Ethics statement

The protocols for this study were approved by the Institutional Ethics Committee of the Institute of Public Health (IPH IEC), Bangalore (Study ID, IEC-FR/04/2017) and received a Favourable Ethical Opinion from the Liverpool School of Tropical Medicine Research Ethics Committee (research protocol 17/062). All workshop and interview participants were adults and provided informed consent via email through acceptance of the workshop invitation or verbal consent for interviewees.

Reviewing current management recommendations and empirical support: A coproduction approach

This review is part of our interdisciplinary One Health Indo–UK partnership, the MonkeyFeverRisk project (https://www.monkeyfeverrisk.ceh.ac.uk/), which together with disease managers and policy makers across the public health, animal health, and environmental sectors, aims to improve our understanding of risk factors and coproduce guidance and decision supports tools for KFD [23]. Embedded in this One Health network, we reviewed the current management practices undertaken to prevent human cases of KFD, identifying guidance documents and sources in the grey literature available to disease managers. In order to review the empirical evidence for these management practices for the KFD system, we conducted a literature search for the search term “KYASANUR” using Web of Science and PubMed for peer-reviewed literature and discussed with key stakeholders to ensure that further grey literature was not missed. In order to identify literature pertaining to the ecological evidence underpinning management practices for tick control for the main tick species involved in KFD transmission, we also conducted searches for “HAEMAPHYSALIS AND INDIA.” Where there was no direct empirical evidence for a specific management practice impacting on KFD or Haemaphysalis tick vectors, we searched for additional evidence from other tick-borne infections, for example, from well-studied systems such as Lyme disease. However, a full review of topics, for example, repellent and acaricide efficacy, was beyond the scope of this paper. Therefore, searches via PubMed and Web of Science were made primarily for review papers on specific topics. Search terms used to identify literature are provided in Table 1. Some recommended management practices, such as wearing protective clothing and washing of clothes to remove ticks, are well-accepted interventions that are endorsed by global health bodies such as the World Health Organization (WHO), and we cite such web pages where appropriate (Table A in S1 Appendix).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Search terms used to identify literature in Web of Science and PubMed providing empirical evidence for disease ecology and transmission of KFD and for the effectiveness of current management recommendations.

https://doi.org/10.1371/journal.pntd.0009243.t001

We supplemented the review of documents and published literature on KFD management with a series of key informant interviews to ascertain views and experiences with KFD surveillance and control. Interviews were conducted (between July and August 2019) with district and taluk public health managers (N = 11) from Shivamogga headquarters, Sagar and Thirthahalli taluks in Karnataka, directly responsible for KFD management. The selection of interviewees was based on purposive and snowball sampling, which are advantageous techniques in situations where existing networks of relevant people are lacking. Purposeful sampling identifies a key set of people with the relevant expertise and/or experience to provide information and insights on the topic. Snowball sampling then broadens the net of potential interviewees when further relevant people are identified by the interviewees from the purposive sampling. Interviews were conducted in Kannada or English based on participant preference and were transcribed and analysed using an open thematic coding approach [49]. In addition, we also collated and analysed messages shared on our MonkeyFeverRisk WhatsApp platform, a group numbering 121 participants set up in 2018 at the request of our stakeholders to facilitate knowledge exchange on KFD between researchers and practitioners working in different sectors and affected areas. The messages reported here were specific questions posed to researchers by participants about KFD management that illustrated evidence and knowledge gaps about how environmental or wildlife management aligns with KFD management guidance (Table C in S1 Appendix). For the interview data, identified patterns and contradictions from the data were coded and organised into emergent themes. These themes were reviewed to ensure they accurately reflected the meanings evident in the data. The initial findings were noted and subsequently compared with the document review (of current management practices) based on which recommendations for effective and integrated management are made [50]. Participation was voluntary, and interviewees gave their full prior informed verbal consent before the conduct of the interviews, which lasted between 30 and 45 minutes. The interview data were anonymised to protect the confidentiality of participants. Table A in S1 Appendix outlines the main themes and exemplar quotes identified through the key informant interviews, and Table B in S1 Appendix identifies key informants.

