Zero human deaths from dog-mediated rabies by 2030: insights from modelling

Disclaimer

The views expressed in this article are those of the author(s). The opinions expressed herein are those of the authors and do not necessarily reflect the views of the World Health Organisation. Publication in Gates Open Research does not imply endorsement by the Gates Foundation.

Background

In over 120 countries around the world, rabies presents a significant threat to human lives and a considerable public health burden. Around 60,000 people die from rabies annually, whilst tens of millions receive costly post-exposure prophylaxis (PEP)¹. The vast majority of these human rabies cases (>99%) are contracted through the bite of a rabid dog². Although fatal following the onset of clinical disease, rabies is entirely preventable. Administration of PEP to rabies exposed persons according to WHO guidelines prevents disease progression and death³. However, PEP does not reduce transmission in source domestic dog populations. Mass dog vaccination has been the foundation for the successful elimination of rabies from dogs in North America, Western Europe, Japan and much of Central and South America². In contrast, mass dog vaccination has barely started in most low- and middle-income countries (LMICs) in Africa and Asia, where the disease remains widespread.

Within the context of neglected tropical diseases (NTDs), zoonoses like rabies have been especially neglected. Rabies was not amongst the diseases prioritized in the 2012 London Declaration, endorsed by international, non-government, government and industry partners that committed over US$785 million to the control or elimination of 10 NTDs by 2020. Investment in the 2012 roadmap overlooked zoonoses requiring control of disease in animal or vector reservoirs⁴. However, rabies was listed amongst five Neglected Zoonotic Diseases (NZDs) within the 66th World Health Assembly resolution on NTDs in 2013⁵. The lack of progress on rabies control in Africa and Asia reflects the omission from high-level political advocacy⁶, and the consequences of minimal investment. While rhetoric on the need for a One Health approach - recognizing the interactions between human and animal health and the required intersectoral collaboration - has increased, investment in veterinary public health to deal with endemic zoonoses has been woeful, with veterinary services focused on control of livestock diseases considered valuable for trade.

WHO and partners recently launched a global campaign to address the neglect of rabies, with the aim of achieving zero human deaths from dog-mediated rabies by 2030⁷. Some regions already had targets to interrupt transmission in dogs or reach zero human deaths. The Americas have progressed remarkably: human rabies deaths in the region have declined by >99% from endemic levels, many countries have not reported human deaths from dog-mediated rabies in over a decade (only three countries reported deaths in 2016⁸), and have ostensibly eliminated transmission, with mass dog vaccination discontinued from large areas. Although the Americas had to reset their target dates and are still to verify freedom from dog-mediated rabies, their successes demonstrate what is possible through coordinated mass dog vaccination. The Southeast Asia and Western Pacific regions targets for zero deaths from 2020⁹ have both been reset following limited progress. The Pan-African Rabies Control Network (PARACON) and Association of Southeast Asian Nations (ASEAN) united regional groups across Africa and Southeast Asia, respectively, and are involved in pushing the rabies elimination agenda. However, countries on both continents have yet to demonstrate significant case reductions. Development of a new NTD roadmap for 2030 that includes rabies provides an opportunity to build on successes and reflect on failures to constructively identify obstacles and overcome them.

For rabies, the new roadmap, for the first time, identifies time-bound targets for countries to achieve the long-term goal of zero human rabies deaths, intermediate milestones of 50% mortality reduction and operational prerequisites of 70% mass dog vaccination coverage in high-risk areas. Modelling contributed to policy development in the build up to and subsequent launch of the ‘Zero by 30’ campaign. In this letter, we describe insights from modelling that can inform progress towards the 2030 NTD roadmap and lay out challenges to meeting these targets.

Insights from modelling supporting policy development

The historic lack of investment in NZDs has previously been attributed to an absence of estimates of their burden on society to motivate policy^10–12. Human and animal rabies is vastly under-reported in most endemic regions due to a lack of surveillance and reporting capacity, contributing to this cycle of neglect^13,14. With this in mind, the Global Alliance for Rabies Control commissioned a study to estimate the global burden of dog-mediated rabies. Building on decision tree models^15,16 and using updated country data, burden estimates were generated that have been widely cited in advocacy^1,17.

