Study setting

This study was conducted as part of a larger research project assessing Ghana’s current ZDSR systems across three metropolitan areas (Greater Accra, Kumasi and Tamale) and their potential convergence points for optimizing intersectoral collaboration at the operational level. The broader project comprised multiple objectives including: (1) a rabies sectoral surveillance evaluation conducted in select districts in each area; (2) in-depth examination of current intersectoral collaboration mechanisms [23] and (3) identification of barriers and opportunities for operational-level One Health implementation. This paper uses the interview dataset from objective 2 of the study to identify specific factors influencing collaboration dynamics among human, domestic animal and wildlife health sectors in the Greater Accra Metropolitan Area (GAMA).

Given the complex nature of intersectoral collaboration, qualitative research methods offer deeper insights into stakeholder perspectives, institutional culture, and governance structures, uncovering aspects that quantitative approaches may overlook [24–27]. For this study, stakeholders refer to human, animal and wildlife health practitioners who play a direct role in detecting, reporting and responding to zoonotic threats. The term ‘stakeholders’ may be used interchangeably with ‘actors’ ‘participants’ or ‘professionals’ throughout this study.

Key institutions included the Ghana Health Service (GHS), the Veterinary Services Department (VSD) and the Wildlife Division (WD) of the Forestry Commission. Participants were selected from the district and subdistrict level of GHS and VSD. However, unlike GHS and VSD, the Wildlife Division did not have district-level officers, thus participants were drawn from the regional and national level. While the GHS and VSD officers are based at the district level, their surveillance and response activities span from community to district levels, making them key informants for understanding intersectoral dynamics in practice. Although site selection was determined by the broader project, the inclusion of over half of GAMA’s districts provides sufficient scope to examine intersectoral collaboration dynamics and inform the study’s conclusions.

Study design and participant selection

This qualitative study employed semi-structured in-depth interviews to explore factors influencing intersectoral collaboration in ZDSR. Although the GAMA area comprises 25 districts, only 21 had both human and animal health officers necessary for examining intersectoral collaboration across sectors.

A combined purposive and convenience sampling approach was used. Districts and district heads were purposively selected based on their authority and direct involvement in ZDSR, while convenience sampling allowed authorities to either participate directly, nominate subordinates, or involve additional team members depending on availability, provided all participants had direct ZDSR responsibility. Participants were recruited for this study from 21/12/2022 to 31/05/2023.

Following ethical approval, initial invitations were sent to all 21 eligible districts. Due to incomplete or outdated contact information, invitations successfully reached 20 human health district heads and 17 animal health district heads (one of whom oversaw two districts). Non-respondents received, a follow-up call to introduce the study, explain its purpose, seek their participation and refer them back to the email for full details.

In total, 17 human health and 17 animal health district heads confirmed participation, representing 19 districts. Ultimately, interviews were conducted in 16 districts: human health participants were interviewed in 13 districts, animal health participants in 10 districts, with only seven districts having both sectors represented. The reduction from confirmed to interviewed districts occurred due to data saturation, scheduling conflicts, participant availability and time constraints. All three wildlife health participants contacted agreed to participate and were interviewed.

A total of 66 participants were interviewed which corresponds to 41 interviews. Most participants were interviewed alone (26 interviews) but participants from the same sector and district could opt for group interviews which comprised a maximum of 4 people (15 interviews). A total of 37 interviews were conducted in person, two via zoom and two over the phone. The human health sector contributed 43 participants, including medical directors, physicians, public health officers, disease control officers, and public health nurses. The animal health sector provided 20 participants: veterinarians, para-veterinarians, one recently retired animal health practitioner (included due to the considerable institutional knowledge of this person), two laboratory veterinarians and three agriculture officers (who were invited by the veterinarians during interviews). Although laboratory veterinarians operate at the regional and/or national level, they were included due to their essential role in supporting operational-level ZDSR activities through diagnostic services and their direct collaborative relationships with operational-level practitioners. The wildlife health sector included three participants: one national level and two regional level wildlife veterinarians from the Wildlife Division of the Ghana Forestry Commission. Most participants (86%) had over five years of experience in ZDSR which ensured substantial exposure to intersectoral collaboration.

District-level participants were distributed across all 16 districts, with the majority of districts contributing between 1–4 participants per sector, ensuring balanced representation. Animal health teams were small at the district level (typically 1–4 officers per district), and in most cases, all available animal health officers were interviewed, providing near-complete sectoral representation. For human health sectors, which have larger teams, key personnel with direct ZDSR roles were purposively selected. One district’s human health sector, however, contributed a larger number of participants (n = 16) due to its involvement in the rabies surveillance evaluation of the broader research project. This evaluation required visiting all public health facilities in that district, which enabled broader sampling of ZDSR personnel beyond the typical 1–4 key stakeholders interviewed in other districts.

