies. Core elements will also be covered in hands-on practical sessions. These sessions will be taught in R and STATA. After the class, the participants should be able to perform a standard meta-analysis by themselves.
Instructors
Olaf M. Dekkers, MD, PhD
Leiden University Medical Center
Saskia le Cessie
Leiden University Medical Center
DAY 5
Design of pharmacoepidemiological studies
In this class, participants will be asked to design a specific pharmacoepidemiologic study, address strengths and limitations of the proposed design, and present the results in the forum. The participants will randomly be assigned to a group, and each group will complete its project and lead the discussion on its specific aspects.
Instructor
Irene Petersen, MSc, PGCert, P. Grad Dip, PhD
University College London and Aarhus University
Format: Lectures, group work, and exercises.
Prerequisites: Basic training in epidemiology and biostatistics is assumed.
Equipment: Laptop. All participants will get a short-course STATA license, which needs to be installed on the laptop.
DAY 1
Pruning the research question (half day)
This class aims to teach the principles of thinking about ‘finding and refining a research question’. After a lecture on the principles, participants will have the opportunity to try out crafting a research question in an exercise based on data from an investigation that happened in real life. It is assumed that participants know general principles of epidemiology and understand general ideas about disease causes. There is no programming or calculation involved.
Instructor
Jan Vandenbroucke, MD, PhD
Leiden University Medical Center, Aarhus University, and London School of Hygiene and Tropical Medicine
Study design in infectious disease epidemiology (half day)
This class covers an introduction to infectious disease epidemiology, and to the use of the case-control study design specifically applied to infectious diseases. The aim is to understand what is specific to infectious disease epidemiology. After the introductory lecture, participants will make an exercise in study design on risk factors for infections with antibiotic-resistant bacteria.
Instructor
Christina Vandenbroucke-Grauls, MD, PhD
Amsterdam University Medical Centers and Aarhus University
DAY 2
Validation studies in routinely-collected data (half day)
Causal interpretation of associations is warranted once alternative explanations are credibly ruled out. Information bias stemming from measurement error is one such alternative explanation. Epidemiologic studies based on routinely-collected data have the benefit of hypotheses-free data accumulation, the flip side of which is unknown amount of measurement error. Using lectures and practical group work (Word/Excel), this workshop will cover basics of validity measures, gold standard, conducting validation studies, prioritizing and transportability of event-finding algorithms, and implications for interpretation.
Instructor
Vera Ehrenstein, MPH, DSc
Aarhus University
Fundamental concepts and current thinking about studying interactions (half day)
In this class, you will first learn how interaction is thought about from a statistical and epidemiological point of view, and how these views differ. Next you will make a hands-on exercise on calculating interactions, based on a real life paper about a gene-drug interaction (a simple calculator suffices, e.g., on your phone with a button for logarithms). Thereafter, current thinking about interaction, according to component cause theory and counterfactual theory, will be explained - based on the same example.
Instructor
Jan Vandenbroucke, MD, PhD
Leiden University Medical Center, Aarhus University, and London School of Hygiene and Tropical Medicine
DAY 3
Missing data and multiple imputation
After a discussion of missing data mechanisms, the course will review ad-hoc methods of dealing with missing data and focus on theory, practice, and reporting of studies using multiple imputation. Computer exercises will be based on Stata.
Instructor
Irene Petersen, PhD
University College London and Aarhus University
DAY 4
Self-controlled study designs
Amongst many ways to account for confounding in observational studies, self-controlled methods have been gaining popularity over recent years. Designs such as the self-controlled case series and the case cross-over make comparisons within individuals, and remove the need to account for between person differences.
This course is designed to introduce the self-controlled case series and case cross-over designs to participants. We will explore both methods, their underlying assumptions, some examples of their application, and more recent methodological developments.
Participants will have an opportunity to gain practical experience in conducting analyses for both study designs and will gain experience in thinking through self-controlled design features in small group discussions. The analytical part of the course will involve the use of Stata and will form ~ a quarter of the course.
A small amount of pre-course reading material will be circulated to participants, and it is assumed that participants are already familiar with basic epidemiological study design (e.g. cohort and case control designs) and the statistics underpinning them.
Instructor
Ian Douglas, MSc, PhD
London School of Hygiene and Tropical Medicine
DAY 5
Case studies of vaccine safety, including emergency epidemiology (half day)
This final session will solidify the previous material on study designs by showcasing real-world projects applying different study designs to study adverse events following HPV vaccination and COVID-19 vaccination, respectively. Traditional cohort designs, self-controlled designs, and time trend studies will be discussed, together with the possibility of doing rapid “emergency epidemiology” studies based on near-real-time medical databases.
