Mathematical models are a central tool for forecasting the impact of infectious disease interventions. Many existing models, however, do not explicitly consider how an intervention such as a vaccination campaign may change social behaviour and contact rates within the population. This omission can alter conclusions about transmission dynamics. The authors present a transmission model that integrates changes in contact rates that occur in response to vaccine rollout, with applications to COVID-19.
The transmission framework introduced distinguishes two conceptual scenarios for how contacts respond to increasing vaccination coverage. In the first, a homogeneous mixing scenario, contact rates increase uniformly across the entire population as vaccination coverage rises. In the second, a heterogeneous mixing scenario, the probability that any two individuals make contact depends on their respective vaccination statuses. The heterogeneous formulation therefore allows for preferential mixing patterns between vaccinated and unvaccinated individuals.
Both scenarios are used to explore how vaccine-driven behaviour change interacts with vaccine characteristics (for example, transmission-blocking effectiveness) to shape epidemic trajectories.
Within this modelling framework the authors derive an expression for the effective reproduction number as a function of two principal inputs: vaccine coverage and the vaccine's transmission-blocking effectiveness. This derivation allows identification of parameter regimes in which increasing vaccination coverage does not necessarily reduce transmission; under some conditions an increase in coverage combined with higher contact rates can lead to a net rise in infections. The derived expression formalises how behavioural change tied to vaccination alters the relationship between coverage and transmission potential.
To examine empirical relevance, the model is parameterised with United Kingdom COVID-19 data spanning 2020 to 2022. The selection of this period permits assessment of how contact patterns changed in the context of the UK vaccine rollout and epidemic history. Specific numerical parameter values and fitting procedures are reported in the full text; the abstract documents the time window used for model parameterisation.
Vaccine-dependent contact rates are estimated using contact survey data. These estimates quantify the relationship between observed contact frequency and vaccination coverage during the study period. The authors report that estimated contact rates increased concordantly with vaccination coverage in the UK data used to parameterise the model.
Analysis of the UK contact survey data shows a temporal association in which contacts rose in tandem with increasing vaccine coverage. When this empirically observed relationship is incorporated into the transmission model, the results indicate that rising contact rates tended to increase transmission pressure; however, the overall epidemiological impact depended on the population mixing structure.
Implementing the empirically estimated vaccination-dependent contact increase within the model, the authors infer that the potential negative effect of rising contact rates on disease spread was tempered by a mixing pattern in which contacts predominantly involved at least one vaccinated person. In other words, preferential mixing that includes vaccinated individuals reduced the net transmission consequences of higher overall contact rates under the scenarios considered.
The abstract highlights conceptual and empirical findings without detailing all model assumptions, numerical parameter values, or the full fitting methodology; those details are available in the full article. The study underscores the importance of incorporating behaviourally driven contact changes into infectious disease models and demonstrates that vaccination campaigns can alter contact patterns in ways that materially affect transmission dynamics. Accounting for both the transmission-blocking properties of vaccines and the vaccination-dependent structure of contacts is necessary to interpret how coverage changes will influence epidemic outcomes.