Gavi, the Vaccine Alliance, has considered investing in rabies as part of its mission to increase the equitable use of vaccines for children in the world’s poorest countries. The Gavi Learning Agenda on rabies, the outcome of Gavi’s 2013 Vaccine Investment Strategy decision, complemented key policy processes: a Strategic Advisory Group of Experts (SAGE) Working Group reviewed evidence on rabies vaccines and immunoglobulins¹⁸ (July 2016-April 2018) and a review of models on the impact and cost-effectiveness of rabies prevention strategies was undertaken through the Immunization and Vaccines related Implementation Research Advisory Committee (IVIR-AC) in 2017. Both SAGE and IVIR-AC models highlighted that one strategy under consideration, rabies pre-exposure prophylaxis within a routine Expanded Programme on Immunization (EPI), would be expensive and much less cost-effective than direct prevention through PEP¹⁹. For post-exposure vaccination, intradermal (ID) administration was under all conditions more cost-effective than the widely used intramuscular route¹⁹. An abridged one-week ID regimen was universally preferred over other regimens recognized by WHO as safe and effective, with potential to reduce costs and be more resilient to stockouts during outbreaks¹⁹. A dose sparing approach to administration (wound infiltration only) of rabies immunoglobulins was also demonstrated to be more cost-effective than previous recommendations¹⁹. These outcomes considerably simplified recommendations for rabies PEP, which had previously been regarded as unwieldy and complex, and informed the latest WHO position³, providing a preferred strategy for improved PEP access under Gavi investment.

Gavi recently announced their support for rabies post-exposure vaccination beginning 2021, subject to funding availability. Models projected that under improved PEP access (Gavi investment), total human rabies vaccine use could remain similar to the status quo across Gavi-eligible countries from 2020–2035 (^~73 million), but with millions more (17.4 million) vaccinated by switching to the abridged one-week ID regimen²⁰. This would prevent around half a million rabies deaths, making improved PEP access an extremely cost-effective intervention at around $600 per death averted and costing approximately $403.7 million for the 46 currently Gavi-eligible countries. The counterfactual prediction was that although current PEP provisioning should save around 900,000 lives from 2021–2035, over one million people will otherwise die from rabies. This difference is attributed to addressing the market failure in vaccine supply, with the introduction of free point-of-care access increasing bite victim care seeking and compliance. Models compellingly showed how investing in PEP would be a step to improve equity and commitment to universal health coverage²¹. Nonetheless, they also starkly highlighted that without concomitant investment in dog vaccination, the 2030 target of zero dog-mediated rabies deaths cannot be met. Scaling up dog vaccination according to the Zero by 30 strategy is predicted to avert a further 300,000 deaths in addition to those prevented through PEP²⁰. But without dog vaccination, transmission in dogs will become more entrenched and rabies deaths and PEP use are expected to steadily increase in line with population growth.

Vaccinating 70% of dog populations in high-risk areas is a cost-effective and recognized measure to break the rabies transmission cycle^22,23. Empirical work shows that achieving 70% coverage is possible^24,25 and reveals advantages and disadvantages of strategies according to setting^26–28. Models identified heterogeneity in vaccination coverage, and specifically areas of low coverage, as a major impediment to progress^28,29. Coverage gaps can result from limited resources, poor planning, neglect of particular communities, non-participation by local government units and a lack of monitoring and evaluation. Models further suggest that rabies persists through metapopulation processes^30,31. Limited geographical extent of dog vaccinations can therefore leave areas vulnerable to incursions from neighbouring endemic populations^31–33.

Another challenge is that even with dog vaccination, PEP costs will likely remain high unless strategies to rationalize PEP are implemented. Integrated Bite Case Management (IBCM) is advocated by WHO, to reduce the costs of PEP once rabies has been controlled³⁴. IBCM is a strategy that formally engages the medical and veterinary sectors to assess the risk of exposure to rabies and requirements for PEP. Implementation of IBCM in Haiti, a resource poor endemic setting, improved patient care by identifying and treating those at risk³⁵, whilst reducing PEP use by 40–60%³⁶. Current PEP access differs dramatically by setting. Chronic shortages occur in the poorest countries where the majority of patients seek care for bites by rabid dogs, whereas access is improved in some middle-income countries and sometimes hundreds of patients are treated for healthy dog bites for every rabies exposure^37–39. In endgame settings, where rabies incidence is very low or absent, PEP savings could be much higher. Models with conservative implementation of IBCM, i.e. judicious PEP only after countries reach zero human deaths, predict reduced PEP costs by 20–70%²⁰.