Data collection – Key informant interviews

The interview guide (S1 Appendix) incorporated two frameworks: (1) The Evaluation of Collaboration in a multisectoral surveillance system (ECoSur) tool which details organizational (at both governance and operational level) and functional attributes essential for effective collaboration in multisectoral systems [28]; and (2) The Components of Surveillance and Response Systems for Monitoring and Evaluation Framework of the WHO which outlines key indicators for assessing communicable disease surveillance and response systems [29].

The interview guide was designed to capture participants’ experiences with intersectoral collaboration in ZDSR, including their motivations and expectations for collaboration, as well as the barriers and facilitators influencing their ability to collaborate. Probing questions explored reasons for collaboration or its absence. While the study primarily focused on the human, animal and wildlife sectors, participants were encouraged to share collaborative experiences involving any other sector, provided it was within the context of ZDSR.

All interviews were scheduled over the phone but only after the initial email. Before each interview, written consent was obtained, and confidentiality was assured. All interviews were conducted in English and audio recorded with participants’ explicit consent. Participants were allowed to pause the recording at any time, such as when discussing sensitive information.

Data analysis

Interviews were pseudonymized, and then transcribed by professional third-party transcribers. Data were analysed using Reflexive Thematic Analysis (RTA), a qualitative technique for systematically identifying and interpreting patterns within interview data [30]. The analysis followed Braun and Clarke’s six step framework:

1. (1) Familiarisation – we ensured transcription accuracy through multiple readings of the transcripts against the original audio recordings;
2. (2) Coding – initial coding was done with MAXQDA Analytics Pro 2022 software [31];
3. (3) Generating themes – we grouped similar codes into themes;
4. (4) Reviewing themes – we refined and merged overlapping themes;
5. (5) Defining themes – final themes were identified from the refined codes and organised to ensure coherence and relevance;
6. (6) Producing the report – themes were synthesized into the study’s findings [30].

Inductive coding was used to allow themes to emerge from the data. This included both explicit responses to reasons for collaboration (or lack thereof) and more nuanced, implicit insights. A second independent re-coding was performed to identify any new or missing codes.

All themes reported in this manuscript were derived from multiple participants’ discourse; no single theme was based on one individual’s perspective. This approach ensures that the analysis reflects shared patterns across the sample rather than isolated comments.

Ethical considerations

Ethical approval for the main project which included all sub-objectives was obtained from the University of Bonn – Centre for Development Research Ethics Board (22c_Joannishka Dsani), and the Ghana Health Service Ethics Review Committee (GHS-ERC: 023/09/22). Formal permissions were granted by the Veterinary Services Directorate of the Ministry of Food and Agriculture and the Wildlife Department of the Forestry Commission of Ghana.

Participant confidentiality

To maintain transparency while protecting participant anonymity, quotes are labelled only by sector (HH = human health, AH = animal health, WI = wildlife health). Many participants shared contextually specific experiences that could reveal their identity to readers familiar with the local health system. Once identified through one distinctive quote, any identifier system would allow readers to attribute all other comments made by that same participant. This risk is particularly acute for animal health participants, as most districts have only one professional in this sector. Unique identifiers, district codes, or other demographic descriptors were therefore not used. Direct quotes from wildlife health participants were not included given their very small number (n = 3). Their perspectives are reflected in the analysis and echoed through shared themes with other sectors, but not illustrated with individual quotations. Finally, to avoid overrepresentation, human health quotes from the district with the highest participation (n = 16) were excluded from illustrative examples presented in this manuscript (though included in the analysis), while animal health quotes from this district were retained. Overall, the illustrative quotes presented in the manuscript represent participants from 15 of the 16 districts and across all organisational levels. One district is not represented in the quotes because only one sector was interviewed in that district and their verbatims were not optimal for illustrating the themes identified.

Individual factors

Structural factors

Positive collaborative outcomes.

Unlike the previous factors that fell primarily within individual or structural categories, positive collaborative outcomes from past partnerships operated across both. Participants described concrete benefits that not only encouraged continued collaboration but also created reinforcing cycles, where past success made future engagement more likely. In this way, outcomes bridged individual efforts and structural conditions, becoming predictive factors for sustained collaboration.

