Facilitators and barriers to community-based livestock abortion reporting for Rift Valley Fever surveillance in Uganda: A COM-B analysis

Abel W. Walekhwa; Andrew J.K. Conlan; Lydia N. Namakula; Brenda Nakazibwe; Salome A. Bukachi; James L.N. Wood; Lawrence Mugisha

doi:10.1371/journal.pone.0359872

Abstract

Rift Valley Fever Disease (RVFD) causes substantial economic losses via livestock abortion storms, yet community-based abortion reporting for early warning remains suboptimal in endemic settings. This study analysed the behavioural determinants of livestock abortion reporting to inform RVF surveillance interventions in Uganda. We conducted a cross-sectional study using qualitative data collection methods in Isingiro District, Uganda, employing the Capability, Opportunity, Motivation–Behaviour (COM-B) framework. We conducted 29 key informant interviews with national/district policymakers and technical officers and 17 focus group discussions involving livestock owners, abattoir operators, and local leaders. Transcripts underwent deductive thematic analysis using NVivo-12, with independent coding verification and member checking. Barriers and facilitators were ranked by frequency and thematic emphasis, and a comparative analysis was conducted across participant groups. The most frequently cited barrier was poor community knowledge about RVF, while the most powerful demotivator was the absence of post-reporting feedback. Key facilitators included mobile phone platforms, community governance structures, and fear of economic and cultural loss. Our study recommends that sustainable abortion reporting requires simultaneous intervention across capability, opportunity, and motivation domains. Priority interventions could include: integrating reporting into Uganda’s electronic surveillance platforms, co-developing local RVF names with communities, establishing reliable feedback mechanisms, addressing veterinary staffing gaps in remote areas, and implementing no-blame reporting protocols. Effective RVF surveillance depends on reinforcing existing community structures while ensuring responsive feedback that validates reporters’ contributions.

Citation: Walekhwa AW, Conlan AJ, Namakula LN, Nakazibwe B, Bukachi SA, Wood JL, et al. (2026) Facilitators and barriers to community-based livestock abortion reporting for Rift Valley Fever surveillance in Uganda: A COM-B analysis. PLoS One 21(10): e0359872. https://doi.org/10.1371/journal.pone.0359872

Editor: Sikiru Adetona Olurode, Federal University of Agriculture, NIGERIA

Received: October 20, 2025; Accepted: September 18, 2026; Published: October 8, 2026

Copyright: © 2026 Walekhwa et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All data are in the paper and/or supporting information files, which have all been included in this submission, and these have unrestricted access. These include the codebook, Ethical Approval letters, and Data collection tools (FGD Guide and KII guide).

Funding: We are grateful for funding for this fieldwork from the Biotechnology and Biological Sciences Research Council (BBSRC) Impact Acceleration Account award, the Public Engagement Starter Grant, and the Nigeria Travel Grant from the University of Cambridge Centre for African Studies. The Alborada Trust supported AJKC and JLNW. The funder provided support in the form of salaries for authors AJKC and JLNW, but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section. AWW was supported by the Cambridge Trust for his doctoral studies at the University of Cambridge, United Kingdom.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Rift Valley Fever disease (RVFD) is caused by the Rift Valley Fever (RVF) virus, transmitted mainly via mosquito vectors (Aedes and Culex spp) [1]. The most notable symptom in livestock with RVF is abortion “storms” [2]. The occurrence of abortion storms could be used to institute syndromic surveillance for RVF [3,4].

Surveillance for livestock abortions, as proposed by Thurmond and Picanso [5], hinges on effective reporting. Community-based reporting refers to a participatory surveillance approach that empowers local livestock stakeholders to recognise and report livestock abortion events in a timely and structured manner, thereby strengthening early detection, response and control of diseases like RVF [6]. This reporting, if consistently done, enables real-time or near real-time detection [3,4,7].

In our study, we sought to understand the facilitators and barriers to community-based reporting of livestock abortions for RVF surveillance in Isingiro District, Uganda. A number of short-term research studies in East Africa, including our own demonstration of feasibility in Uganda [4], have highlighted the potential of this approach. A notable research study in the arid and semi-arid lands of Kenya used self-reporting by phone to quantify the burden of livestock abortions. The leadership of Kenya Veterinary Services (KVS), the availability of phones, and the training of local animal health practitioners were the enabling factors for its success [8]. However, there is scant information on what happened after this research study concluded. This highlights a critical gap that Worsley-Tonks et al. (2025) also highlight: the need for sustainable, long-term integration of community-based surveillance into national health systems [9]. Another research study by Thomas et al. enabled the detection of ten different pathogens, with RVF identified as the leading cause of abortion [3,10,11]. While these studies have demonstrated technical feasibility and quantified the burden of abortion-related diseases, a critical gap remains in understanding the specific behavioural determinants that influence sustained community reporting. Without this understanding, interventions risk failing to address the underlying factors that shape reporting decisions. To address this, we adopted an implementation science approach using the COM-B model [12] to systematically analyse the facilitators and barriers to abortion reporting from multiple stakeholder perspectives. This study advances current knowledge by: first, providing theory-driven identification of modifiable behavioural determinants, second, comparing how barriers are understood and experienced across policymakers and community levels, and third, revealing the interconnected nature of capability, opportunity, and motivational factors that single-domain interventions fail to address. This work contributes to a growing body of knowledge advancing sustainable community-based surveillance systems. This study aimed to systematically identify and analyse the facilitators and barriers to community-based livestock abortion reporting for RVF surveillance in Isingiro District, Uganda, using the COM-B behavioural framework.

