Integrating Genomic Data into Test-negative Designs for Estimating Lineage-specific COVID-19 Vaccine Effectiveness

Area of Science:

• Virology
• Epidemiology
• Immunology

Background:

• SARS-CoV-2 evolution necessitates ongoing evaluation of COVID-19 vaccine effectiveness (VE) against severe disease.
• Lineage assignment methods in test-negative design (TND) studies include sequence-based, proxy-based, and time period-based approaches.

Purpose of the Study:

• To summarize the benefits, challenges, and methodological considerations of different lineage assignment approaches for estimating SARS-CoV-2 lineage-specific VE.
• To model the impact of lineage misclassification error on VE estimates in TND studies using period-based versus sequence-based assignment.

Main Methods:

• Review of benefits, challenges, and methodological considerations for lineage assignment in VE TND studies.
• Deterministic modeling to assess the impact of lineage misclassification on VE estimates under various variant emergence scenarios.

Main Results:

• Period-based lineage assignment may underestimate VE differences between lineages due to misclassification, particularly during prolonged co-circulation or early variant takeover.
• Increasing predominance thresholds in period-based analyses can reduce bias but may decrease precision or preclude estimation.
• Sensitivity analyses are recommended for period-based VE studies to assess robustness to different predominance thresholds.

Conclusions:

• Test-negative design studies using various lineage assignment methods have advanced understanding of SARS-CoV-2 variant-mediated vaccine escape.
• Biases inherent in each study design vary, highlighting the need for careful analytic considerations for robust VE estimation.
• The identified principles for robust estimation may be applicable to other pathogens with continuous antigenic drift.