Improved response and control measures emerged as an important outcome of collaboration. Participants described how information shared across sectors triggered timely interventions. One human health participant explained how reporting a dog bite case led to district-wide vaccination efforts: “When we inform, I remember [a recent dog bite case]... they were able to purchase vaccines and then immunized some of the dogs” (HH). The phrase “when we inform” suggests that this process has occurred repeatedly, creating a reliable expectation that reporting will lead to action. This dependability was incorporated into participants’ decisions to collaborate.

Enhanced surveillance and early detection were other outcomes participants linked to collaboration. Relationships across sectors enabled alerts that might otherwise have been missed. One animal health participant noted: “There have been a lot of cases where, because of [our] collaboration, we’ve been prompted. The cat case that we had; it was the health people that called me... I would not have even heard of the case” (AH). The reference to “a lot of cases” indicates that collaboration provided a consistent and dependable mechanism for identifying threats early. This reliability created mutual dependencies that encouraged participants to maintain these connections.

Improved access to resources, particularly post-exposure rabies vaccines, was also identified as a tangible benefit of intersectoral collaboration. Human health participants described the animal health sector’s ability to provide vaccines as especially valuable: “Veterinary has been helpful as far as dog bite and rabies is concerned. Once we send someone there, they give the vaccine [for] free” (HH). The confidence in “once we send someone” shows that this support has become a dependable arrangement rather than a one-off occurrence. The predictability of this resource sharing provided a practical incentive to sustain collaboration. These experiences demonstrate how positive outcomes become influential factors; participants who experienced benefits were more likely to engage in future collaboration, creating self-reinforcing cycles of engagement.

All factors discussed above are illustrated in Fig 1, which provides a visual summary of the individual and structural factors that influence intersectoral collaboration in ZDSR.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Conceptual framework of factors influencing intersectoral collaboration in zoonotic disease surveillance and response in Greater Accra Metropolitan Area, Ghana.

https://doi.org/10.1371/journal.pone.0347471.g001

The Venn diagram illustrates individual contextual factors (orange circle) and structural contextual factors (blue circle) that influence intersectoral collaboration. Positive outcomes from successful collaborations (green, positioned at the intersection) operate across both dimensions, creating reinforcing cycles that sustain engagement. Green ovals represent factors that exclusively facilitate collaboration, red ovals represent barriers, and yellow ovals represent factors that can function as either facilitators or barriers depending on their presence, absence, or degree. Individual factors are driven by personal characteristics and actions, while structural factors represent systemic conditions that operate independently of individual efforts.

Stakeholder aspirations for future collaboration

While the factors discussed previously reflect the conditions and experiences shaping current collaboration, participants also expressed aspirational expectations about what intersectoral collaboration could achieve in the future. These perspectives represent hoped-for benefits rather than demonstrated influences on collaborative behavior.

Participants consistently anticipated that collaboration would improve efficiency and effectiveness by leveraging different sectoral strengths, leading to smoother coordination and enhanced disease surveillance. They expected collaboration to enable early detection and prompt action, with the ultimate goal being holistic population protection: “The ultimate goal is to protect the public, animals included. Because if you control it from the animal’s side, that is when the public [can benefit]” (AH). The COVID-19 pandemic reinforced participants’ conviction that intersectoral collaboration is essential for ZDSR. One participant expressed hope that the pandemic would catalyze greater recognition of the animal health sector’s critical role: “And sorry to say, but I’m happy that the pandemic is happening and it is all originating from animals. So now, the world will sit up and take this sector serious” (AH).

A key aspiration was leveraging intersectoral collaboration as a platform for public awareness and education. Participants believed that a coordinated, unified voice was necessary to raise the visibility of zoonotic disease risks and prevention among the general public. Beyond public education benefits, participants also viewed collaboration as a platform for learning, skill-sharing, and capacity building for themselves (the professionals), anticipating mutual learning benefits that improve future independent performance:

“The more outbreaks or surveillance work you are doing, the more you are improving your knowledge. […] you collaborate and at the end of the day, you have ideas on how to handle another[case]. Even if the [other sectors] were not there, you would handle it better [because of] prior experience.” (HH).

Resource sharing aspirations focused on overcoming sectoral limitations through systematic pooling, envisioning coordinated multi-sector responses based on past successful experiences. One participant provided a practical example of this vision in action during a crisis: “Collaboration is good because I remember… we had a bird flu [case]. We went to the Fire Service people for their fire tender to do the depopulation” (AH).