Methods

Ethics Statement

Ethical approval for this study was sought and received from Makerere University School of Public Health (Reference No: SPH-2022–364). Administrative clearance from the Isingiro District local government’s administration was also required and provided by the Chief Administrative Officer’s office (dated 20^th March 2023). All methods were performed in accordance with relevant national and local guidelines and regulations.

Respondents were taken through a consenting process and were provided with information about the study. Their voluntary participation and consent were recorded in writing. To enhance understanding, we translated all the data collection tools and informed consent forms into the local language (Runyankole), which was the most spoken language in the Isingiro District. All the data were de-identified to protect the identity of our informants and stored in an encrypted and password-protected external drive of the corresponding author. The emerging quotes were anonymised by using aggregated descriptors.

Study design

We employed a qualitative research design with deductive thematic analysis.

Data Collection Period

Data were collected between 22^nd March 2023 and 4^th July 2023.

Description of the study sites

The location of this study, Isingiro District, Uganda, is an area where previous RVF outbreaks have been reported (Bakamutumaho et al., 2025; Muema et al., 2021; Ndishimye et al., 2024; Nyakarahuka et al., 2023). The specific 17 sub-counties and town councils of Masha, Bugango TC, Kikagati, Kashumba, Kakamba, Endiizi, Endiizi TC, Mbaare, Ngarama, Ruborogota, Rushasha, Rugaga, Rugaga TC, Rwanjogyera, Rwetango, Nyakitunda and Nakivale refugee settlement were selected as they carry out livestock rearing practices associated with a high risk of RVF [13]. Most of the respondents were community representatives from the study area and Isingiro District local government workers. Isingiro District is in southwestern Uganda (0.8435° S, 30.8039° E), about 297 kilometres from the capital city, Kampala (Fig 1).

Download:

Fig 1. Map of Uganda showing Study sub-counties, Isingiro District, Uganda.

https://doi.org/10.1371/journal.pone.0359872.g001

Isingiro District is characterised by livestock rearing as a major livelihood for the majority of residents [14–16]. The district also neighbours Lake Mburo National Park, which harbours wildlife species susceptible to RVF [13]. Both wildlife and livestock share common water sources along the shores of Lake Mburo and the Kagera River, which spans the district. Isingiro District has an estimated cattle, goat, and sheep population of 368,246, 422,108, and 88,621, respectively (UBOS, 2024), and a human population of 635,077 based on the human census (UBOS, 2024).

The district experiences a tropical savanna climate with an average annual rainfall of 1,200 mm, and a temperature range of 17–30° C [17,18]. The district has two rainy seasons each year: March to April and September to November [19]. These climatic conditions play a critical role in the spread of RVF [20].

Theoretical framework of this study

We employed an implementation science (IS) approach grounded in the COM-B model to investigate the facilitators and barriers to community-based abortion reporting for RVF surveillance. IS provides a structured approach for translating evidence-based interventions into real-world settings and is especially relevant for strengthening health systems in low - and middle-income countries (LMICs) [21]. The use of a robust theoretical framework like the COM-B model is a significant strength of this study, as it allows for a systematic analysis of the factors influencing behaviour change.

The COM-B model, developed by Michie et al. [12], posits that behaviour change occurs when individuals or groups have the Capability, Opportunity, and Motivation to perform a behaviour. In the context of this study, the target behaviour was the reporting of livestock abortions by community members.

The selection of COM-B over alternative behavioural frameworks (Health Belief Model, Theory of Planned Behaviour, Social Ecological Model) was guided by several considerations. First, our preliminary fieldwork described elsewhere [4] revealed that knowledge deficits (capability) were demotivating, but more importantly, the absence of response opportunities created motivational failures that training alone could not address. COM-B captures such dynamic interconnections between capability, opportunity, and motivation. Second, COM-B connects directly to the Behaviour Change Wheel, providing a systematic pathway from behavioural analysis to intervention design – a feature absent in purely descriptive frameworks like the Health Belief Model. Third, COM-B can be applied simultaneously at individual, community, and system levels, enabling comparative analysis of how barriers are experienced differently by policymakers and communities. Fourth, COM-B has been validated in diverse health behaviour contexts and increasingly used in Low- and Middle-Income countries (LMIC) settings, demonstrating its utility in resource-constrained environments where multiple barriers must be addressed simultaneously.

Capability refers to the psychological and physical capacity to engage in a behaviour. In the context of RVF surveillance in Uganda, this includes knowledge and training of veterinary and health officers on RVF case recognition, abortion reporting systems, and sample handling procedures. It also involves awareness of surveillance and reporting channels such as the electronic Integrated Disease Surveillance and Response (eIDSR) system and the World Organisation for Animal Health – World Animal Health Information System (OIE-WAHIS). These factors determine whether frontline personnel have the skills and understanding necessary to detect and report RVF cases accurately.

Opportunity encompasses external factors that make the behaviour possible or prompt it. For RVF surveillance, these include the availability of logistics and resources (e.g., transport, sample collection kits, and laboratory networks), effective communication infrastructure (phones and radios) and coordination between human and animal health sectors under the One Health framework. Community-based surveillance systems and supportive governance structures also create opportunities for timely detection and information sharing.

Motivation involves the mental processes that energise and direct behaviour, including both intrinsic and extrinsic incentives. In the Ugandan context, this is reflected in the perceived importance of RVF surveillance, fear of economic losses due to livestock deaths, cultural obligations to protect community livelihoods, and perceptions of disease severity. Motivation can also be influenced by recognition or incentives for timely reporting, as well as by emotional factors such as fear of blame or burnout associated with high workloads.

