Influence of correlated vaccination behaviors on estimates of COVID-19 vaccine effectiveness in older adults - VISION network, October 2023 - March 2024.

Test negative design studies allow for COVID-19 vaccine effectiveness (VE) estimation while minimizing selection bias from healthcare seeking and testing practices. However, failure to consider correlation between vaccination behaviors may result in biased COVID-19 VE estimates. Using different methods to account for potential correlation between COVID-19, influenza, and respiratory syncytial virus (RSV) vaccination status, we investigated variability in VE estimates of the 2023-2024 COVID-19 vaccine against COVID-19-associated emergency department and urgent care (ED/UC) encounters and hospitalizations during the 2023-2024 respiratory virus season. Data were leveraged from VISION, an electronic health record-based platform. VE estimates ≥7 days post vaccination did not increase by more than 5 percentage points against ED/UC encounters and 3 percentage points against hospitalizations when accounting for influenza and RSV vaccination status or excluding influenza and RSV positive controls. As the magnitude of this influence depends on season-specific factors, ongoing monitoring is warranted.