- The courses will be given on the 15th September, 2025.
- They will be open when ten or more participants are enrolled.
- Each course is limited to a maximum of thirty participants.
- The available courses are listed below.
- The course fee and the deadline for Early Bird Registration are available on the Conference Fees page.
- The deadline for the registration is the 30th June, 2025.
- You must fill out the form to register for the courses.
Instructors: Per Kragh Andersen - University of Copenhagen (Denmark) and Henrik Ravn - Novo Nordisk A/S (Denmark)
This course will review analysis of multi-state survival data based on the book 'Models for Multi-State Survival Data: Rates, Risks, and Pseudo-Values' (Chapman and Hall/CRC, 2023). This includes models for intensities, the basic parameters in a multi-state model, and models for marginal parameters (e.g., state occupation probabilities and expected lengths of stay in states). For the latter, both plug-in of intensities and direct marginal models will be studied, a general approach being based on pseudo-values. Methods will be illustrated via real world examples and exercises using R.
Instructors: Haiyan Zheng - University of Bath (United Kingdom) and Giulia Risca - University of Milan Bicocca (Italy)
This course covers the principles of statistical designs, monitoring, and analysis of modern clinical trials. A strong emphasis will be placed on Bayesian methods that permit dynamic borrowing of information for enhanced decision making. To start with, we will provide an overview of the evolving landscape of clinical trials methodology. Key topics include adaptive designs for mid-course modifications made to an ongoing study, and master protocols for simultaneous evaluation of multiple treatments, patient subgroups, or both. Participants will gain a comprehensive understanding of Bayesian inferences, particularly when a mixture prior distribution is used, in enhancing the decision-making throughout the planning, conduct, and analysis of clinical trials. Advantages over traditional frequentist approaches will be demonstrated. Real-world case studies that use Bayesian methods to guide dose (de-)escalation, early stopping, and the incorporation of external-trial information into small population clinical trials will be presented. Participants will further benefit from the hands-on practical sessions to learn implementation of the state-of-the-art methods using R and JAGS.
Intended audience: This full-day course is aimed at applied statisticians working in industry, academia, government, and graduate students who are interested in Bayesian methods and adaptive designs for modern clinical trials. Individuals who understand basics of clinical trials (e.g., purpose, design, ethical considerations) and seeks to expand their expertise in Bayesian methods would benefit the most from this course. Participants with good knowledge of Bayesian methods can delve into the inferences and explore various applications in medical research.
Instructor: Altuna Akalin - Max Delbrück Center (Germany)
The participants will learn how to use AI-assisted data analysis platform, and learn about correct prompting strategies to produce the desired results and troubleshoot potential problems.
Instructor: Therese Andersson - Karolinska Institutet (Sweden)
This course will cover the modelling of survival (time-to- event) data using flexible parametric survival models, with practical examples in Stata. Within flexible parametric survival models, the baseline hazard (or a function of the baseline hazard) is modelled using cubic splines. The advantages of this approach over the commonly used Cox model are the ease with which smooth predictions can be made, the modeling of complex time-dependent effects and investigation of absolute as well as relative effects. The course will make use of the stpm3 command (released on SSC in June 2023). The course will cover general modelling issues that are useful when using survival models for either description, prediction or understanding causality. Specific topics that will be covered are e.g., the advantages and disadvantages of using flexible survival parametric models, the choice of number and location of knots to model the effect of time, relaxing the proportional hazard assumption and predictions of the survival, hazard and other useful functions.