Behaviour refers to specific actions an individual or group performs (or does not perform) that interventions aim to change. In our case, this was abortion reporting by community-level stakeholders.

The strength of the COM-B model lies in its ability to illustrate how these components are interdependent and dynamic [22,23]. This implies that interventions targeting one domain (such as training to improve capability) must also ensure that opportunities for action and motivational reinforcements (such as feedback and visible responses) are present to sustain behavioural change. This integrative understanding is particularly crucial in the surveillance of zoonotic diseases, where both human and animal health systems interact. This resonates with the core message of Worsley-Tonks et al. (2025), who argue that successful community-based surveillance requires a holistic approach that addresses not only the capacity of community members but also the broader systemic and structural factors that enable or hinder their participation [9].

Intervention (Training)

Before qualitative data collection for this study, a community sensitisation training programme was conducted in Isingiro District as part of the broader abortion surveillance pilot study [4]. This training is not the “intervention” being evaluated in this study; rather, it established a contextual foundation that informed participants’ awareness and experiences, which were subsequently explored through the qualitative inquiry.

The office of the District Veterinary Officer (DVO), responsible for animal health surveillance at the sub-national level, co-developed the training material with the corresponding author. The developed instructional material included the local authority DVO’s telephone number, which farmers could contact in case of abortions at their farms. Local leaders distributed these targeted messages during community funerals and barazas (public community meetings where residents gather to discuss local issues), intending to encourage the reporting of abortions. The instructional materials used in this training are provided (S1 Appendix).

The purpose of describing this training here is threefold: First, to provide transparency about the implementation context in which the qualitative data were collected. Second, to acknowledge that some participants in the KIIs and FGDs could have been exposed to this sensitisation through neighbouring sub-counties/locations, which may have influenced their knowledge and attitudes, and third, to situate the facilitators and barriers identified in this study within a real-world setting where initial capacity-building efforts had already been undertaken in a neighbouring sub-counties. This is especially important because the farmers/stakeholders interact in social gathering events, and they tend to discuss livestock-related interventions by the government or different stakeholders like ours. Although efforts were made to first interview the stakeholders before being trained, we cannot fully rule out the possibility of contamination from neighbouring sub-counties.

A one-day training intervention was developed and delivered to stakeholders across multiple sub-counties, followed by three weeks of reinforced messaging. The training was designed to build practical skills and knowledge for syndromic surveillance. The core curriculum covered: (a) the clinical recognition of livestock abortions and their importance as a sentinel event for zoonotic diseases such as RVF. (b) Signs and symptoms of RVF in both animals and humans, according to national surveillance guidelines. (c) The specific action to take upon identifying an abortion: calling a dedicated telephone number for the local authority DVO’s office, which was printed on the instructional materials. (d) The purpose and benefits of a community surveillance system. We conducted in-person training sessions in English, with simultaneous translation into the local language (Runyankole) provided by a native technical officer to ensure comprehension. This training was conducted among 900 stakeholders, including seven district officials, 72 sub-county technical officers, 38 community health workers, 788 livestock owners, abattoir staff, and local council leaders (S1 Table). Qualitative data collection for this study commenced approximately three to four days after the conclusion of this sensitisation campaign in the neighbouring sub-counties.

Data collection

Participants and sampling.

Those who accepted the invitation were invited to participate in the study following the training intervention described in section 2.4.1.

We acknowledge potential selection bias arising from our recruitment strategy. Using community leaders (gatekeepers) to identify participants may have preferentially included individuals more connected to community governance structures and excluded those marginalised from or distrustful of these systems. This could have resulted in overrepresentation of individuals with greater social capital, potentially influencing our findings in two ways: first, by potentially underestimating the role of LC1 meetings as reporting pathways, which may not be accessible or trusted by all community members. Additionally, the requirement to attend one-day training before recruitment may have excluded individuals with time constraints, caregiving responsibilities, or limited mobility, potentially biasing the sample toward those with greater socioeconomic resources and fewer competing demands. However, our use of multiple gatekeepers across the 17 sub-counties, purposive sampling to ensure diversity in gender, occupation, age, economic status, and geographical distribution, and the 100% participation rate among those invited partially mitigated this risk. This consistency of our findings with regional literature [3] suggests that the identified barriers are likely systematic rather than artefacts of selection bias, though we acknowledge this limitation in interpreting the generalisability of our findings.

Procedure of focus group discussions

A total of 17 FGDs (from 17 different sub-counties) were conducted. We chose the FGDs to enable understanding of group-level factors for the intervention and to create an interactive environment that would allow for further probing (Setia, 2017). We considered respondents of similar socio-economic status in the community so that they could freely engage in the discussions. The choice of respondents was purposeful, with stakeholders selected given their active roles in human and animal surveillance efforts in Uganda. The interviews and discussions were held in various sites that were convenient to the respondents. All FGDs were facilitated by a leader, who was selected at the start of each session. The sessions were held in the local language (Runyankole) and moderated by a research assistant who was a native speaker. Each FGD session lasted approximately 60 minutes (range: 52–74 minutes), providing adequate time for all participants to share their perspectives while maintaining engagement throughout the discussion. The use of FGDs is a key strength of this study, as it allows for the exploration of shared experiences and social norms related to abortion reporting [24]. This aligns with the emphasis in Worsley-Tonks et al. (2025) on the importance of understanding the cultural and social context of community-based surveillance [9].