Finally, participants aspired to achieve holistic, intersectoral approaches to disease control that address environmental and policy dimensions. They envisioned comprehensive surveillance systems where various sectors coordinate to track disease risks. One participant described an elaborate vision of collaborative disease prevention driven by integrated surveillance and alerting:

“If we collaborate well, we’ll be able to reduce episodes of outbreaks. [...]Imagine if the people in the wildlife start tracing. [...] Then they can alert the assembly […] Then their environment (unit) starts looking […] [Then human] health can also start community engagement.. That way, we will prevent anyone from getting the disease” (HH)

Participants also envisioned collaboration driving policy enforcement for disease prevention. They anticipated that intersectoral engagement would enable local government bodies to enforce existing bylaws, as one participant illustrated with a hypothetical scenario: “Animal [health] can tell assembly that ‘this group of animals [e.g., stray dogs] are vectors... we shouldn’t allow them to be at residential areas.’ Then assembly [will] ensure that their bylaws are enforced, because now, they [understand] the reasons behind it” (HH)

These aspirational visions reflected participants’ understanding of ideal intersectoral collaboration, even when current structural constraints prevented their realization.

The interplay between individual agency and systemic structural barriers

In the absence of formal structures, personal networks become the primary drivers of collaboration. While interpersonal rapport facilitates rapid coordination where formal systems are weak [32,33], reliance on informal arrangements creates three interconnected barriers:

First, inequitable access to collaborative opportunities. Participation depends on who knows whom rather than relevant expertise. Systems requiring constant self-advocacy privilege assertive individuals while systematically excluding others [34,35], burdening marginalized sectors to repeatedly prove their worth and reinforcing the inequities One Health aims to address.

Second, fragility. Informal networks lack standardized procedures or institutional memory, making collaboration vulnerable to staff turnover. Even where committee reconstitution requirements exist, weak enforcement allows structures to dissolve during personnel transitions.

Third, perceived hierarchies. Participants reported perceived human health dominance likely stemming from visible disparities in workforce capacity, and resources, given that the human health sector is typically better resourced [36–38]. Without clearly defined cross-sectoral roles, actors’ default to interpreting hierarchies subjectively, creating barriers even where objective power differences are minimal.

Beyond relational dynamics, structural fragmentation creates operational barriers. Financial mistrust, with participants fearing resource exploitation, discourages collaborative initiatives [39,40]. Weak governance and limited support from district authorities fragment efforts. Event-driven rather than routine collaboration lack of follow-through on submitted plans (such as rabies control proposals) disrupt continuity and limit the institutionalization of collaboration [21,25,41].

Workforce scarcity compounds these problems: limited veterinary and wildlife health professionals create capacity issues and reduce sectoral visibility [42–44]. Human health actors were often unaware animal health counterparts existed, while others had never even considered collaborating with wildlife health. Physical inaccessibility further reduced visibility. While research suggests that physical proximity increases collaboration likelihood [45], this study observed instances where sectors shared the same office building but had never collaborated due to absent formal mandates and defined roles. This demonstrates that adequate staffing and physical presence are essential for sectoral visibility and accessibility, and while proximity may create opportunities for collaboration, these conditions remain insufficient without formal institutional frameworks to mandate and structure engagement.

Pragmatic drivers and virtuous cycles of engagement

This study challenges assumptions that One Health collaboration is driven by shared, collective goals. Engagement is often motivated by pragmatic concerns: instrumental benefits and ‘covering one’s back’ to avoid professional setbacks. When participation is determined by institutional incentives rather than proactive choice, collaboration remains opportunistic and unsustainable [12,46–49].

However, successful experiences create reinforcing cycles. Participants experiencing concrete benefits, improved outbreak response, early disease detection, enhanced resource access, were more willing to engage in future collaboration, a dynamic underexplored in One Health literature which emphasizes barriers [21,25,41]. When collaboration delivers visible value, expectations of future benefits increase.

Yet these cycles require consistent institutional support. Failed collaborative attempts, such as when sectors lacked facilities to follow through on commitments, discouraged future engagement. This creates an expectation-reality gap: participants held aspirational visions of improved efficiency and comprehensive resource-sharing that remained largely unrealized, contributing to frustration and disengagement [21,50–52]. Initial positive experiences cannot substitute for structural conditions enabling consistent benefit delivery [16,53].

Structural solutions – Optimizing institutional architecture for one health

Addressing these challenges requires modifying existing institutional structures rather than creating new One Health platforms, which often remain non-functional [54,55]. While complete resource equity is unlikely in resource-constrained contexts where human health receives greater political attention, modest structural adjustments can mitigate power imbalances.

