Moving Beyond the 'Pro' or 'Anti' Binary: Insights into Vaccination Attitudes and Behaviour, with a Case Study from a London, UK Survey

Background: Responses to vaccination programmes vary widely, with contrasting perceptions of benefits and harms. This study aimed to provide a contemporary characterization of distinct groups of respondents by moving past traditional "pro" versus "anti" vaccination binaries. The results captured the complex spectrum of vaccine sentiment to directly inform the next strategic phases of a multi-year public health campaign regarding vaccine perceptions and behaviours within the diverse population of London. Methods: Data came from the 2024 Why We Get Vaccinated survey conducted in London, UK using a community-based participatory approach. Participants answered vaccine-related questions based on previous vaccine uptake, future willingness to vaccinate (WTV), perceived effectiveness, and safety concerns. We employed Bayesian Binomial logistic regression to identify sociodemographic predictors of individual responses. Subsequently, a Bayesian Nonparametric Latent Class Analysis (BNP-LCA) using a Dirichlet Process Mixture model was used to identify distinct groups and their sociodemographic compositions. Results: The LCA identified five distinct classes. Class 1, the Consistent Uptakers (37.2%), exhibited high vaccine uptake (89.2%) and near-universal belief in efficacy (98.0%) with minimal safety concerns. Class 2, the Concerned Uptakers (17.5%), maintained high WTV (72.2%) despite significant concerns regarding adverse effects (71.2%). Class 3, the Consistent Refusers (18.4%), demonstrated uniform rejection of vaccines and high levels of safety concern (74.4%). Class 4, the Unconcerned Refusers (20.8%), acknowledged vaccine effectiveness (98.4%) but showed very low uptake (1.7%), possibly due to low perceived personal risk rather than active resistance. Class 5, the Undecided/Uninformed (6.0%), was characterized by pervasive uncertainty and "Don't Know" responses regarding efficacy (80.3%) and vaccination intent. Sociodemographic analysis revealed that residential instability, i.e renting, and caregiving responsibilities were significant barriers to uptake, while older age and retirement were the strongest positive predictors. Conclusions: Vaccine behaviour and attitudes in London are a multidimensional construct. The discovery of "Unconcerned Refusers" and "Concerned Uptakers" highlights that "one-size-fits-all" public health campaigns may not be effective. Tailored public health messaging is required to address the specific uncertainties and socio-economic barriers identified across these distinct profiles. These findings provide a functional framework that will directly guide upcoming resource allocation, messaging adaptations, and policy decisions for the next iterations of the ongoing campaign