Procedure of key informant interviews

For KIIs, 29 policymakers and senior technical officials from district (sub-national) and national levels, and from both public and private agencies (ministries, academia, Food and Agriculture Organisation) were enrolled. Some of the respondents included the incident commander for disease outbreaks; the director of the National Institute of Public Health – Ministry of Health; ministry technical staff, animal health surveillance officers, laboratory technicians, environmental health officers, senior veterinary officers and field epidemiologists. The use of KIIs aimed to explore high-level expert opinions on this subject, providing strategic and contextual explanations for facilitators and barriers to community-based reporting, as recommended by Tumusiime et al. [25]. For the KIIs, apart from four informants who requested a virtual interview, the rest of the interviews were conducted physically in locations convenient to respondents. As part of the participant recruitment process, each participant was taken through a consent process (using a prepared consent form), and only those who gave written informed consent were included in the study. KIIs lasted an average of 47 minutes (range: 32–68 minutes). The four virtual interviews were comparable in duration to in-person interviews (mean: 44 minutes vs 48 minutes, respectively). No participant was interviewed more than once. Each key informant participated in a single interview, and each focus group participant attended only one FGD session. Repeat interviews were not conducted as the study aimed to capture perspectives at a single time point following the training intervention, and saturation was achieved within the initial round of data collection.

Data saturation was assessed through an iterative process conducted alongside data collection. After each KII and FGD, the research team (AWW, LNN and the FGD facilitator) convened for debriefing sessions where field notes were reviewed and emerging themes were discussed. The team used a saturation grid to track the emergence of new codes across successive interviews. Saturation was considered achieved when three consecutive interviews (for KIIs) or two consecutive FGDs yielded no new substantive codes relevant to the COM-B framework. This was reached at the 29^th KII and 17^th FGD. To validate this, we conducted an additional KII (30^th) and FGD (18^th) as a confirmatory check; both confirmed that no new information emerged, as the same themes continued to be discussed without novel content. This systematic approach aligns with the established guidance on saturation assessment in qualitative research [26]. All individuals invited to participate in the study accepted the invitation, resulting in a 100% participation rate with no refusals.

For the KIIs, all 29 invited policymakers and technical officers consented to participate. For the FGDs, all 17 scheduled focus group discussions proceeded with full attendance from invited participants. This high acceptance rate can be attributed to the prior training intervention, which established rapport and demonstrated the study’s value to the community, as well as the use of respected gatekeepers (community leaders) who facilitated trust between researchers and potential participants. No participants withdrew from the study after providing consent.

Interviewer characteristics and team composition

The data collection team comprised five interviewers: two males and three females. The gender-balanced team was intentionally composed to facilitate comfortable discussions with diverse participant groups, as community norms in the study area sometimes influence disclosure based on interviewer-participant gender dynamics. The team included:

The corresponding author (AWW) – a male senior epidemiologist who led KIIs with policymakers and technical officers.
A female research assistant (Mellon Ainembabazi) – a native Runyankole speaker who also served as a female youth local council leader for Bugango Town Council- listened and guided the moderators of all FGDs.
Two additional female research assistants supported notetaking and logistics during FGDs.
One male research assistant supported district-level KIIs and technical coordination.

The gender composition proved advantageous: female participants in FGDs appeared more forthcoming when discussing sensitive topics (such as cultural practices around livestock and dowry) with the female moderator, while male policymakers engaged readily with the male corresponding author during KIIs. All interviewers received standardised training on qualitative data collection techniques, maintaining consistency in approach while leveraging gender-appropriate facilitation.

Data processing and quality assurance

During the entire data collection process, all interviews were recorded. The audio recordings were later transcribed verbatim by two trained, experienced research assistants. For quality assurance in audio transcriptions, the corresponding author sampled six audio files (three from each research assistant) and listened to the audio in comparison to the transcript, which was found to be consistent. Beyond this sampling, we conducted multiple rounds of validation to ensure credibility. We cross-referenced each theme with verbatim quotations to ensure it accurately represented the respondents’ perspectives. Multiple researchers reviewed themes independently to eliminate potential biases and inconsistencies. This refinement ensured that the final themes were precise and accurately reflected the data. The transcripts were then analysed using deductive thematic analysis with NVivo 12 software (Edhlund and McDougall, 2017) guided by the COM-B model.

To enhance the credibility and trustworthiness of findings, we conducted member checking with a subset of participants following initial analysis. This process involved:

Presentation of preliminary findings at three intervals. First, on the last day of data collection, findings were presented to 23 Isingiro district-level officials as part of the exit from fieldwork. Secondly, six months later, findings were presented to 18 district-level stakeholders and sub-county veterinary staff, including six livestock owners. Finally, findings were presented to 13 staff of the National Animal Diseases Diagnostics and Epidemiology Centre (NADDEC) at the Ministry of Agriculture, Animal Industries and Fisheries.

The first two sessions were conducted in both Runyankole and English, facilitated by the district veterinary officer. The national-level sessions with the Ministry of Agriculture were facilitated by the male facilitator (the corresponding author).

Validation process

During these meetings, we presented the key themes organised by the COM-B framework (facilitators and barriers in each domain) using simple language and illustrative quotes (anonymised). Participants were encouraged to:

Confirm whether the themes accurately reflected their experiences.
Challenge any interpretations they felt were incorrect or incomplete.
Provide additional context or examples to enrich the findings.
Prioritise which barriers they considered most critical.

Outcome of member checking

Participants unanimously confirmed that the identified themes captured their experiences accurately. Importantly, they strongly emphasised that the absence of feedback mechanisms was the single most demotivating factor – a perspective that validated our analytical emphasis on this theme. Several livestock owners added nuanced examples of how delayed responses affected their reporting decisions, which were incorporated into the final analysis. No participant contested any themes, and there was consensus that the facilitators and barriers presented reflected community realities.

