Modelling targeted rabies vaccination strategies for a domestic dog population with heterogeneous roaming patterns

Australia is currently canine rabies free. However, communities located on the northern coastline–such as the Northern Peninsula Area (NPA), Queensland–are at risk of an incursion due to their large populations of susceptible free-roaming dogs and proximity to rabies-infected Indonesian islands. A rabies-spread model was used to simulate potential outbreaks and evaluate various disease control strategies. A heterogeneous contact structure previously described in the population of interest–explorer dogs, roamer dogs and stay-at-home dogs–was incorporated into the model using six spatial kernels describing contacts between dog roaming categories. Twenty-seven vaccination strategies were investigated based on a complete block design of 50%, 70% and 90% coverage for each of the three roaming categories to simulate various targeted vaccination strategies. The 27 strategies were implemented in four population structures in which the proportion of dogs in each category varied–explorer dominant, roamer dominant, stay-at-home dominant and a field population (based on field estimates of population structure). The overall vaccination coverage varied depending on the subpopulation targeted for vaccination and the population structure modelled. A total of 108 scenarios were simulated 2000 times and the model outputs (outbreak size and duration) were compared to Strategy 14 (a standard recommended overall 70% vaccination coverage). In general, targeting explorer dogs–and to a lesser extent roamer dogs–produced similar outbreaks to Strategy 14 but with a lower overall vaccination coverage. Similarly, strategies that targeted stay-at-home dogs required a higher vaccination coverage to produce significantly smaller and shorter outbreaks. This study provides some theoretical evidence that targeting subpopulations of dogs for vaccination based on their roaming behaviours (and therefore risk of rabies transmission) could be more efficient than blanket 70% vaccination campaigns. Such information can be used in preparedness planning to help improve control of a potential rabies incursion in Australia.Author summary: Australia is currently free of canine rabies. However, northern communities are at risk of an incursion. The free-roaming dogs in the at-risk communities of the Northern Peninsula Area, Queensland, Australia, display various roaming behaviours which affect how they interact with each other and subsequently affects how rabies could be spread within the population. The presence of these different roaming behaviours provides an opportunity to assess targeted vaccination strategies. Using a rabies spread model, this study provides evidence that targeting roaming dogs (that cause rapid rabies spread) for vaccination instead of dogs that remain around their owner’s residence, can be more efficient than the recommended overall 70% vaccination coverage to control rabies. Furthermore, if dogs that stay at home most of the time are targeted versus other roaming dogs, a higher overall coverage and more resources would be required to control a potential rabies outbreak than the overall 70% vaccination coverage. Comparing various vaccination strategies to find efficient and effective response options is beneficial to contribute to emergency planning and increase preparedness to control a rabies outbreak, should it occur.