Prioritising older individuals for COVID-19 booster vaccination leads to optimal public health outcomes in a range of socio-economic settings

The rapid development of vaccines against SARS-CoV-2 altered the course of the COVID-19 pandemic. In most countries, vaccinations were initially targeted at high-risk populations, including older individuals and healthcare workers. Now, despite substantial infection- and vaccine-induced immunity in host populations worldwide, waning immunity and the emergence of novel variants continue to cause significant waves of infection and disease. Policy makers must determine how to deploy booster vaccinations, particularly when constraints in vaccine supply, delivery and cost mean that booster vaccines cannot be administered to everyone. A key question is therefore whether older individuals should again be prioritised for vaccination, or whether alternative strategies (e.g. offering booster vaccines to the individuals who have most contacts with others and therefore drive infection) can instead offer indirect protection to older individuals. Here, we use mathematical modelling to address this question, considering SARS-CoV-2 transmission in a range of countries with different socio-economic backgrounds. We show that the population structures of different countries can have a pronounced effect on the impact of booster vaccination, even when identical booster vaccination targeting strategies are adopted. However, under the assumed transmission model, prioritising older individuals for booster vaccination consistently leads to the most favourable public health outcomes in every setting considered. This remains true for a range of assumptions about booster vaccine supply and timing, and for different assumed policy objectives of booster vaccination. Computational modelling is used increasingly to assess interventions against infectious diseases. For COVID-19, booster vaccines are being deployed worldwide to mitigate the effect of waning immunity following past infections and vaccinations, and to protect against new SARS-CoV-2 variants. However, booster vaccine availability is limited, making it necessary to consider who should receive booster vaccinations. It is essential that booster vaccine prioritisation is considered not only in the context of high income countries, but also low and middle income countries where the population age structure is very different. We have therefore developed a computational model to guide booster vaccination prioritisation in different socio-economic settings. We show that vaccinated populations can have different compositions between countries, even when identical booster vaccination strategies are adopted. Despite these differences, our modelling results suggest that booster vaccination strategies prioritising the oldest individuals lead to the best possible public health outcomes, due to the substantial skew in the risk of severe outcomes of infection towards older (and other vulnerable) hosts. This remains true even when quantifying the extent of premature, rather than absolute, mortality. This research therefore supports policy advice recommending prioritisation of older individuals for booster vaccines in countries worldwide when vaccine supply is limited.