Independent validation by research assistants

Two research assistants who were involved in data collection but not in the coding team (the female FGD facilitator and male note taker) independently reviewed the final thematic framework. They confirmed that the themes aligned with their field observations and recalled specific discussions that supported each major finding. This external verification provided an additional layer of credibility, as these individuals had direct exposure to participant narratives during data collection but were uninvolved in the analytical decisions that could introduce bias.

Reflexivity and researcher positionality

This stage involved a researcher-led thematic analysis informed by the Braun and Clarke [27] framework. We conducted a primarily deductive thematic analysis guided by the COM-B framework, while remaining open to inductive sub-themes that emerged from the data. The process followed a structured, iterative approach:

a) Familiarisation and Initial Coding: We immersed ourselves in the data by repeatedly reading all the transcripts while listening to the audio recordings. The initial coding was conducted using both deductive and inductive approaches. The COM-B model provided the primary deductive framework, with Capability, Opportunity, and Motivation serving as parent codes. Within this framework, we remained open to emergent codes that captured nuanced aspects of each domain. For example, while the deductive framework anticipated “knowledge” as a capability factor, inductive coding revealed specific sub-themes such as the absence of local RVF terminology and the confusion of RVF with Ebola insights that emerged from participant narratives rather than being predetermined by the framework. This dual approach allowed us to maintain theoretical grounding while remaining responsive to unanticipated findings [28].
b) Systematic Code Development: Using NVivo-12 software [29], we developed a structured coding framework. The transcripts were systematically coded by identifying meaning units (text segments expressing a complete thought or concept) and assigning the appropriate codes. For example, text segments discussing “knowledge of abortion causes” were coded under Capability → Psychological Capacity, while segments about “mobile phone access” were coded under Opportunity → Physical Environment. The systematic coding process generated a comprehensive codebook containing 99 distinct codes with 514 total references across all transcripts, demonstrating the depth and breadth of the qualitative inquiry (S4 Appendix for the complete codebook).
c) Theme Development and Refinement: The coded data were collated and analysed for patterns within and between COM-B domains. The preliminary themes were developed by grouping related codes and examining their relationships. This process involved iterative refinement through team discussions with Lydia N. Namakula and Brenda Nakazibwe, who independently reviewed a subset of transcripts (n = 8, 27.6% of the total) to challenge initial interpretations and enhance analytical rigour.
d) Theme Review and Validation: The preliminary thematic framework was reviewed for internal homogeneity (whether data within themes cohered meaningfully) and external heterogeneity (whether themes were distinct from each other). This involved cross-checking themes against the original transcripts to ensure that they accurately represented participants’ meanings. Discrepancies in interpretation were resolved through consensus discussions.
e) Theme Definition and Narration: The final themes were clearly defined and named to capture their essence. A codebook (S4 Appendix) was developed, documenting each theme’s definition, characteristics, representative quotes, and frequency of references. This codebook guided the final analysis and writing of the result narrative (Table 1).

Download:

Table 1. Distribution of codes and references across COM-B Domains.

https://doi.org/10.1371/journal.pone.0359872.t001

To ensure analytical rigour, multiple quality assurance measures were implemented. For transcription quality, the corresponding author used a systematic validation approach, randomly selecting and thoroughly reviewing six transcripts (13% of the total) by comparing audio recordings against transcripts. This exceeds the minimum 10% validation rate recommended for qualitative research [30]. Independent coding verification by two analysts (co-authors) achieved 87% inter-rater agreement on primary codes, with discrepancies resolved through consensus discussion, thereby enhancing the trustworthiness of the thematic analysis.

To address potential bias arising from researcher characteristics and prior experiences, we maintained reflexive journals throughout data collection and analysis and engaged in systematic positionality reflection. The following researcher characteristics are relevant to interpreting this study:

a) Corresponding Author (AWW): Abel W. Walekhwa is an epidemiologist with 12 years of field experience in Uganda’s livestock and health systems. His professional roles include serving as a member of the High-Level Expert panel on One Health with four UN agencies (FAO, UNEP, WHO, WOAH), which shapes the perspective toward recognising systemic barriers and policy-level solutions. He previously worked on disease outbreak responses for eight years, including in Sheema District – a neighbouring district to the study area – providing him with a grounded understanding of the health systems, cultural norms, and operational challenges in southwestern Uganda. In addition, he has over 17 years of experience with livestock rearing (grazing cattle during upbringing). This insider knowledge gave him credibility with participants and enabled an understanding of the practical realities of livestock management, economic pressures facing farmers, and the cultural significance of cattle (including dowry practices). However, this background also risked bias: It was possible to overemphasise technical barriers (diagnostic capacity, transport) based on his epidemiology training or make assumptions about farmer experiences based on his own background rather than letting participants’ narratives guide interpretation.

Mitigation strategies

Independent coding verification: Two co-authors (Lydia N. Namakula and Brenda Nakazibwe), who were not involved in data collection, independently reviewed themes and coding decisions. Brenda’s social sciences perspectives challenged Abel’s technical biases – for example, she noted early in analysis that he was undervaluing motivational and cultural factors, which prompted re-examination of the data and ultimately led to foregrounding feedback mechanisms and economic sensitivity as primary barriers.

Social Sciences mentorship: Prof. Salome A. Bukachi (anthropologist) provided technical mentorship throughout analysis, consistently drawing attention to cultural dimensions that Abel had minimised.

Reflexive journaling: The corresponding author documented after each interview and during the analysis phases how his positionality might shape interpretations. For instance, he noted after early KIIs with policymakers that he was empathising strongly with their resource constraint narratives; journaling helped the corresponding author to consciously balance these perspectives with community-level views from FGDs.

