Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
July 10, 2023Journal of the American Statistical AssociationOpen Access

Novel double negative control method corrects hidden biases in COVID-19 vaccine effectiveness estimates.

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Test-negative designs evaluate vaccine effectiveness but remain subject to residual confounding and collider stratification bias from healthcare seeking behavior, occupation, or prior infection.

Population

Individuals with data from the University of Michigan Health System

Comparison

COVID-19 vaccination vs unvaccinated

Design

Methodological paper with simulations and empirical application using a test-negative design

Key result

A novel double negative control inference approach was developed to account for hidden biases in test-negative design studies and applied to estimate COVID-19 vaccine effectiveness.

Authors

KLKendrick Qijun LiXSXu ShiMWMiao Wang

Discussion

Loading...

Member takes

Overview

May refine bias adjustment in test-negative vaccine studies; leaves open effects on effectiveness estimates.

Structured PICO

P
Population
Individuals experiencing symptoms and seeking care, tested for infectious diseases (application to COVID-19 using University of Michigan Health System data)
I
Intervention
COVID-19 vaccine (in the application)
C
Comparator
Unvaccinated (in the application)
O
Outcome
Vaccine effectiveness

A novel statistical method using double negative controls can help account for hidden biases like healthcare seeking behavior in test-negative design studies of vaccine effectiveness.

Cite This Study

Li et al. (2023) studied Infectious diseases (COVID-19). Double negative control inference method was evaluated on Vaccine effectiveness. A novel double negative control inference approach was developed to account for hidden biases in test-negative design studies and applied to estimate COVID-19 vaccine effectiveness.

synapsesocial.com/papers/6a169ea883b2be9fec6b4560https://doi.org/10.1080/01621459.2023.2220935
View Full Paper
Ask AI
Bookmark
Share