Causal inference methods for vaccine sieve analysis with effect modification.

Area of Science:

• Epidemiology
• Biostatistics
• Immunology

Background:

• Vaccine protective effects can be modified by individual characteristics (e.g., age).
• Effect modification can also depend on pathogen characteristics measured post-vaccination.
• Traditional methods like subgroup analyses have limitations in capturing complex effect modifications.

Purpose of the Study:

• To develop a causal framework for evaluating effect modification within sieve analysis.
• To assess the magnitude of sieve effects and their modification by individual characteristics.
• To provide a robust method for real-world vaccine effectiveness studies.

Main Methods:

• Developed a causal framework integrating pathogen genetic data with individual-level data.
• Incorporated methods to handle competing risks, nonrandomized treatments, and differential dropout.
• Integrated modern machine learning techniques for enhanced analysis.

Main Results:

• Demonstrated the validity and efficiency of the proposed approach through simulation studies.
• Successfully applied the methodology to a malaria vaccine study.
• The framework can quantify sieve effects and their modification by individual factors.

Conclusions:

• The developed causal framework offers a powerful tool for understanding vaccine effectiveness.
• This approach enhances hypothesis generation regarding vaccine protection pathways.
• The method is applicable to complex epidemiological settings and diverse vaccine studies.