In this study, we simulated mpox outbreaks under various hypothetical scenarios to predict the potential spread of Clade Ib MPXV in Singapore, and assessed the impacts of vaccination and NPIs inclusive of case isolation, quarantining of close contacts and contact tracing. Our simulations revealed a high possibility of large-scale transmission with steep epidemic curves, especially in scenarios where the virus was highly transmissible through sexual routes. In the scenario with moderate disease transmissibility through non-sexual means and when the general population, rather than specific subpopulations, were affected, outbreaks would result in 30% and 25% of the total population being infected with and without control through NPIs, respectively.

Nevertheless, at least a one-month delay was observed from the initial importation event before exponential growth in infections across various scenarios assessed. This could provide decisionmakers with sufficient time to set up emergency infrastructure, testing centres and create treatment guidelines, and to also then begin containing any potential spread before the number of infections become a burden on the healthcare system. In addition, mandatory home isolation of all diagnosed cases, covering 60% of local infections, was projected to reduce the peak number of new infections by 35% and the cumulative infection size by 11%. While incomplete adherence could lead to secondary infections of close contacts, particularly if non-sexual transmission be an established pathway of transmission [6], isolation at home remains more feasible than in healthcare settings. Given the much longer infectious period of mpox compared to COVID-19 [18]. extended isolation could easily strain pre-existing infrastructure with the surge in case counts.

Although pre-departure testing was shown to be effective in preventing importation during the COVID-19 pandemic [19]. its limited effectiveness in our simulations might be explained by the high transmissibility of the MPXV through sexual contact, which resulted in rapid local disease spread, thereby overshadowing the role of importation in driving the outbreak. In contrast, the extra benefits of quarantining close contacts beyond isolation and screening was projected to be much greater in delaying the outbreak, reflecting its extensive use during the COVID-19 pandemic [20,21]. Nevertheless, it is important to note that contact tracing would require substantial human resources for identifying the close contacts, which may not always be available. Isolation is therefore recommended as the primary intervention strategy with the addition of screening or contact tracing as additional measures. In the ongoing mpox outbreak of 2024, most identified patients infected with Clade Ib MPXV outside Africa were imported cases with travel history from high-risk countries [22], making it plausible that any mpox case outbreaks in APAC would be driven by importation with minimal local spread. However, as demonstrated by our simulations, it is also possible that an outbreak could occur depending on mpox transmissibility (Figure 3). We therefore examined the proposed NPIs’ effects under these two scenarios in a subsequent sensitivity analysis, in which the virus’s transmissibility for non-sexual transmission were set to be nearly zero or significantly higher than the baseline scenario. The marginal differences observed in the proportion of infections and hospital admissions averted by these measures consistently demonstrated their effectiveness in containing the outbreak (Figure S7–8, Table S10–11).

While our model outputs supports the pivotal role vaccination played in controlling the epidemic, the results also emphasised the importance of its strategic implementation [23]. Prioritising the most vulnerable group, the sexually active individuals in this context, would maximise its reduction in the overall disease burden. Nonetheless, vaccination might not be optimal in our setting of an outbreak in an APAC city, as its projected benefits were no higher than those brought by various NPIs. Although these limited discrepancies in effects between the two types of interventions might be explained by the relatively low vaccination scale (5%) in our proposed strategies, NPIs could be more sustainable considering the high prices of approved mpox vaccines and the time required for them to take effect in individuals [24]. Moreover, mass vaccination campaigns often face barriers such as the limited vaccine supply and shortage of healthcare workers [25]. Another challenge may be potential vaccine hesitancy in the APAC region, driven by historic public health events, safety concerns, and spread of misinformation [26–28]. Additionally, as the 2022 mpox outbreak primarily affected the GBMSM community, the general public might perceive a low risk of infection, resulting in diminished recognition of the vaccine’s importance and hence reduced motivation in immunisation [29]. These obstacles would collectively make it unlikely to increase vaccine uptake rapidly enough to prevent a forthcoming outbreak. Vaccination for the general APAC population at this time may also result in a lack of vaccines being available for the high-risk groups, especially among countries with ongoing outbreaks, facilitating the continued mitigation of global transmission risks.

One major limitation of our study pertained to the paucity of informative statistics. Our simulated trajectories suggested that the current outbreak potentially might be in an early stage when compared to the surveillance data available at the time of this study, as importation from African countries accounted for most of the globally detected cases. This posed great challenges to extrapolating the disease transmissibility, the key parameters in our model, in an APAC setting where the contact patterns might differ significantly from Africa. To address the large uncertainties in model parameters, we performed a sensitivity analysis in which we simulated a range of outbreaks using various combinations of the basic reproduction numbers for transmission through sexual (R_s) and non-sexual (R_c) activities. This aimed to capture both the potential extremes caused by super-spreaders and the most probable epidemic trajectories (Figure S1). Furthermore, the predominance of the available surveillance data sourcing from Africa, where healthcare and monitoring systems might face constraints, could lead to an underestimation of infection sizes and an overestimation of disease severity. Consequently, the simulated HCRU projected from the African data in our study might far exceed reality in APAC, offering a potential silver lining. With the availability of more case data and epidemiological parameters, these models require updating.

Another limitation was the simplification of real-world scenarios in our model, such as hospital discharge and readmission of patients who had not fully recovered, as well as variations in population size due to natural birth and deaths. However, this was believed to have a minimal impact owing to the large population affected by the outbreak. We also streamlined the clinical testing and quarantine processes, assuming accurate testing results and no tests or quarantine for uninfected individuals. In addition, we did not explicitly account for international arrivals who were exposed to the disease but not yet infectious, as the lack of relevant travel data made it challenging to quantify the exact population. Nevertheless, we have built in such possibilities by adjusting sensitivity of the pre-departure screening to the relatively low 60%.

Despite these limitations, the transmission model proposed in this study is sufficiently flexible to accommodate for a range of scenarios with diverse disease transmission dynamics, enabling decisionmakers to plan for different potential outbreak possibilities. This study explores the impact of NPI impacts on mpox outbreaks, an area much less explored compared to the virus’s transmission patterns [15,30]. This has provided evidence for adapting the containment measures widely applied during the COVID-19 pandemic to effectively managing new outbreaks with probably high community transmission potential.

Our simulations suggest that while mpox outbreaks triggered by importation could lead to substantial morbidity and mortality in an APAC city with low immunity to the virus, they would still remain controllable provided adequate preparation and a timely response from decisionmakers, underscoring the significance of robust surveillance systems. In addition, the NPIs explored are able to delay and suppress outbreak sizes and curb disease spread up to 36%. Based on this, mass vaccination of the general population is not recommended but surveillance systems, infrastructure for contact tracing, quarantining of close contacts and institutional isolation of cases is urgently required.