Maximizing the impact of limited vaccine supply under different epidemic conditions: a two-city monkeypox modelling analysis

O_SCPLOWBACKGROUNDC_SCPLOWIn the current global monkeypox outbreak, many jurisdictions have been faced with limited vaccine supply, motivating interest in efficient allocation. We sought to explore optimal vaccine allocation between two linked transmission networks over a short-term time horizon, across a range of epidemic conditions. O_SCPLOWMETHODSC_SCPLOWWe constructed a deterministic compartmental O_SCPLOWSVEIRC_SCPLOW model of monkeypox transmission. We parameterized the model to reflect two representative, weakly connected O_SCPLOWGBMSMC_SCPLOW sexual networks (cities) in Ontario. We simulated roll-out of 5000 vaccine doses over 15 days, starting 60 days after epidemic seeding with 10 imported cases. Within this model, we varied: the relative city (network) sizes, epidemic potentials (R0), between-city mixing, and distribution of imported/seed cases between cities. In each context (combination of varied factors), we then identified the "optimal" allocation of doses between cities -- resulting in the fewest cumulative infections by day 120. O_SCPLOWRESULTSC_SCPLOWUnder our modelling assumptions, we found that a fixed supply of vaccines could generally avert more infections over short-term time horizons when prioritized to: a larger transmission network, a network with more initial infections, and/or a network with greater R0. Greater between-city mixing decreased the influence of initial seed cases, and increased the influence of city R0 on optimal allocation. Under mixed conditions (e.g. fewer seed cases but greater R0), optimal allocation saw doses shared between cities, suggesting that proximity-based and risk-based vaccine prioritization can work in combination to minimize transmission. O_SCPLOWINTERPRETATIONC_SCPLOWPrioritization of limited vaccine supply based on network-level risk factors can help minimize transmission during an emerging epidemic. Such prioritization should be grounded in an understanding of context-specific drivers of risk, and should acknowledge the potential connectedness of multiple transmission networks.