We extended the framework of Sokolow and colleagues [10] to consider additional barriers that operate for vector-borne zoonotic diseases compared to those that are directly transmitted. Vector-borne pathogens not only have to evade the immune systems of reservoir hosts and humans to effect spillover, but also have to overcome several tissue barriers and immune responses within the body of the arthropod vector in order to survive and replicate to transmissible levels (see review by [51]). A key prerequisite for spillover of vector-borne zoonotic pathogens is that their arthropod vectors are sufficiently abundant, widespread [52], and overlap in the same habitat with key reservoirs and with people [53]. Furthermore, spillover host exposure depends on how human activities within the ecosystem interface with vector and reservoir host habitats and behaviour [54] and whether vectors prefer to bite both people and reservoirs [55]. We then mapped the existing key management recommendations identified from our review onto our new extended framework for vector-borne zoonoses.

We summarised our overall assessment of the validity of current management practices by scoring each recommendation using separate traffic light scales for the degree of empirical support and management effectiveness. For assessment of empirical support, red indicates no or poor support; amber indicates some support from observations and laboratory studies but lacking rigorous empirical data in a field setting; and green indicates good empirical support including rigorous empirical field data. For management effectiveness, red indicates that the management practice is unlikely to significantly reduce human cases of KFD; amber indicates that it is unknown whether the management practice will reduce human cases; and green indicates that the management practice will reduce human cases of KFD. Our proposed research and surveillance priorities were targeted at addressing knowledge gaps (a) to improve the existing management practices (short-term priorities); or (b) to allow more integrated or ecological interventions to be implemented in the future (long-term fundamental research priorities).

Preventing tick bites on people through protective measures

Recommended protective measures, such as wearing long clothing during visits to the forest, application of tick repellents to the body and clothing, and washing the body and clothing after returning from the forest, are supported as being effective for preventing tick bites by robust empirical data from other tick-borne disease systems (refs in Table A in S1 Appendix). These measures are endorsed by health organisations such as WHO. However, there is limited empirical data for the KFD system, particularly around the effectiveness of locally available repellents, leading to the identification of several short-term research priorities (Table 3). Indeed, there is an urgent need to systematically review the natural repellents currently used by people in the Western Ghats [58] and test their efficacy compared to repellents recommended and distributed by the Indian Government (DHFWS), primarily dimethyl phthalate (DMP) oil, or those recommended by the WHO (Priority 1, Table 3). Natural repellents such as Malabar catmint Anisomeles malabarica have good efficacy against Haemaphysalis bispinosa ticks [59] but have not been tested robustly on H. spinigera and other known KFDV vectors. Numerous studies have assessed repellence of natural remedies and report variation in efficacy between vector species (reviewed in [60–62]), but comparisons between studies are hampered by a lack of robust standardised tests [60,63]. Standardised assays are needed (Priority 2), including toxicity tests (via in vivo and in vitro methods [63,64]), to assess safety and tick repellent efficacy of substances in laboratory and field conditions against tick species relevant to KFDV transmission. Such testing would require development of a bioassay facility in which pathogen-free colonies of potential tick vector species in the Western Ghats can be reared. Furthermore, rigorous experiments are needed (Priority 3) to determine whether ticks survive washing of clothes, whether such ticks pose a risk to humans if they then drop off clothes that are being dried in the environment around houses, and for what length of time clothes need to be hung up outside under different conditions before they are free of ticks and safe to be worn again.

