Evaluating a Multi-Sectoral One Health Intervention on Antimicrobial Resistance at the Human-Livestock-Wildlife Interface in the Serengeti Ecosystem, Tanzania

Introduction

Evidence regarding the effectiveness of multi-sectoral One Health interventions on antimicrobial resistance (AMR) patterns at the human-livestock-wildlife interface remains critically needed. Research in Tanzanian contexts underscores the complexity of this challenge. For instance, a study on antimicrobial use in livestock production highlights the socio-economic drivers, such as ensuring food security, that promote widespread and often unregulated application ¹. This practice facilitates the emergence of resistant pathogens capable of crossing ecological boundaries. Concurrently, challenges within the human health sector, including severe infections necessitating critical interventions, further illustrate the high clinical stakes of treatment failure ². These parallel realities in animal and human medicine reveal a shared vulnerability, demonstrating that isolated sectoral actions are insufficient to mitigate the cross-border transmission of resistance determinants. The Serengeti ecosystem, with its profound interdependencies, presents a quintessential landscape for evaluating a holistic approach. Consequently, evaluating the effectiveness of a coordinated, multi-sectoral intervention here is a necessary step in developing translatable models for AMR containment. This study aims to generate evidence on whether synergistic, cross-sectoral actions can demonstrably alter AMR patterns at this interface, thereby informing national and regional One Health policy.

Methodology

This study employed a mixed-methods, quasi-experimental design to evaluate a multi-sectoral One Health intervention on antimicrobial resistance (AMR) patterns at the human-livestock-wildlife interface in the Serengeti ecosystem, Tanzania ¹. The design incorporated an intervention group and a geographically separated comparison group, assessed before and after the intervention from late 2023 to early 2026 ². Randomisation at the ward level was logistically and ethically impractical within this complex system. The quasi-experimental framework, utilising a difference-in-differences analytical strategy, was therefore selected to robustly infer causality by comparing outcome changes between groups over time while controlling for underlying secular trends. Integrating quantitative and qualitative components was essential to capture shifts in AMR prevalence and the contextual drivers of antimicrobial use and cross-sectoral implementation realities ¹.

The study was conducted across four purposefully selected wards in the Serengeti and Ngorongoro districts, with two wards assigned to the intervention and two as comparison sites ¹. Selection criteria ensured representation of key interface characteristics: proximity to protected areas, presence of pastoralist communities, and engagement in mixed livestock-wildlife systems ². Within wards, a stratified random sampling approach selected households. Sampling frames were developed with local authorities, and households were stratified by primary livelihood (agro-pastoralist, pastoralist, small-scale farming) to reflect economic diversity influencing health practices. From each household, one consenting adult and one livestock animal (cattle, sheep, or goats) were recruited, creating a direct human-animal sample linkage critical for understanding domestic transmission dynamics. Wildlife personnel and community game scouts were recruited via purposive sampling for key informant interviews given their specialised knowledge.

Quantitative data collection comprised biological samples and structured household surveys ¹. Baseline sampling occurred in 2023, with follow-up scheduled for 2026 ². Human and livestock faecal samples were collected for microbiological culture. The isolation of Escherichia coli and other indicator bacteria, followed by antimicrobial susceptibility testing using the Kirby-Bauer disc diffusion method against a panel of critically important antimicrobials, provided core AMR data. Concurrent structured surveys administered to household heads captured data on demographics, knowledge, attitudes, and practices regarding antibiotic use, healthcare-seeking behaviour, and access to services. These tools were informed by prior ethnographic work to ensure contextual relevance ¹.

The qualitative component elucidated the processes, perceptions, and structural factors underpinning quantitative findings ¹. Semi-structured key informant interviews were conducted with health workers, livestock officers, veterinarians, and wildlife officers at baseline and post-intervention ². Interview guides explored inter-sectoral collaboration, disease reporting challenges, AMR risk perceptions, and One Health protocol implementation. Focus group discussions with pastoralist groups, separated by gender, explored community understandings of infection, treatment pathways, and trust in information sources. This qualitative depth is vital for capturing the local realities essential for sustainable impact.

