Towards “Zero by 30”, the Kenyan case: a systematic exploration of barriers to rabies policy implementation through systems thinking

Abstract

Introduction

Rabies is a neglected tropical disease (NTD) that disproportionately affects marginalized communities, despite the availability of effective vaccines. In response, the “Zero by 30” global strategic plan aims to eliminate dog-mediated human rabies deaths by 2030. However, localized barriers such as underreporting, fragmented vaccine supply and insufficient governance often impede progress, particularly for settings in Kenya.

Methods

To address these complexities, we employed qualitative system dynamics and group model building (GMB). We combined data from expert interviews (34 participants), field observations (five locations), and GMB workshop discussions (20 participants) to identify the feedback loops that link surveillance capacity, effective vaccinations, trust, prioritization and budget allocations in Kenya’s health and veterinary sectors. This approach allowed us to map the interdependencies that lie at the basis of rabies persistence and to pinpoint systemic entry points for policy intervention. This study followed the Consolidated Criteria for Reporting Qualitative Research (COREQ) to document the methods.

Results

Our findings reveal multiple reinforcing and balancing loops involving vaccine accessibility, surveillance capacity and workforce readiness. Underreporting, budget deficits and the risk of so-called success-induced reductions in vigilance repeatedly surfaced as key impediments. Additionally, economic considerations (for example, high out-of-pocket costs and reliance on donor funds) undermine sustained rabies-control efforts. These dynamics directly intersect with the “Zero by 30” objectives, highlighting the need for stable, multiyear funding, cross-sectoral collaboration and innovative vaccination strategies (for example, streamlined pre-exposure prophylaxis regimens).

Conclusions

Stakeholder-validated causal-loop diagrams show that Kenya’s rabies-control bottlenecks – such as vaccine stock-outs, fragmented surveillance and fluctuating political will – form self-reinforcing cycles that stall progress toward “Zero by 30”. By exposing these dynamics, the study identifies policy entry-points that can break the cycles – shifting the conversation from what fails to why it persists. Therefore, the study offers decision-makers a structured systems map for turning systemic insights into effective, context-specific action.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12961-025-01437-2.

Introduction

Neglected tropical diseases (NTDs), including rabies, are important indicators of persistent inequities in access to high-quality health services and universal health coverage [1]. Accordingly, these diseases have been formally recognized as targets for the Sustainable Development Goals (SDGs; specifically SDG target 3.3), highlighting how their control is linked to broader social determinants of health and economic disparities. Rabies in particular is a preventable yet lethal zoonosis that, once clinical symptoms manifest, is almost invariably fatal. It is categorized as an NTD because it disproportionately affects marginalized and impoverished communities, receiving insufficient resources (for example, information and vaccine availability) relative to its considerable health and economic burden [1, 2].

In response to the ongoing threat of rabies – transmitted primarily through dog bites – the World Health Organization (WHO), Food and Agriculture Organization of the United Nations (FAO), World Organization for Animal Health (WOAH) and the Global Alliance for Rabies Control (GARC) launched the “Zero by 30” global strategic plan [3], aiming to eliminate human deaths from dog-mediated rabies by 2030. The rationale behind this plan lies in the One Health concept, wherein human and animal health are deeply intertwined through a shared environment [4]. Rabies exemplifies this interconnection because dogs live in close communal proximity to humans, enabling viral transmission under conditions of poor vaccination coverage and limited surveillance capacity [5]. Even when preventive vaccines exist, they demand multiple boosters, and preemptive, nationwide vaccination is often deemed costly without robust health and veterinary infrastructure [6]. Hence, the global strategic plan underscores five key pillars to end dog-mediated human rabies: (1) mass dog vaccination targeting at least 70% coverage, (2) universal access to postexposure prophylaxis (PEP), (3) strategic pre-exposure prophylaxis (PrEP) for high-risk groups, (4) rabies immunoglobulin (RIG) for severe exposures and (5) raising public awareness and training of health workers. However, the plan emphasizes awareness, PEP and mass dog vaccination as its three core priorities, with PrEP and RIG positioned as complementary measures. Additionally, the “Zero by 30” plan also emphasizes the need for robust surveillance systems across human and animal health sectors through intersectoral communication, integrated data systems, real-time reporting and harmonized metrics to track rabies incidence. Realizing this ambition, however, demands robust One Health collaborations linking veterinary and public health sectors [6, 7] and consistent financing [8].

Kenya’s Strategic Plan for Rabies Elimination [9] adapts this “Zero by 30” agenda to address region-specific logistical and policy challenges [5, 8] – as presented in Table 1. While both the global and Kenyan strategic plans share the goal of eliminating dog-mediated rabies deaths, the Kenyan strategy explicitly accounts for rural vaccine access barriers, local veterinary capacity shortfalls and deep-rooted socioeconomic constraints that reduce community participation in rabies control [10]. Yet, bridging the last mile requires more than technical fixes. Complexity – manifesting in irregular dog populations, patchy vaccine supply chains and remote communities [11] – calls for flexible, context-driven solutions. Grassroots campaigns in Laikipia County demonstrate how challenges such as limited post-vaccination monitoring, inadequate education programs, funding constraints and diverse community contexts can hinder the achievement of 70% coverage, underscoring the need for stronger surveillance, data sharing, community engagement and sustained financing [12, 13]. Although not unique to Kenya, these factors underscore an essential question for Kenya’s rabies strategy: how do systemic barriers spanning both the health and veterinary sectors hinder progress toward national rabies goals, and what policy or programmatic shifts are needed to overcome these barriers?

Table 1.

Operationalization and adaptations of the “Zero by 30” global strategy to Kenya’s Strategic Plan for Rabies Elimination

Policy aspect	“Zero by 30” global strategy [3]	Kenya’s Strategic Plan for Rabies Elimination [9]
Mass dog vaccination	Recommends universal 70% vaccination coverage in dog populations across all rabies-endemic regions to interrupt transmission broadly	Kenya proposed a phased approach to target high-incidence areas first between 2014 and 2019, aiming for more than 70% coverage in these pilot counties for three consecutive years before expanding nationally. The pilot counties are: Kisumu, Siaya, Machakos, Kitui and Makueni. [4] Additionally, Kenya aims to reduce the proportion of stray dogs by 25% in 5 years
Postexposure prophylaxis (PEP)	Advocates universal, affordable access to high-quality PEP, with competitive pricing and establishment of biological banks to support availability, especially in underserved areas	The objective is to avail PEP in 90% of vaccination facilities and to increase PEP completion rate among eligible rabies exposed cases by 80% in 5 years. However, the strategic plan acknowledges high costs and limited PEP access, particularly in rural regions. Therefore, the focus is on increasing PEP stock in high-risk or underserved facilities, reducing travel costs for patients
Pre-exposure prophylaxis (PrEP)	Recommends PrEP for high-risk groups (for example, veterinarians, lab workers), noting it as resource-intensive and not suited for the general population	PrEP focus aligns with “Zero by 30” but includes targeted outreach for the most at risk workers in areas with limited healthcare access. The aim is to provide PrEP to 95% of these workers in the pilot areas in 5 years
Rabies immunoglobulin (RIG)	Included as part of PEP for severe exposures (category III), recommended primarily for high-risk cases owing to high cost and limited supply. WHO-GARC suggests prudent use with biological banks to improve accessibility	Acknowledges RIG for severe exposures but notes high cost and limited availability, especially in rural areas. The objective is to increase RIG use in 80% of the designated health facilities in high endemic areas. The focus of all other areas is on wound care and PEP availability as primary interventions
Public awareness and training of health workers	Emphasizes global awareness campaigns, community education and general capacity building among health workers in rabies-endemic countries, without prescribing specific local structures	The objective is to increase public awareness on the risk of rabies, dog-bite prevention and the importance of local wound treatment after a bite within households in pilot areas by 95% in 5 years. Additionally, they target to train staff in 100% of healthcare facilities on proper dog bite wound cleaning and management within 5 years