Current management recommendations and surveillance strategies assume that people only need to protect themselves from exposure to infected ticks within forests. There is some empirical support that forests support tick species implicated in KFDV transmission [65]. However, evidence from other (sylvatic) tick-borne disease systems indicates that fragmented and ecotonal habitats at the interface with human habitation may pose a risk of infection to humans, due to such habitats supporting reservoir hosts or substantial densities of infected vectors [66–69]. Other habitats featuring in the agroforest landscape mosaic of the Western Ghats, such as fallow land, paddy, and plantation, may also pose a risk, but there is little empirical evidence on comparative densities of vectors and hosts in these habitats. A key research priority therefore (Priority 4, Table 3) is to quantify the abundance and infection rates of different tick vector species across a broad spectrum of habitats within agroforest mosaics, including human habitation, to understand how risk of exposure to infected ticks extends out from forests. Results of such empirical research in 2 affected districts in the Western Ghats is forthcoming from the MonkeyFeverRisk project of which this evaluation of management strategies is a part (https://www.monkeyfeverrisk.ceh.ac.uk/). Current tick surveillance simply targets forests [65], but stratified tick surveillance sampling and analysis by habitats and season could be integrated into the routine tick surveillance.

Overall, we recommend a new integrated approach at the local community level to prevent KFDV infection from tick bites. For example, current guidance on the use of protective clothing fails to include the need to cover up the feet and does not give guidance on the importance of regular full body checks for attached ticks or advice on how to promptly and effectively remove these ticks (Table 3). Educational leaflets and videos should be developed jointly between the health department and local communities that give guidance on best practice for preventing tick bites and removing ticks (recent community guidance material has focused on disease symptoms and vaccination and less on tick bite prevention) while also taking account of potential trade-offs between recommendations and community livelihood and health priorities (Tables A and B in S1 Appendix). Indeed, as part of our MonkeyFeverRisk project, we have codeveloped such community guidance in local languages (Kannada and Malayalam; education materials can be accessed here: https://www.monkeyfeverrisk.ceh.ac.uk/kfd).

Preventing tick spread by cattle through protective measures

Existing management guidance recommends the application of acaricides to reduce tick loads on cattle and to prevent transportation of ticks from forests to areas of human habitation (Table A in S1 Appendix, Table 2). Cattle do support high numbers of ticks, including adult ticks and H. spinigera (the putative main vector species), although the tick species most commonly found on cattle have not yet been incriminated in the KFDV transmission cycle [44,45]. Hence, cattle may act both as an amplifier of tick numbers and as a disperser of ticks between habitats. Moreover, correlative modelling found that the spatial risk of human KFD cases was associated with the density of cattle in areas long-affected by KFD [70]. A human case–control study from the 2011 to 2012 KFD outbreak also identified handling of cattle as a significant risk factor for infection [35]. Conversely, cattle may act to dilute infection, since they do not show systemic infection with KFDV [42,43]. There is evidence from other tick-borne disease systems that increased density of ungulate hosts, which can amplify ticks but do not have systemic infection, may dilute pathogen transmission by diverting tick bites from competent hosts [71,72]. Although cattle grazing is small scale in the Western Ghats, with small numbers of cattle owned by local households (mean 2.5 per household (range 0 to 15, n = 229) from our MonkeyFeverRisk household survey data), there is also potential that cattle grazing may alter habitat composition and reduce or increase habitat suitability for ticks. Resolving these contrasting potential roles of cattle in KFDV transmission, by comparing the abundance, composition, and infection rates of ticks between cattle, wildlife hosts, and people, in varying ecological settings is therefore an urgent research priority (Priority 5, Table 3).

It is clear from the experiences of disease managers (Table A in S1 Appendix) that farmers cannot always exclude cattle from forests to avoid tick exposure, since the resulting shortages of fodder can lead to animal health and livelihood impacts. This highlights the imperative to identify effective personal protective measures for animals and people that cannot avoid the forests. Depending on resistance, acaricides are effective at reducing tick burdens on livestock, which may act as vectors for a range of diseases [73]. Such tick-borne infections have significant impacts on the livelihoods of poor farming communities in tropical areas by decreasing output of animal products, including manure which contributes to crop production loss [74–76]. Therefore, although the role of cattle in KFD dynamics is unclear, current management practices may prevent transmission of other significant mite- or tick-borne diseases of animals and humans which are known to be present in KFD-affected areas, including scrub typhus, Lyme disease, Indian tick typhus, babesiosis, theileriosis, and anaplasmosis [76–80]. Regardless, further testing of the effectiveness of different formulations of acaricides, including optimal doses and frequency of application, is needed, including for natural repellents that may be more readily available and commonly used within Western Ghats communities (Priority 7, Table 3). It is also essential to ascertain whether resistance against acaricides that are commonly applied both to animals and to the habitat is widespread in tick populations in India, particularly in tick species involved in KFDV transmission (see review in [81]; Priority 8).

