An introduction to causal inference for pharmacometricians

Area of Science:

• Pharmacometrics
• Biostatistics
• Epidemiology
• Artificial Intelligence
• Machine Learning

Background:

• Formal causal inference is increasingly relevant in pharmacometrics and related fields.
• Pharmacometricians require foundational knowledge of causal inference for advanced applications.
• Interdisciplinary collaboration necessitates shared understanding of causal reasoning.

Purpose of the Study:

• To introduce pharmacometricians to fundamental causal inference concepts.
• To provide a tutorial on potential outcomes, g-formula, and directed acyclic graphs (DAGs).
• To enhance the application of causal inference methods within pharmacometrics.

Main Methods:

• Explanation of the potential outcomes framework.
• Introduction to the g-formula for causal effect estimation.
• Overview of directed acyclic graphs (DAGs) for modeling causal relationships.

Main Results:

• The tutorial clarifies three core causal inference concepts.
• It provides a foundation for pharmacometricians to engage with causal inference literature and methods.
• Understanding these concepts facilitates more rigorous study design and analysis.

Conclusions:

• Pharmacometricians will benefit from mastering potential outcomes, g-formula, and DAGs.
• This foundational knowledge supports the integration of causal inference into pharmacometric practice.
• The tutorial serves as a starting point for deeper exploration of causal inference in drug development and research.