The spread of canine rabies in eastern Indonesia and risks to Australia

Regional context

Despite a suspected incursion in Tasmania in 1866–67, Australia has remained historically canine rabies free. This freedom is due to strong quarantine regulations on the importation of susceptible species. More recently, there has been increasing recognition of disease threats to northern Australia¹ – a large, susceptible wild and domestic dog population, proximity to a rabies endemic region and illegal activities by which an infected, incubating dog might be introduced – and rabies was included in Australia’s animal disease preparedness plans (AUSVETPLAN) in 1991 to address a potential incursion.²

In Indonesia, suspicion of rabies began in the 1880s with the first confirmed case reported in 1889 from the Batavia (Jakarta) district.³ By the early 1900s rabies was commonly reported from Java, Sumatra, Sulawesi and Borneo (the Greater Sunda Islands). Rabies spread to Flores Island in 1997 (and from there to Solor in 1998, Adonara in 1998 and Lembata in 2001), Ambon Island in 2003 (and from there to Seram in 2003, Buru in 2004, Larat in 2010 and Kisar in 2012),⁴ Halmahera Island in 2005, Bali in 2008 and Sumbawa in 2019. In 2023 canine rabies was reported from West Timor (spreading to East Timor in 2024) (Fig. 1). Within Indonesia, a variety of drivers of the movement of dogs and hence the spread of rabies have been described, including demand for hunting dogs; demand for dogs to guard crops; ethnic, cultural and religious trends; gifting of puppies; dog meat consumption; regional road improvements; and in some areas, the expansion of oil palm plantations.⁵ In 2019, a World Organization for Animal Health (WOAH)-led region-wide risk assessment was undertaken. Making use of techniques from participatory epidemiology and expert opinion from throughout South-east Asia, rabies spread pathways were investigated by the movement of dogs within the region, both between and within countries of South-east Asia (Fig. 2). This activity provided insights such as commercial transportation being the most commonly reported mode, and pet ownership and companionship being the most common reason for dog movements. Better dog demographic data were identified as a strong need within this region to inform risk assessment of rabies spread.

Fig. 1.

The spread of canine rabies in the eastern islands of Indonesia and Timor Leste, 1997 (yellow) to 2024 (red).

Fig. 2.

Examples of methodologies to assess the risk of rabies spread within South-east Asia: the ‘jellybean’ map (left) and scenario tree model (right).

Rabies in Bali

In Bali, rabies was first confirmed in November 2008 in a village on the Bukit Peninsula, immediately to the south of the international airport.⁶ Although never confirmed, it was suspected that the source of the incursion was the importation of an unvaccinated dog from Flores, sometime in May 2008.⁷ A competing hypothesis is that rabies was introduced by fishermen from Sulawesi.⁶ Regardless, human deaths occurred in September (1) and November (3), prompting recognition of a rabies incursion likely 6 months after it had first occurred.⁷ It is notable that dog deaths did not lead to recognition of this incursion, but rather human deaths. Although island-wide vaccination campaigns were initiated, delays in these campaigns, lack of coordination, use of suboptimal canine vaccines, free roaming dog populations with high turnover rates and an ongoing focus on dog culling have prevented the goals of canine and human elimination.^6–8 The goal of elimination from an island is feasible, but this will require time, commitment and implementation of a One Health approach.⁹

Rabies in West Timor

The most recent rabies spread event in Indonesia is the incursion into West Timor (Nusa Tenggara Timur), in 2023. This incursion was reported to WOAH in June 2023, and likely began in April or earlier.¹⁰ The incursion site was South Central Timor district (Timor Tengah Selatan). During a WOAH mission to West Timor, we conducted key informant interviews to better understand the incursion pathway. The first reported case was a 45-year-old in the village of Fenun (Fig. 3), a small (1824 residents) isolated village ~56 km due east of So’e, the district capital.¹¹ A possible disease spread pathway is formed by the strong ethno-social connections between this district and the island of Flores. Flores and West Timor are part of Nusa Tengara Timur province, which might enhance the risk of rabies spread. Anecdotally, movement of dogs as pets is suspected. Ferry connections exist between Flores to Kupang, and it is possible that dogs could be illegally transported via ferries. This potential spread pathway has been documented previously in Lombok, Nusa Tengara Barat.¹² Community attitudes towards dogs are strongly influenced by ethnic and religious affiliations. In Christian areas there is higher dog ownership and more tolerance of free-roaming dogs. Community culture also places value on dogs, including pet and companionship, guard duty, trading, breeding and the exchange of gifts.¹³ Another explanation offered for the potential spread of rabies from Flores to West Timor in 2023 is that in Flores during the COVID pandemic there was a reduction in rabies vaccination coverage: post-COVID people began travelling again and combined with the increased rabies prevalence (reduced herd immunity) there was a substantial increased risk of rabies spread. Transportation of dogs from Bali was also acknowledged as a possible rabies spread pathway, but the smaller volume of trade (frequency and volume) means this is less likely. Other considerations include the dog meat trade in West Timor (centred mostly on Kupang involving local dogs and informal trade) and fishermen from Sulawesi. Potential disease spread pathways can be further informed by phylogenetic information.¹⁴

Fig. 3.

Activities to inform rabies incursion response and risk assessment in West Timor, Indonesia, December 2023. The index village of Fenun and area surveyed to estimate dog population size (left) and sight–resight dog counting in a peri-urban area of So’e are shown.