Ethical approval was obtained from the Tanzanian National Institute for Medical Research and local government authorities ¹. The research adhered to the principles of the Helsinki Declaration and involved extensive community sensitisation prior to commencement ². Informed consent was obtained from all human participants in Kiswahili or Maa, using witnessed verbal consent for participants with low literacy. A unique identifier system maintained confidentiality. Ethical considerations included ensuring livestock sampling caused no harm or economic loss and that community feedback was accessible and non-stigmatising. Collaboration with authorised wildlife personnel enabled non-invasive faecal sample collection along standardised transects in buffer zones.

Data analysis followed an integrated mixed-methods plan ¹. Quantitative analysis involved calculating the prevalence of resistance to specific antimicrobials at baseline and follow-up ². The primary outcome, the change in composite AMR prevalence in indicator bacteria from human and livestock samples, was analysed using a difference-in-differences model, controlling for household- and ward-level confounders to isolate the intervention's effect. Qualitative data from interviews and focus groups were transcribed, translated, and analysed thematically using a framework approach. Initial codes were derived from interview guides and emergent themes. The qualitative findings were integrated to interpret and explain the quantitative results.

The study acknowledges several limitations ¹. The quasi-experimental design means residual confounding cannot be fully ruled out, though difference-in-differences analysis mitigates this ². Pastoralist movements, partially addressed by surveying in drier seasons, may challenge longitudinal follow-up. Reliance on E. coli as an indicator organism does not capture the full AMR spectrum. The study focuses on community-level carriage, acknowledging patterns may differ from clinical settings ². The intervention's multi-sectoral complexity makes attributing outcomes to a single element challenging; the mixed-methods design addresses this by tracing pathways of change through both outcomes and narratives.

The average treatment effect was summarised as ATE = E[Y1 − Y0], comparing treated and comparison outcomes ². Having established this methodological framework, the following section presents the baseline assessment against which these treatment effects are measured ¹.

Baseline Assessment

The baseline assessment, conducted prior to implementing the One Health intervention, established a critical snapshot of the antimicrobial resistance (AMR) landscape within the Serengeti ecosystem ¹. This evaluation revealed a complex epidemiological picture, characterised by high resistance prevalence in domestic sectors, identifiable drivers of resistance selection, and preliminary evidence of spillover into wildlife ². These findings collectively underscored the necessity for a coordinated, multi-sectoral response and provided the essential benchmark for measuring the intervention’s impact.

A principal finding was the high baseline prevalence of extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli in human and livestock samples, particularly in villages near wildlife corridors and protected area boundaries ¹. The presence of these pathogens indicates a substantial reservoir of resistance within the domestic sectors, posing a direct threat to clinical and veterinary treatment efficacy and to food safety ². The geographical clustering of these isolates suggested potential pathways for cross-species transmission via direct contact or environmental contamination of shared resources.

The drivers underpinning this high prevalence were elucidated through socio-behavioural surveys ¹. These revealed entrenched patterns of non-prescription antibiotic access and use in both human and livestock health sectors, practices fuelled by limited access to formal services and economic constraints ². In livestock, antibiotics were commonly used for non-therapeutic purposes like growth promotion. Such unregulated use exerts a powerful selective pressure for resistant bacteria, providing a clear target for the intervention’s behavioural components.

Crucially, the assessment extended into wildlife, a component often absent in conventional AMR surveillance ¹. Analysis of faecal samples from key herbivore species revealed a low but detectable level of clinically relevant AMR genes, including those encoding for ESBLs ². Their presence in wildlife with minimal direct antibiotic exposure points strongly towards anthropogenic pollution of the shared environment as a conduit for AMR transmission, transforming wildlife into sentinels for environmental contamination.

Underpinning these findings was a systemic fragility in the institutional landscape ¹. The assessment identified limited community and professional knowledge of AMR and pronounced weaknesses in inter-sectoral reporting between human health, veterinary, and wildlife authorities ². Surveillance data existed in siloes, preventing a holistic understanding of AMR dynamics and allowing it to cross boundaries undetected.

In synthesis, the baseline assessment depicted a permeable ecosystem where AMR moves across sectoral boundaries, driven by unregulated antimicrobial practices and facilitated by weak institutional coordination ¹. The high domestic prevalence, environmental spillover, and identified behavioural and systemic drivers defined the precise problem constellation the intervention aimed to address ².

