Novel Representations of Vaccine Protection Against Progression to Severe Disease Over Time
Year: 2026
Type: preprint
Abstract: Background Vaccines can prevent severe disease by preventing infection or by reducing progression among those who become infected. Vaccine effectiveness against progression given infection is often used to quantify this second mechanism, but it conditions on infection, which is itself affected by vaccination. As a result, this estimand lacks a clear causal interpretation and may behave non-intuitively over time. Methods We introduce a conceptual framework that models protection against infection and protection against progression as separate components that wane over time. Protection is represented using individual-level threshold-crossing times that depend on covariates and define a time-varying population susceptible to infection. Within this framework, we derive standard vaccine effectiveness estimands and propose two alternative decompositions of protection against severe disease: a progression-risk–weighted multiplicative decomposition and an additive decomposition based on absolute risk reduction. We illustrate their behavior using simulated examples. Results The weighted multiplicative decomposition restores a causal interpretation for progression protection within the doomed principal stratum and avoids negative estimates. The additive decomposition provides a clear representation of the pathways over time. Conclusions Explicitly modeling the infection-to-severe-disease pathway improves interpretation of vaccine effectiveness under waning immunity.
Source: medRxiv
Authors: Natalie E Dean, Veronika I. Zarnitsyna
Institution: Emory University
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7128903752