Kenya, in particular, has recorded dog-associated rabies for over a century [5], highlighting how longstanding structural and social challenges undermine even well-established interventions [6, 10], as presented in Table 1. Despite renewed attention as a priority disease [14] following the “Zero by 30” framework, rabies still remains a significant public health threat in Kenya, most notably in underserved rural regions where the disease is endemic and health infrastructure is thinly stretched [6]. Figure 1 shows a general rise in rabies cases between 2005 and 2012 in Kenya, ahead of the development of Kenya’s Strategic Plan for Rabies Elimination. Implementation planning began in 2013, followed by a steady decrease in reported cases until 2017. The sharp decline in the number of samples submitted after 2014 coincided with the implementation of Kenya’s National Rabies Elimination Strategy in counties with high rabies circulation. While the reasons for this rapid reduction remain uncertain, the simultaneous rise in reported bite incidents since 2016 [15] underscores persistent systemic challenges in surveillance and reporting [5], where improved case detection and public awareness may temporarily increase recorded human bite incidents from suspected rabid animals. However, the country did not expand beyond the pilot counties after 2019, primarily owing to limited resources – even within those pilot areas [5]. Based on previous modelling estimates by Hampson et al. [16], the per capita deaths in Kenya due to rabies are 1.27 per 100 000 people. Although these data concern bite incidence rather than confirmed human rabies deaths, they nonetheless signal a potential rise in exposure risk and sustained challenges for rabies prevention in the country. Hence, the current evidence shows that while technical interventions – such as mass dog vaccination campaigns and accessible PEP – are crucial, they are rarely sufficient on their own. Structural barriers such as fragmented supply chains, inadequate government prioritization and limited coordination between veterinary and health sectors continue to undercut implementation [8]. Consequently, Kenya has struggled to make the expected strides in eliminating dog-mediated human rabies despite the availability of proven interventions [17].

Fig. 1.

Indicators of rabies prevalence in Kenya. Data sources: Bitek et al. [5], Kenya Ministry of Health 705 A&B Outpatient Summary Reports (revised 2020) and Luka et al. [15]. Suspected rabies samples largely represent animal submissions. Recorded human bite incidents refer to bites from suspected rabid animals

Over the past two decades, several authors have suggested that a systems thinking perspective – an analytical framework that considers the complex, interrelated factors within a broader ecosystem – could help address the complexities that hinder effective rabies control [18, 19]. More recently, this perspective has gained traction within broader One Health literature, where systems-based approaches are increasingly used to understand and manage complex zoonotic and public health challenges [20–22]. Although several rabies-related modelling studies employ dynamic modelling approaches, most focus on dog population dynamics to guide management and dog vaccination strategies [23–27]. Three studies utilized dog demographic data from African countries such as Tanzania and Ethiopia [23–25], while the other two studies geographically focused on India [26, 27]. However, these quantitative studies often rely on assumed or hypothesized drivers rather than investigating the actual system-wide factors at play. Overlooking an in-depth examination of the systemic complexity surrounding rabies control ultimately hampers policy development. Only a couple of studies have explored systemic issues relating to dog vaccination [28] and integrated bite case management [29], and among these, only the former study applied Systems Thinking as a systematic approach focusing on dog vaccination.

This paper aims to bridge that critical gap in literature by examining the systemic obstacles Kenya faces in aligning its localized rabies-control approaches with the “Zero by 30” global objectives. Drawing on prior pilot studies [5, 6] and analyses of vaccine logistics [8], we argue that effective rabies elimination requires more than isolated reforms in dog vaccination and human vaccination (PEP). It hinges upon coordinated improvements in governance, resource allocation, stakeholder engagement and community participation across health and veterinary domains [10]. Notably, a recent benefit–cost analysis indicates that coordinated rabies control efforts in Africa can yield substantial welfare gains at the regional level [30]. These findings suggest that addressing Kenya’s internal barriers will be most effective when aligned with broader continental strategies for rabies control. By juxtaposing data from interviews, workshop discussions and group model-building exercises, this study uncovers how Kenya’s health and veterinary systems interact to exacerbate or mitigate rabies prevalence.

Given the multifaceted nature of rabies elimination, traditional symptom-targeted approaches – such as focusing narrowly on dog vaccination coverage or PEP rates without examining the systemic determinants behind them – often fail to capture the feedback loops that shape disease transmission, resource allocation and policy outcomes [31]. To address these complexities, this study employed qualitative system dynamics methods. Moreover, group model building (GMB) provided a participatory framework to involve stakeholders in mapping these intricate interactions [32, 33]. We show how resource shortages, weak surveillance and public mistrust can reinforce one another and perpetuate rabies incidence. Our analysis applies this lens to expose the systemic challenges that stall progress, proposing actionable steps for a more resilient, cross-sectoral approach – one capable of achieving zero human rabies deaths by 2030.

Methodology

Study design

This study employed a mixed-methods system dynamics design, incorporating semi-structured interviews with expert stakeholders, field research and a 2-day GMB workshop. Given the objective to diagnose what systemic root causes contribute to the prevalence of rabies in Kenya, it was important to develop a so-called helicopter view – see Fig. 2. Hence, the choice of a hybrid research approach was motivated by the need to (1) gather diverse individual insights through interviews, (2) conduct field observations to contextualize and ground the understanding of the current realities and (3) to collaboratively map systems insights via a participatory GMB process to refine and validate previously collected data and insights. The approach draws on established best practices in participatory modelling [34, 35] to systematically elicit and integrate diverse perspectives into a consolidated causal loop diagram (CLD).

Fig. 2.

Study design. CLD causal loop diagram, GMB group model-building, ILE interactive learning environment

For the development of an interview-based CLD and a GMB-based CLD, two complementary frameworks were used to guide data collection and analysis – the “Zero by 30” framework [3] and the WHO Health System Building Blocks framework [36], respectively. The Health System Building Blocks framework was applied during data collection to facilitate the group model building (GMB) discussions, providing a systems-oriented structure that helped participants articulate interconnections between key components of the health and veterinary sectors without delving into the technical specifics of policy documents. By contrast, the “Zero by 30” framework was used during data processing and thematic analysis of interview data to organize findings around globally recognized rabies-elimination policy categories. This dual-framework approach ensured both contextual depth and alignment with international rabies-control objectives.

To explore systemic barriers and challenges to the implementation of Kenya’s Strategic Plan for the Elimination of Human Rabies through expert interviews, systems thinking was triangulated with a hybrid approach of inductive and deductive coding for reflexive thematic analysis – drawing on reference works [37–40]. Given the complexity and variation in implementing rabies control measures, inductive coding is used to add depth to existing themes from the “Zero by 30” framework, enhancing understanding without necessarily forming new themes [39]. This approach is well-suited to uncover the finer details or new insights into recurring issues, such as vaccine accessibility, logistical limitations and political engagement, while also allowing us to expand upon existing themes within the “Zero by 30” framework. The hybrid coding approach is further elaborated in the Additional File 1.

The reflexive nature of systems thinking involved many iterations in the NVivo 14 (Lumivero, LLC) and Stella Architect software (iSee Systems inc.) that facilitated the data analysis and presentation. Moreover, the EQUATOR Network guidelines were used for rigorous reporting of qualitative research. In particular, this study followed the Consolidated Criteria for Reporting Qualitative Research (COREQ) [41] to comprehensively document the methods. The full COREQ checklist for this study is available in the Additional File 1.

Participants and ethical considerations

Participants were selected using purposive sampling to ensure a broad range of expertise and institutional affiliations. Selection criteria were based on their decision-making roles, institutional representation, field experience or direct involvement in rabies policy or initiatives. All participants were either working in Kenya or had professional experience tied to Kenyan rabies control, ensuring that both local and cross-sectoral perspectives were represented. Community-level perspectives were represented both through NGOs involved in organizing local dog vaccination campaigns and through government health officers responsible for community outreach and communication. Demographic details – such as sector affiliation and geographic location – were collected to confirm diversity in the final sample (see Table 2). The final sample involved 34 interview participants and 20 GMB workshop participants, who were a mix of new stakeholders (n = 13) and a subset of interviewees (n = 7). Recruitment occurred via email invitations, followed by virtual meetings to clarify study objectives, obtain informed consent and conduct the in-depth interviews.