Avoidance of human activities

Several of the current management recommendations for KFD seek to reduce human exposure by setting out activities that should be avoided (Table A in S1 Appendix, Table 2). These include avoiding visiting forests where monkeys have died (see next section), not sitting or lying on the ground or in bushy areas of forest, and avoiding bringing dry leaves into villages. Dry leaf litter is important not only for animal bedding and fodder but also for fertilising crops where alternatives are not available [82], and the importance of this practice constituted a key theme of questions from disease managers (Table D in S1 Appendix). Although correlative modelling suggests that the presence of piles of dry leaves around the compounds of the house was associated with human disease cases [35], empirical data are needed on the abundance, infection rates, and survival of ticks found in different types of dry leaf litter, used for animal fodder and bedding, and subject to different treatments in villages (Priority 6, Table 3). More generally, the rate of contact between people and ticks during different activities in and around the forest should be quantified (Priority 9, Table 3) to obtain a clearer picture of which activities cause highest exposure ([83]; https://www.monkeyfeverrisk.ceh.ac.uk/).

Measures to reduce density and distribution of infected ticks

The measures currently recommended to reduce the density and distribution of infected ticks include controlled burning of dry leaves and bushes in forests and around houses; dusting of insecticide (malathion) within 50 feet of a monkey carcass; and burning of monkey carcasses. There are no studies of the impacts of controlled burning regimes on potential tick vectors of KFDV in India. Evidence from other disease systems shows that burning can both reduce or increase tick abundance and infection rates and that impacts may vary with burning regimes, time, and with complex local responses of hosts, vectors, and vegetation [84,85]. Experimental investigations of the impacts of burning regimes on densities of infected nymphal ticks and their contact rates with people over short and long timescales should be conducted in the Western Ghats.

Although infected monkey carcasses may pose a risk to human health through potential direct transmission of infections via contaminated bodily fluids, burning of monkey carcasses and habitat to reduce infected tick densities are perhaps the least supported by empirical evidence among all the recommendations for KFD management and are the least likely to be effective at reducing human cases. The latter 2 measures are predicated on the assumption that humans become exposed to infected nymphal ticks mainly at hotspots around monkey deaths. Although it seems very unlikely that only the immediate area around a dead or dying infected monkey supports high density of KFDV-infected partially fed ticks or fully fed larvae (which would then feed again as nymphs after moulting) that may pass infection to humans, this should be confirmed empirically. The measures further assume that insecticide or burning will kill these partially fed ticks before they bite people (Table A in S1 Appendix). KFDV-infected monkeys are known to have high titres of virus [86] and are likely to bear infected ticks. Although partially fed nymphs have been found in areas around monkey carcasses [87] and engorged ticks have been observed to move up to 30 cm [88], no empirical data exist showing that successful interrupted feeding (intra-stadial feeding) occurs in tick species commonly found to transmit KFDV in the wild. In other tick-borne disease systems, ticks generally feed on only 1 host in each of their life stages [89]. Interrupted feeding has very rarely been recorded and most often under laboratory conditions and in Rhipicephalus spp. where male ticks actively seek female ticks with which to mate [90,91]. Experimental transmission studies on partially fed ticks collected from areas close to monkey deaths suggest that these have limited potential to transmit virus by feeding on a second host [87]. Thus, although dead KFDV–positive monkeys undoubtedly indicate ongoing KFDV transmission in an area, there is no direct evidence that locations of monkey deaths are hotspots of host-seeking infected ticks, compared to surrounding areas of habitat where monkeys or small mammal hosts have spent time. Moreover, evidence is building of malathion resistance in tick populations and off-target effects on human and animal health, such as increased cancer risk, elsewhere in India (Priority 8, Table 2; [92–94]).