The district Public Health Emergency Management Committee (PHEMC) represents an ideal existing platform for this approach. Established under GHS’s Integrated Disease Surveillance and Response (IDSR) guidelines [56], PHEMCs are intended as multisectoral coordinating bodies for surveillance and emergency preparedness, with mandated regular meetings during non-emergency periods. However, implementation remains inconsistent. Participants described committees dissolving after staff transitions, and many human health actors had never met their veterinary counterparts despite working in the same district for years. These gaps reflect structural ambiguity in PHEMC membership guidelines. The IDSR specifies the “District Director of Health Services” and “District Environmental Health Officer” as distinct members, yet veterinary services are designated as “District Director of Veterinary/Agricultural Services” [56], grouping animal health expertise with agricultural leadership. This conflation creates uncertainty over representation and provides no mechanism to ensure veterinary participation during committee reconstitution. Clarifying PHEMC membership to explicitly mandate “District Veterinary Officer” and enforcing regular meetings would address multiple barriers simultaneously: routine meetings would institutionalize contact (mitigating unfamiliarity); formal inclusion would remove reliance on personal relationships or self-advocacy; and planning platforms would reinforce collective ownership.

Simple administrative measures could further strengthen collaboration. Participants’ experiences varied widely depending on whether new personnel were formally introduced to other sector leads upon posting. District Assemblies should implement a standardized introduction protocol, such as circulating new personnel profiles and contact details via email to all relevant departments, thus ensuring early familiarity and reducing reliance on chance encounters.

Beyond administrative reform, institutionalizing One Health education can neutralize hierarchies. Participants noted that human health professionals with postgraduate public health training demonstrated more collaborative attitudes and greater respect for cross-sectoral expertise than those without. While public health training is officially required for district health leadership [57], some leaders reported self-educating on One Health principles, suggesting that these principles are not yet standardized. Consistently emphasizing One Health principles within preservice curricula and in-service training across all relevant health cadres could normalize collaborative attitudes as a professional standard rather than an individual disposition.

These educational efforts must be complemented by formalized communication and data sharing protocols. Rather than implementing complex technologies, routine exchange of aggregated district-level surveillance reports would maintain shared situational awareness while preserving operational control. Clear distinctions between routine updates and urgent notification triggers for notifiable diseases would proceduralize joint actions. Defining roles and response protocols for specific disease scenarios would further institutionalize this collaboration.

These suggested structural reforms require adequate human resources to function effectively. Environmental health’s “visibility” stemmed from workforce numbers, whereas veterinary services were perceived as “just one person” and easily overlooked. Without increased staffing and dedicated infrastructure, such as district veterinary clinics, even well-designed institutional reforms may struggle to achieve an operational impact.

These reforms focus on achievable improvements in coordination and communication rather than comprehensive resource equity, which remains constrained by broader budget realities. By delivering tangible benefits in information exchange, joint planning, and routine engagement, these structural modifications can bridge the expectation-reality gap that currently fuels frustration and disengagement.

Finally, these district-focused recommendations do not address wildlife health marginalization which stems from having only three wildlife veterinarians nationally with no district-level presence. Addressing this gap requires national-level workforce investment and policy intervention beyond the scope of local governance mechanisms.

Limitations and perspectives

This study has limitations. First, self-reported data may introduce bias and precludes establishing causality or measuring actual impacts. Future research could combine interviews with quantitative assessments of measurable outcomes (e.g., response times, vaccination coverage) to establish clearer links between collaboration and public health results.

Second, while we implemented measures to enhance trustworthiness, including independent re-coding of a subset of transcripts, resource constraints prevented dual coding of the entire dataset, which would have further minimized potential single-coder bias.

Third, this study prioritized in-depth insights into operational-level collaboration dynamics within GAMA over statistical generalizability. While systemic One Health barriers are documented regionally [15], evidence on how these operate within Ghana’s specific operational architecture remained limited. This study addresses that gap, offering detailed understanding of dynamics transferable to similar LMIC contexts, though their manifestations and relative importance may vary across settings. Future comparative research across multiple regions and longitudinal studies tracking collaboration over time would strengthen understanding of how these patterns operate in different contexts and evolve with policy interventions.

Finally, while this study focused on domestic governance barriers at the operational level, international funding mechanisms and global health priorities also shape national One Health implementation in ways that may create competition between national actors or prioritize external agendas over local needs. Understanding how international frameworks interact with domestic collaboration dynamics warrants future research.