Table 2.

Categorization of the study participants

Criteria	Number interviews	Number of GMB participants
Organization type
Government	13	14
Intergovernmental agency	7	0
Industry	2	1
Not-for-profit R&D	1	0
NGOs	5	3
Academia	6	2
Total	34	20
Geographic scope
Local/national	16	20
Regional/global	18	0
Total	34	20
Sector representation
Health sector	20	11
Veterinary sector	12	9
Other	2	0
Total	34	20

Stakeholders are categorized by organization type, geographic scope and sector representation. The sample included representatives from both human and animal health sectors, spanning national and county government departments (including the Directorate of Veterinary Services and Zoonotic Disease Unit), as well as intergovernmental agencies such as WHO and WOAH. The term local refers to participants working at county or community level, whereas national refers to those based at institutions with nationwide mandates (for example, KEMSA)

GMB group model-building, NGO nongovernmental organization, R&D research and development

The study protocol was reviewed and approved by the Ethics Review Committee of KNH-UoN (ref. no. P370/04/2023) and SMEC of KU Leuven (dossier no. G-2021-4053-R2(MAR)), aligning with standards for research involving human participants. All participants were informed about the study’s aims and provided written consent before interviews or workshop sessions. Personal identifiers were removed from transcripts and notes to ensure confidentiality.

Semi-structured interviews

Data collection

Between January 2023 and September 2024, in-depth semistructured interviews were conducted in English with a total of 34 stakeholders with relevant expertise and experience (see Table 2). The interview guide included questions on perceived challenges in rabies control, resource constraints, sociocultural factors (for example, traditional medicine practices) and policy enforcement mechanisms (see Additional File 1). Each interview lasted approximately 45–75 min and was video-recorded with participant consent. Transcriptions were produced verbatim, ensuring accuracy for subsequent qualitative coding.

Construction of the interview-based CLD

All interview transcripts were analysed using inductive thematic coding. Two researchers (L.S. and S.J.) independently coded the material and noting any causality statements (for example, “the problem is that rabies is getting out of the mind of people as a disease that poses a risk, and we have problems with people looking for the medical attention or medical care after an event of exposure”) from which the variables were deduced (for example, “perceived risk of rabies by public” and “health-seeking behaviour victims”). Discrepancies in coding were resolved through discussion or judgement by a third senior researcher (C.D.). The codes formed the basis of an initial CLD representing the interview data – for which the causal statements can be traced in the documentation of the model software. The initial CLD was subject to multiple iterations in consultation with senior researchers (N.V. and C.D.). The final interview-based CLD is available in the Additional File 1. Loops were labelled either reinforcing (R) or balancing (B) on the basis of the sign (polarity) of causal links. A double pipeline line symbol (||) on causal links denotes a relative delay. In a reinforcing feedback loop, an increase (or decrease) in one variable feeds through effects in other variables represented in the loop and through those mechanisms ends up increasing (or decreasing) the initial variable even more. Meanwhile in a balancing loop, an increase (or decrease) in one variable produces feedback effects which decrease (or increase) the starting variable. In other words, reinforcing loops cause changes in the same direction as the initial variable change direction, and balancing loops cause changes in the inverse direction.

Field research and the group model building workshop

Data collection

A focused field investigation was undertaken in Makueni County, a known rabies-endemic region and one of the pilot counties for rabies elimination efforts since 2014 [5, 42]. Makueni County was selected in consultation with local experts, who emphasized its geographical complexity due to bordering Tsavo East and Tsavo West National Parks – a factor contributing to significant wildlife interaction. In January 2025, a consultative meeting with government officials was convened at the Makueni Referral Hospital in Wote, facilitated by the county’s designated research focal person, to assess the current rabies situation. Subsequently, five diverse field visits were conducted at key health facilities identified by the county’s principal public health officer responsible for rabies surveillance. These included Makueni Referral Hospital (Makueni Subcounty), Mukuyuni Subcounty Hospital (Kaiti Subcounty), Makindu Subcounty Hospital (Kibwezi West Subcounty), Kiboko Dispensary (Kibwezi West Subcounty) and Sultan Hamud Subcounty Hospital (Kilome Subcounty). The selection of these sites was based on their role in managing recent outbreak hotspots and their strategic relevance to ongoing surveillance efforts. Observations and field notes collected across these locations provided direct input into the subsequent GMB workshop, shaping the discussion around surveillance, resource allocation and cross-sectoral collaboration for rabies control.

After the field visits, a 2-day GMB workshop was organized in Nairobi (Kenya) following recognized GMB scripts [43, 44] which were adapted for the purpose of this study (see Additional File 1). The sessions were facilitated by a lead facilitator (L.S.), a modeler (S.J.) and a dedicated note-taker (C.D.) documenting key discussion points, emerging loops and any consensus or disagreements among participants. The latter was performed by an experienced scientist who was occasionally involved in the facilitation of several GMB sessions. For clarity and transparency, audio recordings were made during plenary sessions to process all insights and refine the model afterwards. Debriefing of the final GMB CLD was communicated over email with the participants to validate and allow for any additional insights or feedback.

Construction of the GMB-Based CLD

During the first day, participants co-created a causal map on a whiteboard, progressively identifying factors driving or hindering rabies control and categorizing them according to the Health Systems Building Blocks. This categorization facilitated linking the factors into preliminary feedback loops (for example, “R1: impact of unreported deaths,” “R2: importance of surveillance”). The research team used the behaviour of key variables (that is, reported human rabies deaths, number of dog bites) as starting points to build the CLD, capturing insights surfaced during plenary sessions, breaks and small-group discussions. Refer to the Additional File 1 for the GMB workshop script. At the end of the first day, the facilitation team embedded the emerging whiteboard map into a Stella Architect CLD, ensuring a digital record of variables, relationships and loops. The second day was devoted to scenario exploration, policy intervention brainstorming and resolving ambiguities (for example, synonymous variable names or uncertain link polarity). In a dedicated plenary session, stakeholders clarified and refined the preliminary CLD. This process of iterative feedback resulted in an integrated, GMB-based CLD (see Additional File 1), highlighting strategic avenues for improving vaccination rates, surveillance efforts and overall rabies management.

CLD comparison and consolidation

For the purposes of internal data validation, data triangulation was used by comparing both interview-based and GMB-based models for model comprehensiveness [45]. The consolidation process involved the following:

Variable definitions

A short descriptive list of definitions was created for each interview-based variable. These definitions were grounded in participant quotes to confirm the context or intended scope (for example, whether “health-seeking behaviour” strictly referred to seeking government health facilities or included private clinics). Subsequently, a variables definition list was extended upon the consolidation of both models.

Loop-by-loop matching

Feedback loops were compared from the interview-derived CLD with those of the GMB workshop focusing on the dynamics. Where the same fundamental mechanism (for example, high perceived risk of rabies → high health seeking behaviour bite victim → low reported human rabies → lower perceived risk of rabies) appeared in both models – albeit under different labels – it was considered a shared loop. Following Ryan et al.’s grounded-theory aggregation⁠, loops repeatedly mentioned by several stakeholders are retained as they signal convergent mental models [46]. Therefore, loops that surfaced in only one source were further evaluated on: stakeholder emphasis (the extent to which the loop was highlighted or corroborated by multiple participants) and relevance to core dynamics (whether the loop captured a critical driver or constraint, for example, major resource bottleneck, central to rabies control). Where a loop was deemed both well-supported and crucial for explaining disease propagation or intervention impact, it was integrated into the preliminary consolidated CLD.

Naming and merging variables

Although a formal one-to-one variable comparison was not carried out, any synonymous or near-synonymous terms (for example, “health-seeking behaviour victims” versus “victim response to dog bite”) were merged to a single, transparent label for the final model. Variables were frequently based on variable names used earlier in the interview CLD. Eventually, a “variable dictionary” was developed for the resulting list of consolidated variable names (see Additional File 1).

