Target Trial Emulation Without Matching: A More Efficient Approach for Evaluating Vaccine Effectiveness Using...

Area of Science:

• Epidemiology
• Biostatistics
• Vaccinology

Background:

• Real-world vaccine effectiveness is often assessed using matching in observational studies, following target trial emulation.
• Matching methods have limitations in estimand clarity and precision, especially when vaccine uptake and infection dynamics change rapidly over calendar time.

Purpose of the Study:

• To propose a novel causal estimand for vaccine effectiveness that summarizes effectiveness over calendar time, analogous to randomized controlled trial efficacy.
• To develop and evaluate simple-to-implement estimators for this new estimand.

Main Methods:

• Proposed a new causal estimand for vaccine effectiveness over calendar time.
• Developed estimators based on two hazard regression models.
• Validated the proposed estimator through simulations and a real-world study of the Pfizer-BioNTech COVID-19 vaccine in children aged 5-11 years.

Main Results:

• The proposed estimator demonstrated similar scientific inferences compared to matching-based estimators.
• Significant efficiency gains were observed with the proposed estimator over commonly used matching methods.
• The estimator was successfully applied to assess Pfizer-BioNTech COVID-19 vaccine effectiveness against SARS-CoV-2 infection in children.

Conclusions:

• The novel causal estimand and associated estimators provide a more precise and clear measure of vaccine effectiveness in real-world settings.
• This approach offers substantial efficiency improvements for observational vaccine effectiveness studies.
• The method is applicable to dynamic epidemiological scenarios and specific vaccine evaluations, such as for pediatric COVID-19 vaccination.