Pre-Conference Workshops

Session Responsibles:

Leader(s): Yu-Han Chiu

Co-leader(s): Shuyuan Yang

Coordinator(s): Therese Haugdahl Nøst

Short Description:

This pre-conference course introduces the target trial emulation framework, a structured approach for drawing causal inferences from observational data by explicitly designing the study as if it were a randomized trial. Participants will learn how to specify the protocol of a hypothetical target trial and then emulate that protocol using real-world data.

Through applied examples and interactive discussion, the course will highlight common sources of bias in observational research (e.g., immortal time bias, selection bias, and time-varying confounding) and demonstrate how careful alignment between design and analysis can mitigate these threats. We will also discuss practical implementation strategies using longitudinal data.

The course is designed for clinical researchers, epidemiologists, biostatisticians, and data scientists seeking to strengthen the rigor and transparency of causal inference in real-world evidence research.

Learning Outcomes:

By the end of the session, participants will be able to:

1. Formulate a clearly defined causal question in target trial terms;
2. Translate that question into an explicit trial protocol; and
3. Identify key design and analytic decisions necessary to validly estimate causal effects from observational data.

Session Responsibles:

Leader(s) – Speaker(s): Heidi Taipale and Pia Vattulainen

Coordinator(s): Nhung Trinh and Jacqueline Cohen

Assistant with R practical: Jinmei Chen

Short Description:

PRE2DUP is an advanced method for generating drug use periods from drug purchases and PRE2DUP‑R is the new open-source R-implementation of it. PRE2DUP-R estimates drug use periods from drug purchases, based on dispensed amount and sliding average of daily dose. It utilizes package-level parameters, hospital care periods and personal purchase regularity over time. PRE2DUP-R can be applied to all medications, any drug form and different dispensing regulations.

A key requirement for using PRE2DUP‑R is the construction of package‑level parameters—often the main challenge when applying the method. The course includes short presentations introducing the PRE2DUP‑R and package-level parameters, followed by hands‑on exercises. Participants will work in small groups (5–10 people) to design package‑level parameters, run the PRE2DUP‑R program in R, troubleshoot common errors, and explore advanced features. Exercises will use simulated Nordic register‑based data. Each group will present their solutions to the full group.

Learning Outcome(s):

By the end of the session, participants will learn:

1. The fundamentals of parameter design;
2. How to apply these parameters in practice; and
3. How to approach special cases and troubleshooting scenarios.

Requirement(s):

Participants should have a basic understanding of pharmacoepidemiology and/or Nordic prescribed drug register data.

A laptop with R and RStudio installed is required. Participants should also install the R packages:

• devtools from CRAN
• PRE2DUP-R from GitHub

Installing instructions: https://piavat.github.io/PRE2DUP-R/articles/introduction.html

Throughout the course, participants will have an excellent opportunity to receive guidance directly from the developers of PRE2DUP and PRE2DUP‑R.

Pre‑reading: Introduction to PRE2DUP-R

https://pubmed.ncbi.nlm.nih.gov/41689954

https://piavat.github.io/PRE2DUP-R/index.html

Program:

13:00-13.15                        Presentation 1: Brief introduction to basic concepts of PRE2DUP‑R
Pia Vattulainen

13:15-13.30                        Presentation 2: Basics of package‑level parameter design
 Heidi Taipale

13:30-14:30                        Group Work 1, followed by review session:

Designing package parameters for specific drugs or drug classes

14:30-14:45                Coffee Break

14:45-15:00                        Presentation 3: How to run PRE2DUP‑R and check your results
 Pia Vattulainen

15:00-16:00                        Group Work 2, followed by review session:

 Hands on running PRE2DUP-R, identifying and solving issues and exploring special features of PRE2DUP-R

Session Responsibles:

Leader(s): Karina Standahl Olsen

Co-leader(s): Marko Lukic and Nhung Trinh

Speaker(s): Marko Lukic, Karina Standahl Olsen, Anick Berard, Ursula Berger

Short Description:
This seminar brings together university teachers to share practical, evidence-informed strategies for teaching epidemiology, pharmacoepidemiology, and biostatistics in a post‑2020 landscape. We’ll examine shifts in learner needs, hybrid delivery, assessment integrity, and the role of generative AI. Sessions spotlight a digital teaching platform for pharmacoepidemiology and approaches to teaching biostatistics that cultivate critical reasoning over rote computation. We conclude with a panel discussion to exchange resources, pitfalls, and next steps.

Learning Outcome(s):

By the end of the session, participants will gain increased awareness and competency in:

1. Post-2020 changes providing new frameworks that affect pedagogy, assessment, and student engagement in quantitative health sciences, including the opportunities and risks introduced by generative AI. A birds-eye view of the landscape starts off the session.
2. The design and use of digital platforms for teaching which have become increasingly popular, promising flexible learning and resource efficient teaching. The Canadian platform for transdisciplinary training in perinatal research on medications complementary to university trainings will be showcased, and experiences will be shared.
3. The increasing availability of statistical software, automated data analysis procedures, and artificial intelligence tools which has fundamentally changed the landscape of empirical research while easy access to powerful methods does not automatically translate into sound statistical reasoning. Teaching strategies will be highlighted, that center interpretation, uncertainty, and critical appraisal in biostatistics education, generalizable to other fields of teaching.

Tentative program

Time	Presenter	Title
13:00-13:45	TBA	Introduction: Teaching and learning post 2020 – changes and challenges
14:00-14:45	Prof. Anick Berard, Faculty of Pharmacy, University of Montreal	A digital teaching platform for pharmacoepidemiology: why, what and how
15:00-15:45	Prof. Ursula Berger, Institute for Medical Information Processing, Biometry and Epidemiology Ludwig-Maximilians-Universität, München	Empowering Students in Epidemiology: Teaching Biostatistics for Critical Reasoning
16:00-16:30	Moderator: TBA	Panel discussion with all presenters