Final validation and refinements

A draft consolidated CLD illustrating both the shared and unique loops was circulated to all participants – interviewees and GMB workshop participants – for stakeholder validation. Two respondents provided feedback, leading to minor revisions in link polarity and variable naming. The final diagram encompassed the most robust feedback loops (common to both data sources) plus selected unique loops deemed critical on the basis of stakeholder emphasis for comprehensive system representation.

Results

The following sections present the consolidated CLD. Through our analysis, five key thematic groups of feedback loops were identified that capture the core challenges and policy opportunities in Kenya’s rabies elimination efforts:

1. Surveillance and reporting dynamics

2. Prioritization and vaccine availability

3. Economic status and funding

4. Workforce capacity and intersectoral collaboration

5. Cold chain equipment, vaccine innovation and vaccine acceptance

Each group reflects a distinct cluster of feedback mechanisms influencing policy, resource allocation and ultimately, the success of rabies control programs. Figure 3 shows an aggregated picture of how these thematic feedback loop groups connect to each other. The sections below detail each theme and report on the loop dynamics. For a comprehensive overview of all connected feedback loops, refer to the Additional File 1.

Fig. 3.

Aggregated consolidated CLD. The dashed arrows represent a weak link, whereas the solid arrows depict strong causal connections. Note: Arrow colours are used solely to distinguish between feedback loops for visual clarity. The label colour of each loop corresponds to the colour of the arrows within that loop; no additional meaning is assigned to colour. Software: Stella^® Architect (version 3.8) from Isee Systems^®. B balancing loop and R reinforcing loop

Surveillance and reporting dynamics

Figure 4 shows a diagram – as part of the consolidated CLD – featuring two reinforcing loops (R1, R2) and one balancing loop (B1). Together, they illustrate how surveillance, reporting and public awareness interact to shape the perceived risk of rabies – and, in turn, influence national policy, funding and community behaviour.

Fig. 4.

Surveillance and reporting dynamics in Kenya. The double pipeline line symbol (||) on causal links denotes a relative delay

B1 – impact of reported deaths on perceived risk by public

When reported deaths increase, it heightens the perceived risk of rabies among the public and decision-makers. Greater concern encourages bite victims to seek medical care. This health-seeking behaviour also depends on external factors such as public education, awareness and engagement with traditional medicine practices, which is often influenced by cultural beliefs and the trusted social status of traditional healers within communities. Intensifying awareness campaigns stimulate health-seeking behaviour, rather than relying on traditional medicine practices. As more bite victims receive timely PEP, the number of reported human rabies deaths decreases. Fewer reported human rabies deaths reduce the perceived risk of rabies by the public, lowering the impetus for seeking treatment – ultimately balancing the system. If perceived risk becomes too low, however, reduced vigilance sets in, allowing reported deaths to rise again and restarting the cycle. Note that in case of reduced vigilance the loop dominance shifts towards unreported rabies deaths given that these bite victims do not seek formal treatment. Nevertheless, they can still end up hospitalized and diagnosed with rabies, leading to reported human rabies deaths and again increasing perceived risk by the public in the balancing loop B1.

R1 – impact of unreported deaths on perceived risk by public

Low health-seeking behaviour (for example, driven by low perceived risk) means more deaths remain unreported which reduces the perceived risk by the public compared with the real risk. Conversely, if more bite victims seek formal medical treatment (PEP) in due time, fewer rabies deaths go unreported. When many deaths go unrecorded, official statistics appear artificially low, causing policymakers and the public to underestimate rabies’ true burden. This further lowers perceived risk, reinforcing a cycle in which underreporting perpetuates reductions in vigilance. However, this reinforcing feedback loop can also spiral into a virtuous cycle of growing perceived risk which enables bite victims to seek treatment and reduce the underreporting.

R2 – impact of surveillance on perceived risk by policy-makers

National political prioritization of rabies control prompts greater national funding for human and animal surveillance programs, including budget allocations for diagnostic tools and laboratory capacity. Kenya’s strategic plan outlines specific targets for improving the sensitivity of human and animal surveillance systems to 80% and 90%, respectively, in 5 years [9]. With increased funding, surveillance capacity expands (for example, more field teams, improved data management), and diagnostic laboratories are built or upgraded. This leads to more comprehensive testing of suspected cases and timely reporting. As labs confirm more cases (or rule them out accurately), epidemiological data become more precise, clarifying the true burden of rabies. This increased awareness boosts the perceived risk by policy-makers, which in turn motivates national political prioritization – closing the loop in a self-reinforcing cycle.

Prioritization and vaccine availability

The second part of the consolidated CLD demonstrates how vaccine availability is a consequence of prioritization, as shown in Fig. 5. This is shown by two reinforcing loops (R5 and R9) and three balancing loops (B2, B6 and B7).

Fig. 5.

Prioritization and rabies vaccine availability in Kenya. KEMSA Kenya Medical Supply Authority, MEDS Mission for Essential Drugs and Supplies, PEP postexposure prophylaxis

B2 – prioritization and animal rabies control budget allocation

In this balancing loop, increased county government budget allocation for dog vaccination lowers out-of-pocket vaccine costs for dog owners, which is otherwise approximately USD 2.50 per dog [9], making the vaccinations more affordable. This reduces the percentage of unvaccinated dogs as more dogs get vaccinated and, in turn, lowers the incidence of infected dog bites, leading to reduced reported deaths. Fewer reported human rabies deaths decrease the perceived risk of rabies by policy-makers, which then undermines national political prioritization and causes the total county rabies control budget to drop. Lower budgets reduce future allocations for dog vaccination, leading the system back toward a moderate equilibrium rather than continued growth in funding and dog vaccinations.

B6 – impact of rabies control prioritization on PEP vaccine affordability

Low national political prioritization for rabies elimination leads to minimal budget allocations at the county level. With fewer resources for PEP vaccine procurement from the Kenya Medical Supply Authority (KEMSA), regional public health facilities often face shortages, forcing dog-bite victims to purchase PEP vaccines on the private market – potentially not WHO prequalified – and incur significant out-of-pocket costs (USD 10–15 per PEP vaccine dose). A crucial issue is the large variability in PEP availability in Kenya. Stockouts have been reported for up to 36 weeks a year in health facilities across the country [8]. PEP vaccine stockouts were also observed at all five health facilities visited during the field research. Poor vaccine affordability, in turn, lowers health-seeking behaviour, resulting in more reported human rabies deaths. As deaths rise, perceived risk escalates, spurring renewed political attention and temporarily boosting county budgets for PEP supply. However, once deaths fall, perceived risk and political will subside again, causing budget cuts and reduced access to PEP. In case of PEP vaccine stockout at KEMSA, health facilities are allowed to procure from the Mission for Essential Drugs and Supplies (MEDS), which comes at significantly higher prices (USD 12.5 per PEP vaccine dose against USD 1.8 per PEP vaccine dose at KEMSA). This impedes the financial sustainability of county governments as the unforeseen PEP vaccine expenses increase – shifting the problem to reinforcing loop R9. This cycle of fluctuating prioritization perpetuates inconsistent rabies control and leaves communities vulnerable to future outbreaks.

B7 – effect of PEP vaccine availability on health seeking behaviour

This balancing loop begins with PEP vaccine supply through the public system (for example, KEMSA) that increases availability at public facilities and lowers out-of-pocket costs. More affordable access to PEP boosts health-seeking behaviour, reducing reported human rabies deaths over time. However, declining reported deaths subsequently diminishes the perceived risk by policy-makers, which weakens national political prioritization of rabies elimination and ultimately reduces the overall county budget for human rabies control. A diminished budget constrains future PEP procurement and supply via KEMSA, pushing out-of-pocket costs back up – making PEP unaffordable, which is what was observed during the field visits. Thus, while initial improvements in PEP accessibility decrease mortality rates, they unintentionally lead to a drop in perceived risk and budget allocations, returning bite victims' situation to a more constrained state.