Participant validation: Member checking ensured that the interpretation resonated with participants themselves, providing an external check on researcher bias.

b) Female FGD Facilitator (Mellon Ainembabazi): As a native Runyankole speaker and community leader, she brought cultural fluency that facilitated open discussions. Her gender was advantageous for discussing sensitive topics with female participants. She maintained her own reflexive notes, documenting instances where participants’ body language or off-record comments provided context for interpreting verbal responses. Her independence from the coding team meant she could provide candid feedback on whether written interpretations matched her recall of discussion tone and emphasis.
c) Research Assistants (Validation Role): The two research assistants who reviewed final themes (female facilitator and male note-taker) had extensive field experience in the region but were not involved in coding decisions [31,32]. This separation of roles meant they served as validators – confirming that themes reflected field realities without being invested in analytical choices.

The combination of insider perspectives (AWW’s livestock background, FGD facilitator’s cultural fluency) and outsider analytical distance (independent coders, social sciences mentorship) created productive tension that strengthened analytical rigour. We explicitly discussed during team meetings how our positionalities might shape interpretation, and we actively sought disconfirming evidence, cases where participants contradicted emerging themes. This systematic reflexivity ensured that findings remained grounded in participants’ accounts rather than the researcher’s assumptions.

Reporting of findings

The findings are reported following the Consolidated Criteria for Reporting Qualitative Research (COREQ) guidelines [33] to ensure transparency and completeness. To maintain participant anonymity while preserving the analytical value of distinguishing between respondent categories, a standard coding scheme was developed and applied to all quotations presented in this manuscript (Table 2).

Download:

Table 2. Participant coding Scheme: KII denotes key informant interviews, FGD denotes focus group discussions.

https://doi.org/10.1371/journal.pone.0359872.t002

Results were organised according to the COM-B framework, with facilitators and barriers presented separately for each domain (Capability, Opportunity, Motivation). Representative quotes are provided to illustrate key themes, with participant codes shown above to indicate the data source while maintaining anonymity.

Results

Comparative perspectives across participant groups

Analysis revealed both convergence and divergence between key informants (policymakers/technical officials) and the focus group participants’ (community members and local leaders) perspectives. Both groups unanimously identified poor community knowledge about RVF as the most fundamental barrier to surveillance (55 references), with KII noting that “all confirmed RVF cases were submitted as Ebola suspects” and FGD participants admitting “we don’t know the disease”. Similarly, both groups highlighted severe understaffing of veterinary personnel, with FGD participants reporting waiting “five months without seeing the doctor” and KIIs confirming one officer covering multiple sub-counties (S5 Appendix).

However, notable differences emerged in how barriers were conceptualised. KIIs emphasised systemic and structural challenges, including inter-ministerial coordination failures, One Health platform weaknesses, and national policy inertia. For example, KIIs discussed “the most challenging part where we have not strengthened is the One Health approach”, a concept absent from FGD discourse. Conversely, FGD participants articulated barriers through lived experiences – the daily reality of inaccessible veterinarians, impassable roads during rainy seasons, and the economic desperation that drives self-medication and high-risk slaughter practices.

The absence of feedback mechanisms illustrated this divergence poignantly: KIIs discussed how unanswered cases “demotivate technical people” while FGD participants repeatedly stated, “we report, we never hear what happened”, expressing personal disillusionment rather than systematic analysis. Economic concerns similarly diverged – KIIs framed RVF as a threat to “international trade” and the national economy, whereas FGD participants focused on household-level impacts: “animals are used for paying dowry, so that means even people can’t marry”.

These complementary perspectives suggest that effective interventions must address both systemic barriers (identified by KIIs) and experiential barriers (identified by FGDs) simultaneously. Community-level sensitisation cannot succeed without addressing the transport and staffing gaps that leave reports unanswered; conversely, laboratory capacity building cannot compensate for the linguistic barrier of RVF having no local name, which prevents the disease from entering community discourse entirely (Table 3).

Download:

Table 3. Facilitators for community-based abortion reporting, organised by the COM-B model.

https://doi.org/10.1371/journal.pone.0359872.t003

Facilitators

a) Capability Facilitators.