Practical implications of the currently proposed goals

To achieve the targets of the 2030 NTD roadmap, considerable progress must be made in the implementation of rabies control and prevention (Table 1). PEP must be distributed in sufficient quantities and equitably to all rabies-exposed people²⁰. Gavi support will be crucial and should provide stability and guaranteed funds to maintain production and supply of human vaccines, and catalyse pre-qualification of new suppliers to address ongoing international shortages. In parallel, in most LMICs, mass dog vaccination must be scaled up then maintained at sufficiently high levels through sustained campaigns to interrupt transmission.

The Zero by 30 strategy ambitiously targets the scale up of dog vaccination in >100 countries over the next decade³⁴. Given the threat of incursions from neighbouring populations, this coordinated approach should accelerate progress. Coordination of dog vaccination across the Americas provides the model for success, but also illustrates how the poorest states impede progress⁴⁰, i.e. circulation persists where implementation of dog vaccination is weakest, seeding outbreaks elsewhere. Dog vaccination campaigns must be effectively targeted, such that high coverage is maintained locally and nationally. Moreover, delivery methods need tailoring to local contexts, which may require door-to-door vaccinations⁴¹, capture-vaccinate-release²⁸, and supplementary oral vaccination⁴², in addition to central point methods²⁶. There is a learning curve to the scaling up of mass dog vaccination. When introduced to communities for the first time, awareness about rabies and trust in dog vaccination may be low, as is the experience and confidence of practitioners, but coverage should increase over consecutive campaigns⁴³. This progressive scaling up is recognized by the operational target for countries to achieve 70% vaccination coverage.

Rabies is highly underreported¹⁴. Validating the targets of zero deaths and 50% mortality reductions will therefore be a challenge. The most recent WHO expert consultation on rabies included, for the first time, chapters with guidance on validation of zero human deaths and verification of dog-mediated rabies freedom². The NTD roadmap highlights timely diagnosis and accurate assessment of risks combined with strategic use of rapid diagnostic testing to improve surveillance, i.e. key components of IBCM. Indeed, IBCM is identified as a strategy that can sufficiently enhance surveillance for verification⁴⁵. Implementation research is, however, urgently needed to provide guidance on operationalization of IBCM across wide-ranging contexts and best practice given anticipated changes over the time horizon until 2030 (Figure 1). Regulatory measures will be needed for quality control of rapid diagnostic tests and affordable production for larger-scale use.

Figure 1. Interventions to reach the Zero by 30 target.

PEP, post-exposure prophylaxis.

Risks

The overwhelming risk to Zero by 30 is the lack of financial support for introduction, implementation and maintenance of mass dog vaccination. Scientific evidence strongly supports the need for sustained high coverage dog vaccination, but most national authorities have not committed sufficient financial resources to support this strategy. Veterinary budgets are typically much smaller than those of health services, and are largely directed towards measures for trade and animal production, not veterinary public health. In most countries that have successfully controlled rabies, the Ministry of Health has led the rabies control programme including financing dog vaccination⁴⁶. Unless Ministries of Health recognize dog vaccination as an essential public health intervention for human rabies prevention, the Zero by 2030 target will not be achieved. Instead, continued circulation of rabies in dogs may lead to an over-reliance on PEP. International investment in dog vaccine and technical support has demonstrably catalyzed action and is needed to overcome barriers that otherwise prevent rabies activities being initiated and adopted as routine by LMIC governments.

To support effective mass dog vaccination, canine vaccines must meet potency and safety standards⁴⁷. Use of cheap (substandard) vaccines is a false economy given the much greater costs of vaccine delivery. Furthermore, cheaper, lower quality vaccines can have fatal consequences and dramatically set back elimination programmes⁴⁸. Countries in Latin America, with long-standing mass dog vaccination programmes and slow progress towards elimination are notably those using vaccines that do not meet international standards⁴⁹. Some countries push for local vaccine production, but quality standards need to be maintained. World Organisation for Animal Health (OIE) vaccine banks are a step towards more affordable procurement of high quality dog vaccines, but budgets need to be committed to dog vaccination for pharmaceuticals to scale up production, and this remains the most enduring challenge⁵⁰. Limited regulatory oversight for animal rabies vaccines, which would not be tolerated for human vaccines, is a failure of One Health.

Even in countries with mass dog vaccination, interruptions are common, from financial and logistical challenges to emergencies that divert funds from rabies. Lapses in mass dog vaccination prolong time to elimination²⁸. Rapid dog population turnover causes coverage to decline and recurrent outbreaks can lead to loss of confidence in vaccination⁴⁸. If dog vaccines are considered expensive, governments may instead promote culling, which is visible but ineffective, or dog owners may be charged for vaccination, leading to ineffectual low coverage⁵¹.