R5 – accountability of PEP vaccine availability

This reinforcing loop centres on how record-keeping and financial accountability influence PEP vaccine supply. When KEMSA procures sufficient PEP to supply county governments, public health facilities receive ample stock and administer more doses. Thereafter, PEP vaccine usage is recorded in the inventory of health facilities. This lowers the volume of unaccounted vaccine inventory, reducing unforeseen PEP expenses for counties and preventing the accumulation of pending bills to KEMSA. In turn, timely bill payments maintain a stable procurement cycle, ensuring consistent supply of PEP in the future. However, poor record-keeping, for example, healthcare workers who administer PEP vaccines without keeping records, inflates unaccounted inventory and unforeseen expenses, driving up pending bills and destabilizing procurement – thereby undermining vaccine availability in public health facilities over time.

R9 – accountability-trust-prioritization conundrum

Similar to the R5 feedback loop, accurate PEP inventory management helps to maintain financial transparency, reducing the likelihood of unforeseen vaccine expenses for county governments. As these costs become predictable, trust among decision-makers in lower-level health systems strengthens – encouraging national political prioritization for rabies elimination. Greater political will, in turn, elevates the total county rabies control budget, enabling more consistent PEP procurement and supply through KEMSA. Consistent PEP availability leads to more vaccines being administered and recorded; however, introducing incorrect record-keeping impacts county government PEP vaccine expenses and impedes timely bill settlements. These unforeseen expenses negatively reinforce the cycle of accountability and trust between lower and higher-level county government workers and KEMSA.

Economic status and funding dynamics

Negative feedback loops affecting the economic status of Kenya show potential repercussions for rabies control efforts over the long term. The next diagram (see Fig. 6) illustrates the interactions between four reinforcing loops (R3, R4, R7 and R10) and one balancing loop (B3).

Fig. 6.

Kenya’s economic status and funding dynamics. The double vertical lines crossing the arrow (||) indicates a relative delay

R3 – impact of rabies dog vaccination on economic status

This reinforcing loop illustrates how expanded dog vaccination improves public health, thereby supporting broader economic stability, through reduced healthcare expenditure and productivity losses. At the outset, national political prioritization boosts the total county rabies control budget and, ultimately, budget allocation for dog vaccination. As dog vaccines become more affordable, the percentage of unvaccinated dogs falls, lowering the incidence of infected dog bites and thus reducing reported human rabies deaths. Fewer deaths and reduced PEP demand alleviate healthcare costs and time losses associated with treatment-seeking, enabling greater productivity. With a more stable economic outlook, policy-makers may assign greater priority to rabies elimination, expanding the rabies control budgets and closing the loop in a virtuous cycle. Note that the affordability of dog vaccines is determined by government funds as rabies dog vaccinations are provided for free given there is budget. Otherwise, the vaccination cost falls entirely on the account of the dog owner – reducing the incentive for vaccination. In addition to the cost, also the dog owner’s willingness to vaccinate the dog is strongly determined by the social value of dogs.

R4 – effect of agricultural economy on rabies prioritization

Rabies control in dogs can reduce infections in livestock – particularly cattle – thereby helping to prevent economic losses for farmers and improving Kenya’s overall economic status. The resulting economic improvement reinforces national political prioritization of rabies elimination, yielding greater county rabies control budgets and subsequent budget allocations for dog vaccination. However, in light of other competing animal diseases, for example, foot and mouth disease, rabies is not considered economically relevant and is therefore not prioritized (weak link indicated by dashed arrows in Fig. 6) – contributing to the neglected status of the rabies disease. As a result, dog vaccination efforts remain underfunded, leading to an increased proportion of unvaccinated dogs, more rabies transmission to livestock and escalating economic losses for farmers. These losses, often underreported or misattributed, obscure the true burden of the disease – reinforcing the misperception that rabies control yields little return and is therefore not worth prioritizing. In Ethiopia, rabies-related losses in pastoral cattle herds have been estimated at an average of USD 477 per affected herd annually [47], underscoring the significant economic burden of the disease in livestock-dependent communities. Given the comparable pastoral systems in Kenya, similar economic impacts are likely, highlighting the need for targeted rabies control measures in these regions.

R7 – budget competition between county government departments

Reinforcing loop seven is a small loop that captures how funding allocations for human rabies control inadvertently limit the resources available for animal rabies programs, and vice versa. As the human rabies control budget increases, fewer funds remain for the animal rabies control budget, potentially leading to rising dog rabies incidence and eventually greater human rabies burdens. If not adopting a One Health approach, that surge in human cases further justifies expanding the human rabies control budget while constraining the animal rabies control budget. Therefore, the loop reinforces a skewed distribution of resources: whichever side of the competition gains initial momentum tends to pull more funding over subsequent budget cycles.

R10 – impact of rabies human vaccination on economic status

This reinforcing loop demonstrates how improved PEP vaccine accessibility not only decreases the number of human rabies deaths but also bolsters Kenya’s overall economic status, which in turn feeds back into stronger political support for rabies elimination. This enhanced economic condition, in turn, generates stronger political prioritization for the elimination of rabies. However, stakeholders have indicated that the prioritization of rabies elimination efforts is influenced not only by Kenya’s economic status but also by factors such as external donor funding and the perceived risk of rabies. Consequently, short-term improvements in economic conditions or external funding do not necessarily translate into increased prioritization of rabies elimination if the perceived risk remains low. By contrast, as county governments invest in human rabies control, KEMSA procures and supplies more vaccines, ensuring lower out-of-pocket costs and greater availability at public facilities. Affordable, accessible PEP encourages health-seeking behaviour, lowering reported human rabies deaths. However, KEMSA is not procuring enough PEP vaccines because of irregular ordering from counties and pending bills to KEMSA and hence impacting the PEP availability at health facilities. In Kenya, the direct medical cost for a full PEP regimen is estimated at USD 85 per person – imposing a significant financial burden on households in Kenya [9]. Notably, indirect expenses such as travel, medical fees and income loss constitute over 50% of the total costs [48]. In Africa, the average cost per patient is approximately USD 40, representing 4% of the annual per capita gross national income [9]. Therefore, reduced rabies-related deaths ease the economic burdens on families and health systems, elevating the country’s economic performance. A better economy helps sustain national political prioritization and higher rabies control budgets, reinforcing the cycle of investment in human vaccination.

B3 – effect of economic status on external donor funding

This balancing loop highlights how Kenya’s economic status influences the flow of external donor funding (for example, NGOs, Gavi) for PEP distribution. When Kenya’s economy is perceived as weaker – defined by specific thresholds – donors tend to channel more financial support to ensure broad geographical distribution of PEP, including to last-mile dispensaries. This external infusion keeps PEP vaccine accessibility high and reduces human rabies deaths, ultimately supporting incremental economic growth – however, given the geopolitical climate this donor money influx is uncertain. Moreover, as the economy becomes more robust, external donors scale back their contributions, prompting potential gaps in vaccine availability if local budgets do not compensate. This cyclical mechanism seeks to keep the system in balance: rising economic indicators reduce external funding, and reducing funding, in turn, undermines PEP supply, affecting public health, and potentially lowering economic gains.

Workforce capacity and intersectoral collaboration

Figure 7 depicts a diagram consisting of three balancing loops (B5, B8 and B9). These feedback loops illustrate the fundamental importance of workforce capacity and intersectoral collaboration to PEP vaccine accessibility, dog vaccinations and effective coordination of rabies control efforts through increased surveillance capacity.

Fig. 7.

Workforce capacity and intersectoral collaboration in Kenya. PrEP pre-exposure prophylaxis

B5 – workforce capacity for human rabies prevention

Increasing investments in human rabies control efforts raise the healthcare worker capacity to diagnose and prevent rabies, yet – by reducing mortality and perceived risk – eventually dampen the urgency for further investments. Specifically, greater county budgets for human rabies control expand the workforce capacity for prevention, leading to higher PEP vaccine accessibility (through more and better-trained staff and improved logistics). Better accessibility lowers reported human rabies deaths, which, in turn, reduces the perceived risk by policy makers of rabies. A lower perceived risk weakens national political prioritization, trimming future allocations to the total county rabies control budget and ultimately limiting the workforce capacity needed to sustain low rabies incidence.