Psychological Capability Facilitators
1. Knowledge of Standardised Case Definitions (IDSR): The existence of standardised case definitions, particularly those described in the Integrated Disease Surveillance and Response (IDSR) guidelines, was reported as a foundational element for surveillance. A policymaker noted that RVF surveillance in humans used “the standardised case definition that is always applied” (KII-PM–8). This indicates that formal and consistent criteria for identifying and classifying disease events are established at a national level. These definitions provide a common language and framework for health personnel involved in surveillance activities.
2. Knowledge of RVF Clinical Signs in Animals: Knowledge about the clinical signs of RVF in animals was frequently highlighted as a key facilitator, with 28 references in the codebook. Participants indicated that farmers who had received training were better able to identify abortions as a potential sign of RVF. This awareness of specific clinical signs, such as abortions and haemorrhagic symptoms, was reported to enable more accurate and timely reporting. This finding suggests that targeted education on disease recognition can significantly improve community-level surveillance capabilities.
3. Knowledge of RVF as a Zoonotic Disease: Understanding that RVF transmits from animals to humans emerged as a key psychological capability (29 references). Participants demonstrated awareness that “this disease attacks both human beings and animals, but mostly people get it when they milk infected cows and drink unboiled milk or eat half-cooked meat” (FGD-LO-7, Rugaga). This knowledge creates a cognitive link between animal health and personal health, motivating reporting behaviour.
Physical Capability Facilitators
1. Presence of Trained Veterinary and Health Personnel: The presence of trained veterinary staff and health personnel, equipped with technical skills, was found to be a crucial asset. These individuals have the expertise required to understand and apply standardised case definitions, conduct investigations, and manage disease outbreaks. Their training enables them to perform specific tasks within the surveillance system, contributing to the overall capacity for disease detection and response.
2. Diagnostic Competence at National Laboratories (UVRI, CPHL): National reference laboratories, such as the Uganda Virus Research Institute (UVRI) and the Central Public Health Laboratories (CPHL), were identified as possessing significant diagnostic capabilities. These facilities are equipped to perform advanced laboratory tests, including those necessary for confirming RVF. A policymaker mentioned, “an investigation form that is centrally located in the central public health laboratory and UVRI” (KII-PM-8), indicating that these laboratories serve as central hubs for sample analysis and confirmation. This centralised infrastructure provides the authoritative confirmation required for formal responses to disease reports.
3. Analytical Capacity of Emergency Operations Centre (EOC) Epidemiologists: The Emergency Operations Centre (EOC) was described as a central coordination hub with epidemiologists possessing analytical capacity. This capacity involves the ability to process incoming information, verify reports, and coordinate responses. A member of the Ministry of Health technical staff stated: “...when we receive an alert in the emergency operations centre, then we verify the information” (KII-MOH-3). This indicates that the EOC has the personnel and processes in place to analyse surveillance data and trigger appropriate actions.
4. Presence of District Task Forces and Rapid Response Teams: The existence of district-level coordination mechanisms was noted as a critical physical capability (5 references). A district official explained: Isingiro district has an existing district task force which, in times of outbreak like this, sits every week; it has an active district response team… the response capacity to survey is there” (KII-DTO-3). Another KII participant added: “we have fellows on the National Rapid Response Team that are actually capable of responding within 24 to 48 hours” (KII-MOH-13).
5. Innovation Capacity for Diagnostic Development: Uganda was found to have innovation platforms in the country dedicated to diagnostic design and development. A national technical officer highlighted this, stating: “We have an open platform where we can design a diagnostic using PCR. We also have a sequencing platform where we can sequence and be able to tell which genotypes are circulating” (KII-NTO-4). This indicates a sophisticated capability to develop and adapt diagnostic tools, including molecular methods such as PCR and sequencing, which allow the identification of specific pathogen genotypes. This capacity extends beyond routine diagnostics to include the ability to respond to emerging threats and evolving viral strains. This facilitates reporting behaviour.
6. Community Structures (CHWs, Local Leaders, District Rapid Response Teams): The presence and active involvement of various community structures were identified as significant facilitators. These included Community Health Workers (CHWs), local council leaders (e.g., LC1 chairpersons), and District Rapid Response teams. These individuals and groups were observed to use existing communication platforms to bridge information gaps between local communities and higher authorities. A policymaker noted the role of CHWs as “foot soldiers put in the villages who could actually be given some information which they can easily pass on to the community to be able to look out for these signs” (KII-PM– 8). Similarly, a local council leader described the reporting pathway: “When there is a disease, we report to the chairperson of the LC 1 [Local Council], and they take it to the town council to the veterinary doctor” (FGD-LCL-17). These structures integrate surveillance within established community networks, providing accessible pathways for reporting.
7. Decentralised Surveillance System: The surveillance system was described as having a decentralised structure, which was identified as a key facilitator. This structure includes focal points and task forces at the district-level. A National Technical Officer explained, “We have a surveillance team led by a focal person... it is the task force that supports them” (KII-NTO-5). This decentralisation places surveillance coordination closer to the community level, theoretically shortening the communication and response chain between communities and decision-makers.

b) Opportunity Facilitators.

Physical Opportunity Facilitators
1. Mobile Phones and Communication Platforms: Relevant to both Capability and Opportunity. Communication platforms provide the physical infrastructure enabling reporting. A district official noted: “We use radio talk shows to pass on the message to those in far distances who need to know” (KII-DTO-18). This technological presence created physical pathways for information flow to new places or people.
2. Presence of Laboratory Hub System: The existence of a functional laboratory hub system has been frequently cited (12 references) as a significant physical opportunity. This system includes a cold chain (7 references) for sample preservation and transport, and access to reference laboratories for advanced analysis (8 references). These infrastructures provide the necessary physical resources for diagnostic confirmation and support the overall surveillance efforts. Furthermore, the presence of research infrastructure (6 references) was noted as an opportunity to advance diagnostic capabilities.
3. Emergency Operations Centre as Institutional Resource: Beyond its analytical capacity, the EOC functions as an institutional resource that coordinates multi-sectoral response. A KII participant noted: “I will credit the way we set up our surveillance systems, which is quite good now because when we receive an alert in the EOC, then we verify the information” (KII-MOH-18).
4. M-Trac Platform and Electronic Reporting: The existing mobile-based reporting system for human health presents an opportunity for integration. A KII participant from FAO noted: “ We have been having the event mobile application reporting, that is M-Trac… Officers can now send in immediate mobile reports, whenever there is a problem; they can send a signal and reports, and the country can be alerted to where the problem is” (KII-NTO-5).
Social Opportunity Facilitators
1. Existing Contact Tracing for other Endemic Diseases (e.g., Tuberculosis): The established system for contact tracing, particularly for diseases such as tuberculosis (TB), was identified as a pre-existing opportunity that could be leveraged. A health technical staff member stated that “There are also some surveillance activities that are performed, and we could integrate the RVF, for example, we have what we call active case search” KII-DTO-3). This indicates that community members and health workers are already familiar with procedures for disease investigation and reporting, and platforms like M-Trac are in use for human health events.
2. Existing Indicator and Event-Based Surveillance Systems under the Ministry of Health: The Ministry of Health operates a dual system of indicator-based and event-based surveillance, which was noted as a robust opportunity. A member of the Ministry of Health technical staff explained: “We have a robust system now in the country; we do both indicator-based surveillance and then event-based surveillance... alerts... come to the EOC, and then eventually we respond quite early” (KII-MOH-3). This signifies a mature surveillance infrastructure within the human health sector, capable of detecting and responding to health alerts.
3. Community Governance Mechanisms (LC1 Meetings, Community Dialogues): Regular community gatherings, such as monthly meetings of the local council (LC1) and broader community dialogues, were identified as existing social opportunities. A livestock owner reported that, “…we meet every month and discuss and exchange knowledge. We meet there with the veterinarians and report” (FGD-LO-2). These forums provide routine and collective settings for information sharing and interaction between community members and local authorities, including veterinary personnel. Community sensitisation events (3 references) were also highlighted as a valuable opportunity to raise awareness.
4. RVF Being a Priority Zoonotic Disease: The formal prioritisation of RVF as a zoonotic disease in Uganda was recognised as a significant policy opportunity. A member of the Ministry of Health technical staff stated: “…we supported joint risk assessment in 2019, and the country prioritised RVF among the three diseases...” KII-NTO-28). This prioritisation indicates that RVF is officially recognised as a legitimate concern at a high policy level, theoretically attracting political will, resource allocation, and inter-sectoral attention.
5. One Health Coordination and NGO Support: Improved One Health coordination (8 references) was identified as a social opportunity, particularly at the sub-national level, allowing for better integration of human and animal health efforts.