The confusion surrounding the role that monkeys play in KFDV transmission, compared to alternative small mammal hosts, highlights the need for fundamental ecological research. Priority should be given to quantifying the tick burden on dead and dying monkeys, assessing the number, species composition, life stage, and KFDV infection rate of these attached ticks, as well as determining the number, KFDV infection rate, and blood meal identity [95] of ticks sampled from the habitat or off people near the monkey death site (Priority 10, Table 3). It also needs to be ascertained whether and how quickly partially fed ticks begin to quest, the distances they can cover, and the survival rates of non-questing partially fed nymphs and larvae that moult to the next life stage (Priority 10, Table 3).

Sampling of ticks from monkeys, small mammals, and people over a more diverse range of habitats would indicate whether monkeys are sentinels of infection or whether they contribute to KFDV transmission across broader spatial scales, by the systemic infection of large numbers of larvae or via co-feeding (transmission between larvae and nymphs feeding in close proximity; Priority 11, Table 3). Although small mammals are neglected completely in current management recommendations, given their central role in amplifying transmission in other tick-borne systems through processes such as co-feeding [96], it is equally important to clarify their role in the KFDV transmission, by sampling them in similar ways alongside monkeys and humans (Priority 12, Table 3). If small mammals and monkeys are confirmed as being important reservoirs and amplifying hosts for KFDV, then empirical data on their movements and habitat associations are required in order to quantify how far they tend to travel in the landscape (when healthy or sick for monkeys), in which habitats they interact with people and ticks, and the predictability of their movements in time and space (Priority 14 and 15, Table 3). This is vital for matching the scale of surveillance, awareness raising, and vaccination to the scale over which spillover from monkeys or small mammals occurs.

Phylodynamic approaches may be particularly promising for improving the understanding of KFD transmission. As an RNA virus, KFDV is characterised by rapid rates of molecular evolution [97,98]. Nucleotide sequence data can be used to identify genetic variation between virus population samples in time and space, infer transmission pathways, and identify dispersal and broad scale movements of pathogens between areas, offering new opportunities for elucidating the scale of KFD risk and understanding human spillover (Priority 13, Table 2; [99,100]). Moreover, assessment of the genetic diversity of KFDV strains would greatly facilitate the development of more effective vaccines (Priority 18, Table 3), as the existing vaccine, developed using virus isolates from the 1960s, may have reduced efficacy against more modern strains, as has been shown for other viruses [101].

Measures to reduce susceptibility of people as the spillover host

A formalin-inactivated tissue culture vaccine has been used in Karnataka since 1990 [35]. The strategy of vaccinating people within a certain radius of known cases or circulation of KFDV in ticks or monkeys is well underpinned by evidence, and this vaccine is known to give protection against KFD if the correct dose procedure is followed [102,103]. However, duration of immunity from the vaccine is short, requiring multiple repeated doses, and acceptance of the vaccine and coverage rates has been poor in some areas [34,35,102], hampered by limited seasonal vaccine availability and due to social and cultural factors that need to be further investigated ([104,105]; Priority 15, Table 2). This highlights the importance and urgency in developing complementary disease prevention measures alongside vaccination that follow the new framework for disease prevention set out in Fig 1. In recent years, there is also some evidence that vaccine efficacy was reduced compared to previously [34,102], necessitating robust testing of efficacy and duration of protection offered by the current vaccine (Priority 17, Table 3) and parallel development of novel or alternative vaccines for closely related viruses which may offer cross-immunity [106] (Priority 18, Table 3).