B8 – effect of PrEP provision to high-risk workers on safety

Making PrEP affordable for high-risk groups (for example, veterinarians) lowers their fear of conducting dog vaccination, reducing the percentage of unvaccinated dogs and, ultimately, the infected dog bite incidence among humans. As reported human rabies deaths drop, the perceived risk declines, which in turn reduces national political prioritization and overall county budget allocations for rabies control. Consequently, fewer resources lower the PEP/PrEP vaccine availability at regional public health facilities, forcing high-risk workers to resort to the private market. Therefore, high PrEP costs for veterinary staff reduces their willingness to vaccinate dogs, potentially leading to lower dog vaccination coverage and an increased incidence rate of infected dog bites.

B9 – workforce capacity for dog rabies vaccination

In this balancing loop, greater workforce capacity for animal rabies vaccinations reduces the percentage of unvaccinated dogs, thereby lowering infected dog bites in humans and ultimately reported human rabies deaths. As the reported human rabies deaths decline, the perceived risk of rabies by policy-makers drops, which in turn reduces national political prioritization and thus the county budgets for animal rabies control. These dynamics result in a balancing feedback loop where it is observed that initial gains in dog vaccination capacity gradually taper off if risk perception falls. Note that under impetus of intersectoral collaboration, structural investments in rabies control enables the different county departments to safeguard their rabies control budgets through capacity building in surveillance systems – counteracting the balancing loop B5 and B9.

Cold chain equipment, vaccine innovation and vaccine acceptance

The final part of the consolidated CLD highlights the interplay between vaccine attributes and its effectiveness – and show how vaccine product innovation can improve societal challenges like cold chain requirements, public vaccine acceptance and therapy compliance. This is illustrated in Fig. 8 by two reinforcing loops (R6 and R8) and one balancing loop (B4).

Fig. 8.

Cold chain equipment, innovation and vaccine acceptance in Kenya. NTD neglected tropical disease, R&D research and development

B4 – effect of cold chain equipment on the last mile reach of PEP vaccines

This balancing loop focuses on the tension between expanding PEP vaccine distribution to last-mile dispensaries and the cold chain capacity gap that arises when storage infrastructure lags behind delivery needs. Initially, county budgets for human rabies control help strengthen the existing cold chain capacity (for example, new refrigerators, improved transport). As capacity improves, the geographical distribution of rabies vaccines for human use broadens, raising vaccine availability at peripheral facilities and thereby increasing PEP vaccine accessibility. Subsequently, higher accessibility reduces reported human rabies deaths and lowers the perceived risk, potentially slowing the momentum for further funding.

R6 – trust in health service provision

This reinforcing loop highlights how public vaccine acceptance – influenced by factors such as religious belief and the ease of vaccine administration – affects the number of bite victims treated. As more patients undergo treatment and have a positive posttreatment experience, community trust in health services grows, which further strengthens vaccine acceptance. However, religious or cultural views can lead to lower acceptance and fewer victims getting treated, resulting in reduced trust. Therefore, the loop can spiral in a virtuous direction – if each successful treatment enhances community confidence – or in a vicious direction, if initial hesitance and poor experiences reinforce doubts about vaccination.

R8 – demand driven innovation

Improvements in product design qualities (for example, fewer doses, simpler administration, reduced cold-chain requirements) lower the government’s cost of providing PEP, thereby driving PEP vaccine demand. As demand grows, manufacturers realize greater profit from rabies vaccine sales, incentivizing further R&D efforts to enhance the vaccine’s Target Product Profile. In turn, these innovations further lower costs and enhance ease of administration, reinforcing demand in a virtuous cycle of product improvement and market growth.

Discussion

The findings of this study reinforce earlier evidence from Kenyan case work showing that fragmented vaccine supply chains, chronic underreporting and volatile financing streams remain pivotal obstacles to reaching “Zero by 30” in Kenya [5, 8, 10]. At the same time, our feedback-loop analysis advances previous systems-thinking work. We expand the dog-vaccination-focused causal-loop analysis of Savadogo et al. [28] to a broader One Health governance landscape by revealing how trust-dependent accountability mechanisms and the risk of “success-induced” reductions in vigilance reshape those very dynamics. By applying a participatory systems-thinking lens, this research directly answers to the call by Sintunawa [18] and Abbas [19] for complexity-sensitive rabies research.

Systemic policy insights

Referencing the feedback loops side by side (see Results) highlights how interdependent mechanisms across surveillance, funding and community engagement collectively determine the success or failure of rabies-control efforts. The findings illustrate that isolated interventions (for example, vaccination campaigns or reporting improvements) can produce counterproductive effects when not aligned with broader system dynamics, such as variable budget flows or shifts in perceived risk. Coordinated interventions – ensuring stable financing, reliable vaccine supply, effective two-way feedback reporting and sustained public engagement – are therefore essential to achieve lasting progress toward human rabies elimination in Kenya. Consequently, the following sections provide a more detailed discussion of thematic policy insights in alignment with the “Zero by 30” global objectives. Refer to the Additional File 1 for a summary of the objectives and outcomes stipulated by the “Zero by 30” global strategic plan [3].

Surveillance and reporting dynamics

Together, these loops expose how underreporting (R1), inadequate surveillance infrastructure (R2) and reduced public vigilance after initial success (B1) interact to hamper effective rabies control. This dynamic directly undermines the “Zero by 30” objective to ensure reliable data to enable effective decision-making (objective 2.2). R1 emphasizes that underreporting obscures the true burden of rabies, eroding political support and risking a downward spiral if vaccination and awareness campaigns stall. Kenya currently does not have sufficient laboratory capacity and, therefore, suffers from a reporting bias [5]. R2 underscores the importance of building robust laboratories and ensuring transparent use of allocated funds – demonstrable improvements in surveillance can justify ongoing investment and encourage broader stakeholder buy-in. Based on field research, public health officers and veterinarians have reported that the lack of feedback from laboratories regarding their sample submissions causes them to lose faith in the surveillance system, leading to a reluctance to submit samples proactively. Meanwhile, B1 highlights the danger of reduced vigilance once reported cases decline; without consistent public awareness and feedback reporting systems, perceived risk by stakeholders can fall prematurely, triggering funding cuts and enabling rabies to resurge – putting the objectives of “Zero by 30” at risk (outcome 1.1).

Prioritization and vaccine availability

Currently, Kenya lacks consistent availability of rabies vaccines [8]. Although a reduction in infected dog-bite incidence could theoretically lower PEP demand and influence procurement, stakeholder discussions revealed that in Kenya, these processes are largely decoupled. PEP vaccine procurement and budget allocation are driven less by real-time demand than by perceived risk and shifting political prioritization. This structural disconnect contributes to recurrent stockouts and highlights the need for stronger alignment between procurement planning, surveillance feedback and observed demand patterns. R5 and R9 both underscore that prompt bill payments (for example, to KEMSA), robust inventory tracking and transparent communication about vaccine usage are crucial for maintaining trust and a stable supply of PEP. Meanwhile, B2, B6 and B7 show how budget commitments must persist even if human rabies deaths decline; when perceived risk falls, political prioritization can wane and lead to funding shortfalls. Additionally, the requirement for WHO prequalification of rabies vaccines increases procurement costs for government bodies, leading to unintended consequences such as lower prioritization and decreased vaccine availability. These dynamics are impeding the “Zero by 30” objectives of sustaining commitment and resources (outcomes 3.1 and 3.2), and preventing human deaths from rabies exposures by ensuring equitable, affordable and timely access to vaccines (outcome 1.3). Recent global policy developments could further influence these dynamics positively. For example, GAVI’s new investment in human rabies PEP for eligible countries aims to improve vaccine procurement and supply stability [49]. This initiative may help address the recurrent stockouts and affordability issues identified by participants, particularly in low-resource settings where national budgets remain constrained. To warrant baseline resources for both dog vaccination and PEP, coupled with clear data on coverage rates and cost-effectiveness, helps ward off a resurgence. In essence, across all five loops, transparent record-keeping, accountability, sustained budgets and consistent reminders of rabies risk are indispensable to prevent the system from lapsing into underreporting, underfunding or institutional complacency.