Instructor
Reimar W. Thomsen, MD, PhD
Aarhus University
Benefits and risks of treatments are only interpretable in the context of treated patients with a given disease. In this year’s series of master classes, we start by showcasing pharmacoepidemiology challenges and methods in specific diseases and finish by the popular meta-analysis course.
Format: Brief overview lectures followed by hands-on exercises and group work rooted in real studies.
Prerequisites: basic training in epidemiology and biostatistics is assumed.
Equipment: Laptop. All participants will get a short-course STATA license for the duration of the course, which needs to be installed on laptop.
DAY 1
Get real: crash course in pharmacoepi-methods (half day)
Epidemiologic study design, prescription data, and sources of bias in pharmacoepidemiology. Confounding by indication, healthy user effects.
Sugar, statins and other stories (half day)
Brief overview of diabetes/obesity/cardiovascular disease epidemiology and treatment. Studying comparative effectiveness and drug safety.
Instructor:
Reimar W. Thomsen, MD, PhD, Aarhus University
DAY 2
Cancer epidemiology in the era of precision medicine (half day)
Treatment patterns and recurrence, immortal time bias, targeted therapies.
Instructors:
Deirdre Cronin-Fenton, PhD, Aarhus University
Mette Nørgaard, MD, PhD, Aarhus University
Drugs and biomarkers in renal diseases (half day)
Longitudinal data, defining outcomes using laboratory values, longitudinal data with repeated measurements.
Instructor:
Christian Fynbo Christiansen, MD, PhD, Aarhus University
DAY 3
Perinatal pharmacoepidemiology (half day)
Design, analysis, and biaas as applied to studies of drug treatment in pregnancy.
DUS and Don’ts (half day)
Drug utilization and risk minimization studies. Interrupted time series regression. Validation studies of routinely collected data.
Instructors:
Irene Petersen, PhD, University College London and Aarhus University
Vera Ehrenstein, MPH, DSc, Aarhus University
DAY 4
Meta-analysis (part 1)
The course will give a comprehensive overview of topics in meta-analysis. Most standard topics in meta-analysis will be covered, such as risk of bias analysis, fixed versus random effect models, heterogeneity, meta-regression and differences between meta-analysis of randomized trials and meta-analysis observational studies. Core elements will also be covered in hands-on practical sessions. These sessions will be provided in R and STATA. After the course, the students should be able to perform a standard meta-analysis by themselves.
Instructors:
Olaf M. Dekkers, MD, PhD, Leiden University Medical Center
Saskia le Cessie, PhD, Leiden University Medical Center
DAY 5
Meta-analysis (part 2)
The course will give a comprehensive overview of topics in meta-analysis. Most standard topics in meta-analysis will be covered, such as risk of bias analysis, fixed versus random effect models, heterogeneity, meta-regression and differences between meta-analysis of randomized trials and meta-analysis observational studies. Core elements will also be covered in hands-on practical sessions. These sessions will be provided in R and STATA. After the course, the students should be able to perform a standard meta-analysis by themselves.
Instructors:
Olaf M. Dekkers, MD, PhD, Leiden University Medical Center
Saskia le Cessie, PhD, Leiden University Medical Center
The 2021 Pharmacoepidemiology Summer School was cancelled due to COVID-19. See the 2020 program below.
DAY 1
Get real: crash course in pharmacoepi-methods (half day)
Epidemiologic study design, prescription data and sources of bias in pharmacoepidemiology. Confounding by indication, healthy user effects.
Sugar, statins and other stories (half day)
Brief overview of diabetes/obesity/cardiovascular disease epidemiology and treatment. Studying comparative effectiveness and drug safety.
Instructor:
Reimar W. Thomsen, MD, PhD, Aarhus University
DAY 2
Cancer epidemiology in the era of precision medicine (half day)
Treatment patterns and recurrence, immortal time bias, targeted therapies.
Instructors:
Deirdre Cronin-Fenton, PhD, Aarhus University
Mette Nørgaard, MD, PhD, Aarhus University
Drugs and biomarkers in renal diseases (half day)
Longitudinal data, defining outcomes using laboratory values, longitudinal data with repeated measurements.
Instructor:
Christian Fynbo Christiansen, MD, PhD, Aarhus University
DAY 3
Bone disease and device safety (half day)
Measuring adherence and persistence of long-acting agents. Health Technology Assessment. Safety of devices.
Instructor:
Alma B. Pedersen, MD, PhD, DMSc, Aarhus University
DUS and Don’ts (half day)
Drug utilization and risk minimization studies. Interrupted time series regression. Validation studies of routinely collected data.
