This year we are delighted to offer two pre-conference courses, set to run on Sunday 16 June 2024 13:00 - 17:00. They will be held at the same location used for the conference itself – at the Beurs Van Berlage.
Pre-conference Course 1
Bayesian methods for borrowing of information (with applications in clinical trials)
Course description: Using external-trial information such as historical control data in the design and analysis of clinical trials has gained considerable interest in recent years. It is especially desirable
in early phase drug development or situations of small population clinical trials, e.g., rare-disease clinical trials, where classical approaches can fall short. Use of data from historical studies conducted under the same or similar conditions could
mean less information is required in a new study while preserving study power. This course is designed to introduce both fundamental concepts and advanced features of Bayesian methods that permit borrowing of information. Case studies will be presented
to illustrate key points to consider in the planning stage, as well as effective communication of advanced designs to non-statistician professionals.
Learning objectives: This course covers the principles of statistical designs that incorporate external trial information like historical control data in clinical trials. Concepts and several Bayesian methods
are illustrated using case studies. Participants will:
- be engaged throughout with regular poll questions
- understand both the benefits and caveats of methods for borrowing of information
- learn by example how these approaches are easily implemented in R using the RBesT package
- see examples of how to communicate advanced designs with non-statistician collaborators
- hear perspectives on future methodological developments in the field
- participate in an open discussion about the benefits of, and potential challenges around the implementation of techniques for borrowing of information.
Intended audience: This course is aimed at statisticians working in industry, academia, government, and graduate students who are interested in the design and analysis of clinical trials using Bayesian methods
for borrowing of information from external sources. Individuals with good knowledge of trial designs and/or some knowledge of Bayesian statistics would benefit the most from the course.
Course instructors: Dr Haiyan Zheng is a Senior Lecturer in Statistics at the Department of Mathematical Sciences, University of Bath. Haiyan’s research is primarily focused on Bayesian
adaptive methods that improve the efficiency of small population clinical trials, which include precision oncology trials, rare disease trials, etc. Dr Lisa Hampson is a Senior Director in Novartis’s Statistical Methodology
group in Basel. Her research interests are in novel trial designs, real-world evidence and use of Bayesian methods to leverage trial-external information. Dr Sebastian Weber is a Director in Novartis’s
Advanced Exploratory Analytics group in Basel. He has worked on Bayesian evidence synthesis, early phase I dose-escalation trials in Oncology, paediatric extrapolation and is interested in the application of pharmacometrics to sparse clinical data.
Pre-conference Course 2
Applying causal inference for effective implementation of the estimand framework in drug development
Course description: There is increasing interest in estimands and causal inference in RCTs. This is reflected in the ICH E9 (R1) addendum on estimands (2019) which clarifies that effects other than the
traditional intention-to-treat may be of interest; considers post-randomization confounding and selection bias due to intercurrent events; and discusses principal stratum and hypothetical strategies. Estimation aligned with these strategies often
requires causal inference methodology. It is also reflected in the FDA guidance on covariate adjustment (2023) where the difference between conditional vs. marginal effects is discussed and G-computation (standardization) is mentioned as one possible
estimation approach. Application of causal inference can help provide impactful insights into trial data. During the course the following topics will be covered:
- Introduction to Causal Inference
- Overview of causal inference and its alignment with the estimand framework and key points in ICH E9(R1)
- Introduction to causal inference, including potential outcomes and directed acyclic graphs (DAGs), causal effects and common assumptions
- Conditional and Marginal Treatment Effects
- Introduction to conditional and marginal treatment effects
- Appropriate estimators for conditional and marginal estimands
- Estimation Methods of Causal Effects targeting Hypothetical Estimands
- Introduction to common estimation methods, e.g., g-computation, IPW (Inverse probability weighting)
- Introduction to targeted maximum likelihood estimation (TMLE)
- Discussion on how to adjust for time-varying confounding
- Other topics
- In this section we will give an outlook to various application of causal inference in clinical trials including principal stratum estimands and their estimation, generalization of trial results to
a population wider than the trial population
Throughout, case studies and RCT examples will be presented and illustrated using R code.
Learning objectives: At the end of this course, participants will be able to:
- describe basic concepts of causal inference (e.g., potential outcomes, common assumptions, confounders) and appreciate the role of causal inference in RCTs (e.g., in dealing with intercurrent events)
- identify and apply common estimation methods of causal effects
- differentiate conditional and marginal treatment effects, to learn the appropriate corresponding estimation methods, and to be able to implement the learnings in clinical trial designs and analyses.
- understand common causal inference methods (e.g., g-computation, IPW) and their application
Intended audience: This course is aimed at statisticians working in industry, academia, government, and graduate students who are interested in applying causal inference methods in drug development in
general and in clinical trials specifically. Individuals should have a good understanding of the estimand framework as outlined in ICH E9 (R1).
Course instructors: Dr Tobias Mütze (Novartis Pharma AG) is an Associate Director in Novartis’ Statistical Methodology group in Basel. His research interests are in estimands, time to event and recurrent event methodology, and the application of causal inference to clinical trials. Dr Simon Newsome (Novartis Pharma AG) is an Associate Director in Novartis’ Statistical Methodology group in Basel. Simon’s prior research includes the use of causal inference methods to analyse observational data in the presence of time-varying confounding, and his current research focuses on the application of causal inference to clinical trials. Prof Karla Diaz Ordaz (University College London) is Principal Research Fellow in Biostatistics. Karla’s research explores the use of statistical machine learning in causal inference to assist in medical and policy decisions, using both observational and clinical trials data. This is often referred to as causal machine learning (double machine learning and targeted learning are examples of this methodology).
The number of places on these courses is limited, register today!
The registration fees for the pre-conference courses are as follows.
|Early Bird Rate
All amounts are in Euros and include VAT.
The course will be available to register to attend alongside your event registration, in January.