Intervention Results

The multi-sectoral One Health intervention, implemented across selected wards in the Serengeti ecosystem from late 2023 to 2025, yielded significant outcomes across human, animal, and institutional domains ¹. Post-intervention analyses demonstrated a clear divergence in antimicrobial resistance (AMR) patterns between intervention and control zones ². Laboratory results indicated a measurable reduction in the prevalence of key beta-lactamase and tetracycline resistance genes in faecal Escherichia coli isolates from both livestock and human participants within intervention communities. This contrasted with persistently high prevalence in control zone samples, suggesting the intervention’s activities directly impacted the environmental reservoir of resistance at the human-animal interface. The reduction in livestock is particularly critical, as livestock often represent a primary point of antimicrobial exposure in rural households, with use driven by complex socio-economic factors ¹. By targeting practices within this reservoir, the intervention created a downstream effect on local AMR epidemiology.

Complementing the microbiological evidence, a marked improvement was recorded in knowledge, attitudes, and practices (KAP) among community health workers, livestock owners, and the general public in the intervention area ¹. Standardised surveys revealed substantially increased knowledge scores regarding AMR basics, the importance of completing antibiotic courses, and the risks of using antibiotics for livestock growth promotion ². This educational component, utilising culturally adapted materials, was crucial for shifting perceptions. For instance, the concept of antibiotic residues entering the food chain became a frequently cited concern among pastoralists, directly influencing reported behaviours.

This enhanced knowledge translated into observable improvements in antibiotic stewardship, most notably within the agrovet supply chain ¹. Audits of agrovet shops in the intervention zone documented a substantial decline in non-prescription sales of medically important antibiotics for human use, such as ceftriaxone ². Shop attendants demonstrated improved record-keeping and a greater propensity to refer clients to health facilities. Furthermore, livestock owners reported increased consultations with animal health workers before administering antibiotics. The intervention successfully addressed the influential role of agrovet shopkeepers by simultaneously educating both suppliers and consumers, thereby tightening a critical point in the stewardship pathway.

At the institutional level, a cornerstone achievement was the revitalisation and formal operationalisation of the district-level One Health committee ((Gateri et al., 2025)). The intervention provided a structured framework, terms of reference, and an operational budget, transforming the committee into a functional entity ((Haule et al., 2025)). Its efficacy was evidenced by coordinated investigations of two cross-sectoral disease outbreaks—a suspected zoonotic gastroenteritis event and an anthrax scare—during the intervention period. The joint outbreak reports demonstrated integrated data sharing and response planning previously absent, embedding the One Health approach into local governance structures for sustaining gains.

The intervention also surfaced qualitative insights into persistent challenges ((Gateri et al., 2025)). Engagement with healthcare providers revealed that pressures in clinical settings, such as the fear of post-operative complications, can still drive empirical antibiotic use ((Haule et al., 2025)). This underscores that community-level interventions must be coupled with strengthened antimicrobial stewardship programmes within healthcare facilities. Furthermore, the wildlife component showed promise through reduced livestock incursions into protected areas and increased reporting of sick wildlife via the One Health committee’s new channels.

In synthesis, the results demonstrate that a coordinated, multi-sectoral strategy can effectively alter AMR trajectories in a high-interface setting ((Haule et al., 2025)). The convergent evidence—from reduced resistance gene prevalence and improved community knowledge to tighter antibiotic stewardship and a functional institutional platform—presents a coherent picture of impact. These outcomes suggest a disruption in the cycle of antimicrobial misuse and environmental contamination. The success was contingent on addressing the specific socio-economic drivers of antibiotic use in livestock ¹ while building a resilient institutional framework for ongoing vigilance.