During this WOAH mission we estimated the dog population size (192,461; 95% CI 183,746–201,176) in West Timor by sight–resight surveys (Fig. 3). This survey method is an efficient and accurate approach to estimate the size of roaming-dog populations. It is based on conducting two identical surveys, 2 days apart, counting dogs seen outside private property on both days and noting the dogs that are re-sighted on day 2. Sight–resight surveys have been previously used to support rabies preparedness programs in both Timor Leste¹⁵ and northern Australia.¹⁶ Based on the West Timor surveys, the number of dog vaccine doses needed to eliminate this rabies incursion was estimated at 135,000 (95% CI 128,000–140,000). Despite initiation of a vaccination campaign, canine rabies spread to East Timor, where it was first reported in the enclave of Oecusse in February (dogs) and March (human death), 2024.

Risk assessment: the rabies risk to Australia

In 2012, the risk of a rabies incursion in northern Australia was assessed qualitatively using expert opinion.¹⁷ The probability of entry included proximity to Indonesia and fishing, trade and cultural travel routes and considered the potential impacts on communities along the coastline in which domestic dogs roam free. Similarly, in a review of rabies incursion risk, Sparkes et al.¹⁸ considered several entry pathways including the illegal importation of rabid dogs travelling on boats engaged in Australian–South-east Asian cross-cultural traditions, unauthorised fishing vessels and itinerant yachts. This focused on the previously identified high priority regions of East Arnhem Land (Northern Territory) and the Northern Peninsula Area, Cape York Peninsula (Queensland).¹⁷ As noted, the ongoing eastwards spread of canine rabies in Indonesia increases the risk of an incursion in northern Australia,¹⁸ since the prevalence of rabies in the source population is a factor that influences incursion risk.^19,20 Therefore any program to reduce risk needs to include control at the source, i.e. the reduction of rabies spread and incidence in these newly infected areas of eastern Indonesia by population management and vaccination programs.⁵ In addition, the increased risk of rabies spread to New Guinea island – the Indonesian provinces of Papua and the nation of Papua New Guinea – is of great cause for concern.^21,22 Based on the Bali experience, rabies control in this region would be extraordinarily challenging. Ongoing risk of rabies spread to northern Australia via the Torres Strait would be likely, perhaps requiring a modification of policy and pre-emptive vaccination of the local dog population.

To address preliminary estimates of risk, a rabies research program was instigated to investigate the risk of an incursion in northern Australia. Both risk assessment and disease modelling were used.

The risk of a rabies incursion to Cape York Peninsula, Queensland was investigated using quantitative risk analysis.²⁰ To confirm suspected high-risk pathways, structured interviews were conducted with key informants (biosecurity officers and land and sea rangers) from the local Indigenous communities. This confirmed the risk associated with illegal boat entries from Indonesia. Key informants believed that the reasons for travel was fishing, specifically for shark-fin. In addition, they believed the main reasons for a dog on board was for hunting and companionship. The probability of entry of rabies into north-west Cape York Peninsula via an infected dog on this pathway, followed by the probability of exposure of a dog resident in a remote community, was then estimated. The probability of exposure of an Australian resident dog to rabies was influenced by the prevalence of rabies in dogs at the origin of the route in Indonesia.

Several disease spread models have been developed to explore the impact of a rabies incursion in this at-risk location, in both domestic and wild dogs.^23–25 The northern Australian coastline is home to Indigenous communities in which domestic dogs roam free much of their time, and where there is a wild–domestic dog interface with dingoes and feral dogs.^26,27 Whether an incursion occurs in the domestic or the wild dog population, without interventions rabies would likely spread. In the case of the wild dog population, an incursion would spread (defined as a case >120 km from the index case) 21% of the time, and the rate of spread was predicted to be 67 km year^–1.²⁴ In the case of domestic dog populations, the epidemic peak was predicted to be reached on average after 93 days.²³ Both domestic and wild dogs can have great cultural and social importance in these communities.²⁷ The impact of rabies in this region would therefore be far reaching.

In addition to dog population management and restrictions on the movement of dogs where possible, vaccination programs have been a foundation of canine rabies control. The northern Australian context, where a rabies incursion is most likely to occur, presents some unique challenges to rabies vaccination. Vaccine delivery is challenging in remote areas, dog population demographics are often uncertain, and social and cultural norms needs to be carefully considered.²⁸ Modelling studies on options for vaccinating dog populations in this region have suggested efficiencies could be gained by targeting subpopulations of dogs based on their roaming behaviour, compared to the standard approach of blanket vaccination of at least 70% of the dog population.²⁹ Targeted pre-emptive vaccination strategies, in which selected dog populations in northern Australia might be vaccinated in anticipation of an incursion, have also been modelled.³⁰ This raises policy issues for the prevention of a rabies incursion in northern Australia. Should we vaccinate, or can we rely on timely detection of a canine rabies incursion and an eradication policy?

During the past 30 years, the risk of a canine rabies incursion in northern Australia has increased due to the ongoing spread of rabies in the eastern islands of Indonesia (such as Timor and Larat) and regional development and commercial activities. Risk assessment informed surveillance and response is now more critical than ever to prevent such an incursion and spread of rabies on the Australian mainland.