Surveillance in people, ticks, and monkeys to inform interventions

Finally, the current human case surveillance that underpins KFD management is well justified by empirical evidence. Human cases tend to be clustered, and surveillance of human fever cases, recommended to take place between December and May with screening for both virus (by reverse transcription PCR [RT-PCR]) and antibody presence (by ELISA), has been effective in directing vaccination campaigns [34]. Given the shifting pattern of KFD cases, with new geographical hotspots identified each year [107], we additionally recommend that fever case surveillance and serological surveys should be adopted in areas beyond the current known range of KFD cases in order to map the distribution of KFD [40] and predict disease spillover in new areas. However, serological surveys are likely to be more costly than those targeting fever cases. Targeting of human surveillance can again be guided by predictive models (Priority 19, Table 3), as is already happening in Karnataka [70] but also by detailed case–control surveys that can identify high risk livelihood groups and activities (cf. [35]).

Tick surveillance can be an effective way of predicting spillover, as demonstrated in other systems [108,109], and is valuable locally to monitor persistence of KFDV transmission over time. However, as mentioned above, tick surveillance methods could be more standardised, rigorous, and stratified by habitat (extending beyond forests) to build knowledge of tick vector habitat associations (Priority 4, Table 3) to improve targeting of subsequent interventions and surveillance. As with human surveillance, it should routinely encompass areas beyond the current front of human KFD cases. Similarly, monkey surveillance is currently implemented through passive sampling of dead and dying monkeys, and it is restricted to areas where resources are available and awareness of KFD is high (Table A in S1 Appendix). The coverage and value of this passive sampling could be extended by improving education on the importance of reporting dead and dying monkeys (perhaps providing smartphone-based app technology to facilitate such reporting), allowing standardised postmortems and tick sampling around monkey death sites to be carried out more rapidly than is current practice. Additionally, consideration could also be given to targeting small mammal maintenance hosts in future surveillance programmes. However, small mammal surveillance is likely to be time-consuming, infection duration is short lived, and high effort is needed to sample enough individuals to robustly estimate infection prevalence. Such surveillance programmes may be more achievable with improved empirical understanding of maintenance host species in order to hone selection of target species and habitats.

Overall conclusions

Using the tick-borne KFD in India as a case study, we demonstrate how a novel barriers framework ([10,14], Fig 1) can be applied to evaluate the evidence base for current management practices for reducing zoonotic disease spillover. This approach is particularly powerful for KFD, and indeed for other neglected zoonoses such as scrub typhus and leptospirosis in India and other lower middle-income countries, where ecological and social evidence underpinning management strategies is outdated or lacking altogether. We show how this framework can be used to critically evaluate existing disease mitigation measures and focus recommendations for their improvement, identify knowledge gaps and priority areas for research, and highlight potential opportunities for new interventions.

Current management guidance for reducing the risk of human KFD primarily focuses on conventional practices to reduce human exposure and susceptibility to infection, mainly targeting the final, human barriers to spillover (Fig 1). Primary recommendations are for the use of repellents in conjunction with advice for avoiding potentially risky habitats and activities and a targeted vaccination strategy. However, our review clearly highlights the lack of robust, current empirical knowledge on the ecological and social factors leading to human cases of KFD, thus precluding the future development of a broader suite of management interventions, such as ecological interventions. This lack of empirical evidence also means that the effectiveness of current management recommendations is questionable. In particular, the role of cattle and primates in KFDV dynamics is not well understood (Table 2, Table B in S1 Appendix). Thus, management practices such as burning monkey carcasses, malathion spraying, and controlled burning of vegetation around the sites of monkey deaths are particularly unfounded. Such practices are predicated on the untested assumption that monkey deaths reflect localised hotspots of transmission, as opposed to monkeys being sentinels that indicate KFDV prevalence in the area and more widespread transmission in a range of alternative hosts. Some of the current guidance can be improved based on available evidence, and we suggest detailed ways of doing this in Table A in S1 Appendix. Improved ecological understanding of KFD disease dynamics has the potential to lead to a broader and more economical suite of potential interventions that target vector or reservoir hosts and which are more effective in the longer term at reducing human spillover [10]. Indeed, even in similarly complex, but well-studied multi-host tick-borne infections such as Lyme disease, interventions that target nonhuman hosts are possible: There is evidence that oral bait vaccines that target rodent reservoir hosts are effective at reducing the density of infected nymphs [110–112]. Entomopathogenic fungi such as Beauveria bassiana and Metarhizium anisopliae have also shown potential as biological control of tick vectors involved in Lyme disease transmission [113]. Other novel interventions are currently being developed, and much may be learned from the testing of such interventions in tick-borne disease systems where transmission dynamics are well elucidated. Vaccines are being developed, for example, against tick saliva or salivary gland proteins, in order to decrease the ability of ticks to feed on reservoir hosts, thus disrupting pathogen transmission [114–116], and self-disseminating, transmissible vaccines that are capable of spreading through wild animal reservoirs [117,118].