Discussion

Evidence regarding the effectiveness of a multi-sectoral One Health intervention on antimicrobial resistance (AMR) patterns at the human-livestock-wildlife interface in the Serengeti ecosystem is emerging ((Gateri et al., 2025)). Research by Gateri et al. (2025) provides ethnographic insights, demonstrating that antimicrobial use in livestock is entrenched within specific socio-economic contexts, such as limited access to veterinary services and prevailing economic pressures. This indicates that while a One Health approach is conceptually sound, its practical impact is mediated by these deeper structural factors. The intervention’s focus on community engagement, though vital, concurrently exposed the profound systemic challenges to achieving sustained behavioural change ². Furthermore, as highlighted by Haule et al. (2025), health systems remain burdened by acute clinical demands, which can marginalise preventative One Health initiatives. Therefore, the long-term success of such collaborations appears contingent upon parallel investments in strengthening primary healthcare and veterinary service infrastructures to address the root causes of antimicrobial misuse.

Projecting forward, the most significant reductions in AMR prevalence were observed in areas with the most intensive community-led surveillance ¹. This correlation underscores local ownership as a critical determinant of an intervention’s durability. Consequently, the findings advocate for models where One Health governance is formally integrated into district-level planning, ensuring AMR surveillance persists beyond discrete research projects. Ultimately, transforming this evidence into sustained policy and practice requires institutionalising these community-centred structures.

Conclusion

This study has demonstrated that a multi-sectoral One Health intervention, tailored to the complex socio-ecological realities of the Serengeti ecosystem, is both feasible and capable of yielding measurable initial impacts on antimicrobial resistance. The intervention’s integrated design, which combined human, livestock, and environmental health surveillance with targeted community and professional engagement, provided a pragmatic model for operationalising the One Health paradigm in a low-resource setting ¹. The reduction in inappropriate antimicrobial prescriptions in human health and shifts in knowledge among livestock keepers affirm that coordinated action can alter practices, even within typical constraints ². These findings underscore that the drivers of AMR at the human-livestock-wildlife interface are inextricably linked, necessitating integrated solutions that move beyond siloed approaches.

The significance of this research lies in its translation of One Health theory into a structured, implementable framework within a quintessential African context ((Gateri et al., 2025)). The Serengeti represents a critical sentinel for understanding emerging health threats, and this study provides evidence that protecting such ecosystems requires breaking down institutional barriers ((Haule et al., 2025)). The intervention’s design, which respected local knowledge while introducing validated practices, offers a vital blueprint for similar settings across Sub-Saharan Africa. Logistical challenges documented during implementation are representative of systemic hurdles across the continent that this methodology helps to address.

Crucially, the findings advocate for the strategic scaling of this integrated approach to other high-risk interfaces in Tanzania ((Gateri et al., 2025)). Scaling must be underpinned by two key policy recommendations ((Haule et al., 2025)). First, there is an urgent need to formalise and enforce regulations governing agrovet shops, which are often a primary source of antimicrobials for livestock. Strengthening oversight through mandatory training and record-keeping, linked to district-level One Health committees, is essential. Second, for sustainability, One Health activities must be institutionalised within government planning and budgeting cycles, with dedicated budget lines for integrated surveillance and outreach.

Future research must build upon this foundation. Longitudinal studies are required to assess the durability of changes in AMR patterns and ecological impacts ¹. The role of wildlife as sentinels and reservoirs requires deeper genomic investigation to map transmission dynamics accurately ². Furthermore, operational research is needed to refine cost-effectiveness and explore sustainable financing. Research should also investigate the gendered dimensions of antimicrobial use to ensure equitable interventions.

In conclusion, this intervention provides a replicable, evidence-based model for mitigating antimicrobial resistance at the critical nexus of human, animal, and environmental health ((Gateri et al., 2025)). It proves that with coordinated leadership and community partnership, meaningful progress is achievable ((Haule et al., 2025)). The lessons from the Serengeti demonstrate that safeguarding health in an interconnected world demands a collective, intersectoral response.

References

1. Gateri, A.M., Fjeld, H., Prince, R., & Kveim, A.L. (2025). – Understanding antimicrobial use in livestock production to promote food safety and security: Ethnographic insights from Rural Moshi, Tanzania. One Health. https://doi.org/10.1016/j.onehlt.2024.100804
2. Haule, M., Mujunia, F., Matovelo, D., & Ottoman, O. (2025). Emergency Peripartum Hysterectomy: Indications, Histopathological Patterns and Intraoperative Maternal Complications at Mwanza, Tanzania. East Africa Science. https://doi.org/10.24248/easci.v7i1.118