Instructors:
Irene Petersen, PhD, University College London and Aarhus University
Vera Ehrenstein, MPH, DSc, Aarhus University
DAY 4
Meta-analysis (part 1)
The course will give a broad overview of topics in meta-analysis. Most standard topics in meta-analysis will be covered, such as risk of bias analysis, fixed versus random effect models, heterogeneity, differences between meta-analysis of randomized trials and meta-analysis observational studies. Core elements will also be covered in hands-on practical sessions (risk of bias analysis, statistical analysis). After the course, the students should be able to perform a standard meta-analysis by themselves. The course will briefly cover some more advanced topics (network meta-analysis, combining different metrics).
Instructor:
Olaf M. Dekkers, MD, PhD, Leiden University Medical Center
DAY 5
Meta-analysis (part 2)
Instructor:
Olaf M. Dekkers, MD, PhD, Leiden University Medical Center
Advanced topics in pharmacoepidemiology (2 days)
Professor M. Alan Brookhart
University of North Carolina at Chapel Hill
Self-controlled study designs (1 day)
Professor Ian Douglas
London School of Hygiene and Tropical Medicine
Perinatal and reproductive pharmacoepidemiology (1 day)
Professor Irene Petersen
University College London and Aarhus University
Professor Vera Ehrenstein
Aarhus University
Reporting our research using routinely-collected health data better- introducing STROBE, RECORD and RECORD-PE (half day)
Professor Sinead Langan
London School of Hygiene and Tropical Medicine
Pharmacoepidemiology and Comparative Effectiveness Research (2 days)
Professor Til Stürmer & Assistant Professor Jennifer Lund
University of North Carolina at Chapel Hill
Meta-Analysis (2 days)
Professor Olaf M. Dekkers
Leiden University Medical Center and Aarhus University
Group Work: Designing a Pharmacoepidemiologic Study (half day)
Professor Irene Petersen
University College London and Aarhus University
Professor Vera Ehrenstein
Aarhus University
Advanced Topics in Pharmacoepidemiology (2 days)
Professor Alan Brookhart, BA, MAMS, MA, PhD
University of North Carolina
Dr. Leah Sirkus McGrath, BS, MHS, PhD
NoviSci, LLC
Missing Data Issues (1 day)
Professor Irene Petersen, MSc, PGCert, P. Grad Dip, PhD
University College London and Aarhus University
Self-controlled Designs (1 day)
Associate Professor Ian Douglas, BSc, MSc, PhD
London School of Hygiene and Tropical Medicine
Perinatal and Reproductive Pharmacoepidemiology (half day)
Professor Irene Petersen, MSc, PGCert, P. Grad Dip, PhD
University College London and Aarhus University
Professor Vera Ehrenstein, MPH, DSc
Aarhus University
Introduction to Pharmacoepidemiology (half day)
Professor John Acquavella
Aarhus University, Aarhus, Denmark
Questions to Consider Regarding the Application of the Results of a RCT of the Efficacy of Therapy to the Management of Patients (half day)
Professor Noel Weiss
University of Washington, Seattle, USA
Pharmacoepidemiology and Comparative Effectiveness Research (2 days)
Professor Til Stürmer & Assistant Professor Jennifer Lund
University of North Carolina, Chapel Hill, USA
Meta-Analysis (2 days)
Professor Olaf M. Dekkers
Leiden University Medical Center, Leiden, The Netherlands and Adjunct Professor at Aarhus University, Aarhus, Denmark
Introduction to Pharmaceutical Risk Management (half day)
Visiting Professor John Acquavella
Aarhus University, Denmark
Pruning the Research Question (half day)
Professor Jan P. Vandenbroucke
Leiden University Medical Center, The Netherlands and Aarhus University, Denmark
Pharmacoepidemiology and Comparative Effectiveness Research (2 days)
Professor Til Stürmer & Assistant Professor Jennifer Lund
University of North Carolina, USA
Visiting Professor and Visiting Scientist, respectively, at Aarhus University, Denmark
Systematic Review and Meta-Analysis (2 days)
Associate Professor Olaf M. Dekkers
Leiden University Medical Center, The Netherlands and Honorary Professor at Aarhus University, Denmark
Issues in study design (2 days)
Randomized clinical trials
Associate professor Vera Ehrenstein, MPH, DMS
Power and sample size calculation
Associate professor Anders H. Riis, MSc
Cohort studies
Clinical associate professor Christian Fynbo Christiansen, MD, PhD
Issues in randomization and hypothesis testing
Consultant Søren Paaske Johnsen, MD, PhD
Case-control studies
Associate professor Deirdre Cronin-Fenton, BSc, PhD
Clinical associate professor Mette Nørgaard, MD, PhD
Cross-sectional studies
PhD student Morten Schmidt, MD
Drug safety (1 day)
Professor Tom MacDonald, BSc, MD, FRCPE, FRCPSG, FESC, FISPE, FBPharmacolS
Systematic review and meta-analysis (2 days)
Associate professor Olaf M. Dekkers, MD, MSc, MA, PhD
