Event

Apply causal inference and estimands to improve real-world evidence and trial analyses

This introductory course is designed for researchers, analysts, and decision makers involved in health technology assessment (HTA), regulatory science, and evidence-based healthcare. The course explores how causal inference methods and appropriate estimands can improve the design and interpretation of both real-world studies and clinical trials—especially when addressing treatment switching or other biases.

Technical topics include:

• Introduction to causal principles, including causal diagrams (directed acyclic graphs; DAGs) and target trial emulation.

• Common sources of bias in observational studies and trials (e.g., time-zero bias, immortal time bias) and strategies to avoid them.

• Causal methods for baseline confounding (e.g., multivariable regression, propensity scores).

• Methods for time-varying confounding (e.g., g-formula, marginal structural models, inverse probability weighting, and g-estimation).

• Selection and definition of appropriate estimands to directly address decision problems, including in trials with treatment switching.

This course includes tools and concepts that can be immediately applied, including:

• Real-world case examples from HTA, such as external control arms and treatment-switching scenarios.

• Practical guidance on selecting and implementing causal inference methods and estimands.

• Discussion of current perspectives from HTA agencies, including acceptance and barriers to adoption.

This short course is offered in-person at the ISPOR Europe 2024 conference.

Nottingham, UK

Schedule:

LENGTH: 4 Hours | Course runs 1 day

Sunday, 9 November 2025 | Course runs 1 Day
13:00-17:00 Central European Time (CET)