The new framework developed here illustrates the significant number of hierarchical barriers (Fig 1) that must be overcome in order for humans to become infected by KFDV. This further emphasises the need for a cross-disciplinary approach to provide an evidence base and implement appropriate management interventions for tick-borne diseases. Focusing research on a single barrier or barrier type within the hierarchy, within a particular scientific discipline, will not be sufficient for understanding spillover risk or implementing effective interventions [14,119]. For KFD, the ecological processes that underpin seasonal transmission dynamics, such as vector and host activity and habitat associations, seasonal resource use, and movement, need to be quantified alongside the social and cultural processes that influence ecosystem use, livelihoods, and exposure of people. For example, the majority of disease managers interviewed highlighted complex trade-offs between restricting forest access to minimise risk of exposure and safeguarding local livelihoods that need to be further investigated, as well as diverse social, cultural, and techno-administrative barriers to uptake of vaccines, tick personal protection, and surveillance measures, e.g., post mortems (Table B in S1 Appendix). Aligning with the One Health initiative, interdisciplinary approaches require collaboration across diverse disciplines including ecology, epidemiology, animal and public health, health systems, and social sciences and would enable predictive models of pathogen spillover to be refined, new interventions to be developed, and vaccination strategies and surveillance to be targeted more effectively [70].

In order to sustainably and effectively refine management interventions for neglected zoonotic diseases in the face of changing empirical knowledge and environmental and policy shifts, it is vital that strategies and research are codeveloped iteratively and reflexively across disciplines, targeting knowledge gaps and prioritising interventions identified by cross-sectoral stakeholders and involving beneficiaries alongside researchers [70,120]. Within the MonkeyFeverRisk project, empirical research and models were co-created using a coproduction approach, which placed stakeholder engagement at the heart of the research, from joint framing of the problem to knowledge integration and experimentation with resulting knowledge and tools [121–123]. In practice, active science–policy–practice interfaces between a diverse range of stakeholders, researchers, and beneficiaries were established and maintained through multi-stakeholder workshops and focus groups, inclusion of decision-makers as active research partners, researcher membership of government technical advisory committees on KFD, and through cross-sectoral WhatsApp groups setup following a direct request from stakeholders at our first workshop (https://www.monkeyfeverrisk.ceh.ac.uk/sites/default/files/Stakeholder_workshop_report_MonkeyFeverRisk_16012019.pdf). An illustration of the valuable framing of empirical research and subsequent knowledge exchange that can arise through such science–policy–practice interfaces is provided in Table D in S1 Appendix, which shows key ecological questions posed to researchers by practitioners through the 2 recent KFDV transmission seasons. The trade-offs between livelihood benefits and disease disbenefits from forests, the poor uptake of some interventions for KFD (Table B in S1 Appendix), and the wide range of traditional coping methods in use by local communities highlight the importance of meaningful involvement of local communities in the design of management strategies [123], and this is a key priority for effective management of KFD.