A further impediment is that required improvements in surveillance will result in short-term increases in reported cases⁵². This can have political implications in rabies-endemic countries and lead to resistance to policies that enhance surveillance. Addressing such perverse incentives is a challenge for elimination programmes, where pressure to reach the end can disincentivize reporting. Even under strong surveillance, dog vaccination must be maintained for years after cases have ceased to be reported to prevent re-emergence⁵³.

Future modelling priorities

Modelling now has an important role to play within the 2030 NTDs roadmap, requiring both new and better data, as well as technical development. In anticipation of Gavi investment and with country specific data, modelling should inform the spatial targeting of PEP. Considerable uncertainty in how rabies control and prevention measures will be implemented makes their impacts difficult to ascertain. Forecasting stepwise progress towards goals in the context of data on the roll out of dog vaccination is therefore a priority, with the need for expedient and realistic dynamical models that can be scaled to specific geographies. The size and connectivity of dog populations determines progress towards elimination. Nonspatial models are inadequate for capturing the low endemic incidence of rabies (<1% dogs infected/year); thus, more complex, data-intensive modelling approaches are required. There is scope for such models to prioritize areas for scale up, to predict the duration over which dog vaccinations will be required to eliminate disease and to assess when dog vaccination can be relaxed. Improved data available from the roll out of mass dog vaccinations and post-vaccination monitoring should inform model development and calibration. Improved data collection methods will be invaluable to populating these models. Transition to electronic data collection can provide timely spatially resolved data to better understand effective program implementation⁵⁴. Programs utilizing mobile applications for large-scale, real-time guidance of vaccination teams has shown effective in both collecting data to improve methodologies and rapidly implementing high-coverage campaigns²⁷.

Improved dialogue between stakeholders, with a focus on packaging and communicating guidance from models and data, would bring many benefits. Close collaboration between practitioners and modelers to assess the impact and cost-effectiveness of interventions (PEP, dog vaccination, IBCM) could improve implementation, for example, by identifying areas requiring remedial dog vaccination. Models allow the comparison of alternative control strategies, for example, by illustrating the ramifications of low efficacy vaccines or undetected cases from weak surveillance⁵³. Burden estimates rely heavily on models. Improved surveillance, particularly IBCM, will provide better data to inform and validate dynamic models and burden predictions, transparently accounting for PEP use and directly measuring impact.

Increasingly, there are calls for integrated interventions in dog populations to more effectively combat diseases that they vector, such as leishmaniasis, echinococcus and guinea worm. Models could help build cost-effectiveness arguments for programmatic integration and inform how such interventions are targeted. Finally, as the goal of elimination approaches, models can be used to compare alternative endgame strategies and guide contingency planning for rapid response and enhanced surveillance to maintain rabies freedom.

Conclusions

WHO’s 3rd report on NTDs in 2013 highlighted four major obstacles for rabies: improving access to PEP, scaling up of mass dog vaccination, maintaining support for elimination once incidence is no longer a major public health threat, and weak surveillance to monitor progress towards targets⁴. These challenges remain, but Gavi investment in human rabies vaccines addresses the first, representing an unprecedented and long-overdue recognition of the entirely vaccine-preventable burden of rabies. IBCM can strengthen surveillance over the elimination timeline and prevent PEP costs from escalating, providing a sustainable exit strategy for Gavi and a direct measure of the impact of investment. However, to achieve Zero by 30, Gavi investment needs leveraging to secure country-level and international financing for scaling up mass dog vaccination. Veterinary capacity is limited in most LMICs, with small budgets compared to health, which are not investing in veterinary public health despite the ‘One Health’ rhetoric. Regulatory mechanisms for animal vaccines are much weaker than for human vaccines, even though poor quality dog rabies vaccines and poor implementation of dog vaccination results in the deaths of people. The Zero by 30 strategy relies on countries stepping up their dog vaccination programmes over the next five years. The intermediate target of 70% coverage is the critical directly measurable indicator of whether countries deliver on this commitment and investment will be needed to support countries to do so. If countries do not, modelling will remain a counterfactual exercise to demonstrate what could have been achieved and to starkly measure how many deaths from rabies we, the global community, are willing to tolerate.

Data availability

No data are associated with this article.