Economic status and funding

These loops illustrate how economic considerations and cross-sector coordination can be pivotal in sustaining rabies control – as the “Zero by 30” plan aims to achieve through sustained commitment and resources (objective 3). R3 and R10 both show that dog vaccination and PEP accessibility, respectively, reduce disease burden and spark productivity gains – yet unless these benefits (for example, lowered healthcare costs, less livestock loss) are measured and showcased, the resulting financial advantages may not translate into deeper political commitment to eliminate dog-mediated human rabies. R4 highlights similar dynamics within the agricultural sector, where guarding livestock from rabies can protect farmers’ incomes and garner broader economic development support. Meanwhile, R7 underscores that unbalanced departmental priorities undermine holistic control: a One Health approach can prevent budgets from skewing solely toward human rabies programs, which leaves animal vaccination underfunded. Finally, B3 reminds policymakers that reliance on donor funding is fragile without robust domestic planning and transparent reporting; abrupt withdrawal of external assistance risks undermining last-mile vaccine delivery unless local resources are firmly in place. Furthermore, national surveillance gaps make it difficult to quantify the aggregate economic impact, therefore, the cumulative effects of these local-level costs and productivity losses could also have broader implications for Kenya’s economy. By integrating economic data, engaging multiple sectors and balancing domestic and donor resources, these loops can reinforce one another for sustainable rabies control.

Workforce capacity and intersectoral collaboration

These loops all revolve around workforce capacity – whether for human rabies treatment, dog vaccination or protecting at-risk workers (for example, veterinarians). B5 underscores the need for ongoing training and multiyear budgeting so that healthcare staff remain prepared even when case counts fall. Similarly, B9 emphasizes longer-term commitments to veterinary personnel, linking performance metrics (like dog vaccination coverage) to funding so that staff capacity does not erode when perceived risk dips. Meanwhile, B8 highlights how affordable PrEP for high-risk groups can bolster mass dog vaccination by reducing fear of exposure. Furthermore, increasing surveillance capacity necessitates more intersectoral collaboration that would shape concerted control efforts – especially in areas near wildlife reserves. Although Kenya has a national-level department for One Health challenges, it does not have formal county-level intersectoral collaboration and, therefore, lacks systematic coordination. Taken together, these loops point to a critical aspect of dog and human vaccinations: maintaining a well-trained, well-protected workforce. Therefore, effective workforce development for rabies prevention is fundamental for achieving the “Zero by 30” objectives (outcomes 1.2 and 1.3) and, ultimately, requires collaboration across sectors.

Cold chain equipment, innovation and vaccine acceptance

These loops highlight the infrastructure, societal acceptance and product innovation dimensions essential to the objective of “Zero by 30” to generate, innovate and measure impact (objective 2). B4 warns that last-mile PEP distribution can falter if cold chain upgrades do not keep pace with increasing vaccination coverage, underscoring the need for multiyear budgeting and targeted monitoring in remote areas. R6 shows how community trust, influenced by religious beliefs and the ease of vaccine administration, directly affects vaccination uptake; without proactive engagement of local leaders and clear communication of treatment or prevention benefits, acceptance can remain low. Finally, R8 stresses that demand-driven innovation (for example, more thermostable vaccines, fewer required doses) can lower government costs and boost accessibility, provided policymakers facilitate agile research partnerships and share cost-effectiveness data with developers. Recent studies further demonstrate the potential of thermostable rabies vaccines to enable community-led vaccination strategies and improve coverage in resource-limited settings [50, 51]. By advancing cold chain capacity simultaneously with public acceptance initiatives and product innovations, Kenya can strengthen all three loops to ensure that no single constraint – logistical, cultural, or technological – stalls progress toward rabies elimination.

Towards “Zero by 30”

The rabies disease imposes a disproportionately heavy economic burden on vulnerable communities, notably through direct and indirect costs associated with PEP or livestock losses. The “Zero by 30” strategic plan does recognize that rabies primarily affects rural areas in approximately 80% of human cases [3], yet its high-level framework leaves open questions on how to address the broader socioeconomic and political issues in which these neglected diseases persist [1]. Although the plan indicates that a detailed work plan with defined milestones will guide implementation, and progress has been made (for example, through the United Against Rabies forum [52] and providing guidance on the development of national strategic plans [53]), our qualitative insights highlight systemic barriers – such as resource volatility and “success-induced reductions in vigilance” (B1 or B2) – that may erode momentum in low-resource contexts, setting the stage for re-emergence if not strategically tackled.

Importantly, the systemic feedback loops identified in this study intersect directly with the established determinants underpinning the “Zero by 30” strategy, particularly, access to PEP, mass dog vaccination and public awareness. Our findings illustrate how barriers to vaccine availability, weak surveillance, two-way feedback reporting and fluctuating political prioritization can undermine these core areas in practice. Building on these qualitative policy insights, we have formulated practical recommendations to address some of the challenges in the implementation of “Zero by 30” policies. While quantitative modelling would be required to prioritize specific policy interventions, the recommendations presented here point to leverage points for strengthening coordination, resource stability and community engagement within the existing “Zero by 30” framework. Readers are encouraged to consult Table 3 for a detailed overview of actionable policy measures tailored to Kenya.

Table 3.

Thematic discussion of practical policy recommendations for Kenya based on qualitative insights

Feedback loop group	Actionable policy recommendations	Contributes to “Zero by 30” [3]
Surveillance and reporting dynamics	Kenya can invest in local diagnostic capacity (laboratories and/or rabies virus antigen rapid test kits), the harmonization of different reporting software tools for both rabies detection and vaccination and routine community education (even when cases drop) to ensure a stable surveillance-and-reporting cycle	Objective 1 (Outcome 1.1) Objective 2.1 (Outcome 2.2), Objective 2.2 (Outcome 2.3)
	Elaborate on standardized operating procedures (SOPs) for consistency in: (1) feedback reporting mechanisms, (2) training of veterinarians and health workers and (3) educational campaigns (aiming primarily at school children)	Objective 2.1 (Outcome 2.1, Outcome 2.2), Objective 2.2 (Outcome 2.3)
Prioritization and vaccine availability	Besides political prioritization, Kenya could start prioritizing a joint budget spanning the different ministries to ensure vaccine availability. By playing on a sense of accountability, Kenya could lock in multiyear budget commitments attached to realistic goals and publicize rabies risk data to increase trust, awareness and preserve consistent funding	Objective 1 (Outcome 1.3), Objective 2.2 (Outcome 2.3), Objective 3 (Outcome 3.1, Outcome 3.2)
	Institute a post-incidence decline safeguard to reduce the chance of so-called success-induced institutional complacency in financing – for example, requiring counties to maintain baseline vaccine stocks and routine vaccination for a minimum number of years after zero cases have been reported	Objective 1 (Outcome 1.3) Objective 2.1 (Outcome 2.2) Objective 3
Economic status and funding	Mandate joint budgeting or cross-department committees that set allocations for dog vaccination and PrEP/PEP so that no single department’s priorities overshadows a holistic approach	Objective 2.1 (Outcome 2.1) Objective 3 (Outcome 3.2)
	Additionally, Kenya can systematically measure how rabies control reduces healthcare expenditure and safeguards livestock revenue, then use this evidence to justify stable funding in county budgets. Demonstrating cost–benefit payoffs helps secure and maintain multisector buy-in and meets the “Zero by 30” emphasis on sustainable financing	Objective 2.2 (Outcome 2.3), Objective 3 (Outcome 3.1, Outcome 3.2)
Workforce capacity and intersectoral collaboration	Intersectoral collaboration at the national level (exemplified by the Zoonotic Disease Unit) should extend down to the county level into a legislative framework. This is fundamental and facilitates prompt implementation of the policy recommendations in all other categories	Objective 2.1 (Outcome 2.1, Outcome 2.2)
	Kenya could provide consistent training and incentives for both medical and veterinary staff, along with subsidized PrEP for high-risk workers (for example, veterinary teams). In doing so, the workforce capacity remains robust even after immediate rabies threats fade – directly supporting the “Zero by 30” push for sustained human and dog vaccination capabilities	Objective 1 (Outcome 1.1, Outcome 1.2, Outcome 1.3)
Cold chain equipment, innovation and vaccine acceptance	Kenya could pair any scale-up in dog or human vaccine distribution with corresponding cold-chain investments, simultaneously engaging local leaders to dispel myths and encourage health seeking behaviour or therapy compliance	Objective 1 (Outcome 1.1), Objective 2.1 (Outcome 2.2)
	Collaboration with research partners to pilot new vaccine formulations could further lower costs and meet the “Zero by 30” objective of expanding both vaccine reach and acceptance	Objective 2.2 (Outcome 2.3)

Moving forward, integrating these systemic insights into ongoing refinement of the “Zero by 30” plan could help operationalize the proposed solutions and objectives through the use of leverage points. Through quantitative simulation modelling and further local stakeholder engagement, rabies-control efforts would better align with local realities. In doing so, simulation modelling could demonstrate the feasibility of human rabies elimination via a set of different scenarios – such as abrupt donor funding withdrawal or coordinated cross-country strategies for the control of rabies in Africa as proposed by Bucher et al. [30]. This approach can strengthen the global goal of dog-mediated human rabies elimination. Moreover, extending such system-focused approaches to other neglected tropical diseases could prove equally beneficial for sustainable development.

Study limitations

Potential sources of bias may arise from qualitative subjectivity in interviews and workshop discussions which can be influenced by participants’ personal experiences and recall, as well as group dynamics, potentially skewing the findings if certain voices or viewpoints were more dominant. Therefore, the study was designed to limit bias through the combination of interview and GMB workshops – the respective transcripts were subjected to a thorough assessment of the different perspectives. Although the interviews were fully transcribed, only plenary sessions of the GMB workshop were transcribed (owing to the technical availability of equipment), relying heavily on note-taking during smaller group discussions. This approach, while pragmatic, may have omitted nuances expressed during small-group discussions. To mitigate bias during the GMB workshop, the facilitation team used structured scripts [43, 44] and encouraged quieter participants to contribute via the use of breakout groups. Moreover, the availability and willingness of participants to attend interviews and workshops may have introduced sampling bias, as those who participated could be more engaged or have stronger opinions on rabies control compared with the broader stakeholder population.

In addition, for clarity in system mapping, bite victims were treated as an aggregated group to illustrate overarching feedback dynamics between perceived risk, health-seeking behaviour and PEP access. This simplification helped visualize the structural relationships within the system but may obscure subgroup-specific nuances. Future quantitative modelling could further disaggregate victim profiles to examine differential exposure risks and care-seeking behaviours.

Furthermore, we acknowledge that the participant composition leaned slightly toward human health personnel. Although we invited stakeholders from the veterinary and animal health sectors in proportions similar to those from the human health sector, response rates were lower among the former. As a result, participation from animal health professionals was somewhat underrepresented during fieldwork. However, their perspectives were still captured through targeted interviews with veterinary officers from, for instance, the Directorate of Veterinary Services and the Zoonotic Disease Unit. Moreover, during the GMB workshop, veterinary participants were among the most active contributors, ensuring that animal health perspectives were prominently reflected in the causal loop development and interpretation of system dynamics.

Lastly, as the field study focused on Makueni County, the field observations may have introduced a geographic sampling bias – not reflecting contexts elsewhere with different social, economic or ecological characteristics. Despite these potential limitations, the loop-by-loop comparison method – coupled with stakeholder validation – helped ensure the final consolidated CLD accurately reflected the system’s complexity and core stakeholder perspectives.

Conclusions and future research

This study highlights that Kenya’s progress toward rabies elimination is shaped by interdependent feedback across the human and animal health systems, which can undermine long-term commitment to the “Zero by 30” policy objectives. This is an insight consistent with emerging One Health Systems Thinking literature that emphasizes the need to address governance, coordination and resource flows alongside technical interventions. By mapping these loops through stakeholder-validated CLDs, we reframed vaccine stock-outs, fragmented surveillance and wavering political support as mutually reinforcing dynamics. That systems perspective reveals actionable entry points – stabilizing cross-level financing, embedding transparent inventory tracking and aligning human- and animal-health reporting – that governments can target to unlock progress toward zero human rabies deaths. In short, our work frames rabies control as a complex adaptive system that shifts the conversation from what fails to why it persists, offering decision-makers a structured roadmap for turning systemic insights into effective, context-specific action.

As future research is concerned, recent studies have begun to highlight innovations in PrEP regimens, including intradermal administration that shortens vaccination schedules and reduces the need for multiple follow-up visits [54]. Moreover, emerging clinical evidence indicates that a single-dose intramuscular PrEP provides an immune response comparable to current WHO-recommended prime-boost strategies. If proven reliable at scale, such simplified regimens could shift the economic balance in favour of cost-effective PrEP [55], potentially integrating rabies vaccination into routine practices in high-endemic countries. While the present qualitative findings did not identify PrEP availability as a major systemic constraint in Kenya, these innovations exemplify the kinds of new interventions that could be explored using system-dynamics simulation to assess their interaction with broader financial, logistical and policy factors. Building on these insights, our next step is to translate the loop-based findings into a quantitative system-dynamics simulation that evaluates, in a system-wide way, these emerging PrEP strategies. However, a large-scale adoption of PrEP would not eliminate the need for PEP in the event of rabid dog bites, and without concurrent mass dog vaccination, the overall transmission risk would likely persist or increase. Hence, any expansion of PrEP should occur alongside sustained investment in dog vaccination programs and surveillance. Nevertheless, a quantitative assessment of emerging PrEP strategies will inform the ongoing debate on PrEP and One-Health recommendations in resource-limited settings.

Abbreviations

CLD

Causal-loop diagram

COREQ

Consolidated Criteria for Reporting Qualitative Research

FAO

Food and Agriculture Organization of the United Nations

GARC

Global Alliance for Rabies Control

GMB

Group model-building

KEMSA

Kenya Medical Supply Authority

KNH-UoN

Kenyatta National Hospital-University of Nairobi

LMIC

Low- and middle-income country

MEDS

Mission for Essential Drugs and Supplies

NGO

Nongovernmental organization

NTD

Neglected tropical disease

One Health

Collaborative approach integrating human, animal and environmental health

PEP

Postexposure prophylaxis

PrEP

Pre-exposure prophylaxis

R&D

Research and development

RIG

Rabies immunoglobulin

SD

System dynamics

SDG

Sustainable Development Goal

SMEC

Social and Societal Ethics Committee (KU Leuven)

WHO

World Health Organization

WOAH

World Organization for Animal Health

Author contributions

Conceptualization of study design: L.S., C.D. and N.V.; data collection: L.S., S.J., B.M. and C.D.; data analysis: L.S. and S.J.; internal validation, L.S., S.J., V.S., B.M. and C.D.; writing – original draft preparation, L.S.; writing – review and editing, L.S., S.J., V.S., B.M., V.O., C.D. and N.V.; visualization, L.S. and S.J.; funding acquisition, C.D. and N.V. All authors have read and agreed to the final version of the manuscript.

Funding

This research was funded by the Internal Fund of KU Leuven, grant number C24M/19/006, and received no external funding.

Data availability

The datasets generated and analysed for this study can be found in the Additional File 1 (.pdf).

Declarations

Competing interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential competing interest.