Support from non-government organisations (6 references) was noted as a valuable resource. A district official acknowledged: “Capacity building, trainings they were once supported by the Ministry of Health under Uganda Virus Research Institute, we had trainings from OXFAM, yes, trainings in the refugee settlements have been supported by Medical Teams International and Uganda Red Cross Society” (FGD-LCO-12). Easy communication between Village Health Teams (VHTs) and Health Officers (HOs) (3 references) was also highlighted as a facilitator of information flow.

c) Motivation Facilitators.

Reflective Motivation Facilitators
1. Experience from Past Outbreaks: Memory and experience from previous RVF outbreaks were frequently cited (10 references) as a powerful motivator for reporting. A KII participant noted: “Preparedness Uganda I think has done so well in generally outbreaks since Ebola, which was recently contained in a short time… we have built a lot of experience in outbreaks, especially on the human side” (KII-MOH-19). A district official added: “When it comes to a human being, because you know, having had a background of COVID-19 and recently we had this Ebola alert, we strengthened the systems in the communities, so anything suspicious now the community responds very fast” (KII-DTO-25). This collective memory of disease events contributes to increased awareness and willingness to participate in surveillance activities.
2. Perceived Severity of RVF: The perceived severity of RVF, particularly its haemorrhagic manifestations, was identified as a motivator. A MoH Technical Officer noted, “...The issue is the seriousness of the disease because it presents in unusual ways that people get scared, and once they get scared, they report...” (KII-MOH-5). This indicates that the visible and alarming symptoms of RVF, which can be frightening, prompt individuals to report cases.
3. Desire to Access Veterinary Professionals: The accessibility of veterinary professionals was reported to be a key motivator for the reporting. A livestock owner stated that: “Also, the availability of vets is one that we can report to and get help from... they are easily accessible” (FGD-LO-1). This suggests that the perceived ease of reaching expert help and the physical proximity of veterinary services directly encourage reporting behaviour, as farmers anticipate receiving assistance and solutions for their animals.
4. Trust in Government Response Systems: Trust in the government’s ability to respond effectively to disease outbreaks was identified as a motivational factor. When community members believe that reporting will lead to a timely and appropriate response from authorities, their willingness to engage with the surveillance system increases.
5. Sense of Civic Duty to Protect Community Health: It was observed that the sense of civic duty to protect community health contributes to the motivation to report. This suggests that some individuals are driven by a broader responsibility to safeguard the well-being of their community, extending beyond personal economic concerns.
Automatic Motivation Facilitators
1. Fear of Economic and Cultural Loss: Fear of economic and cultural losses was found to be a powerful motivational driving force for reporting. Livestock owners expressed concern about “significant economic losses due to death and abortion” (FGD-LO-13). This economic anxiety was compounded by cultural considerations, as an official explained: “Animals are used... to pay the dowry... which means that... people cannot marry... because there is a quarantine” (KII-MOH-7). This indicates that the potential loss of livestock impacts not only financial well-being but also deeply embedded social and cultural practices.
2. Fear of Cultural Disruption: Closely related to economic loss, the fear of cultural disruption emerged as a distinct automatic motivator. Participants recognised that livestock loss affects social standing, marriage arrangements, and community relationships – creating powerful emotional responses that bypass reflective deliberation and prompt immediate action (Table 4).

Download:

Table 4. Barriers to community-based abortion reporting, organised by the COM-B model.

https://doi.org/10.1371/journal.pone.0359872.t004

Barriers

a) Capability Barriers.

Psychological Capability Barriers
1. i) Poor Knowledge About RVF: The most frequently cited barrier to community reporting, with 55 references, was poor knowledge about RVF among community members. This fundamental lack of understanding significantly impedes the ability to recognise and report potential cases. The participants noted that farmers often studied RVF the last time in school and were only reminded during outbreaks. A key informant from academia stated:
  “What stops us to curb down the speed of this disease is limited knowledge about the disease” (FGD-LO-5). Another academic key informant further elaborated on the community-level capacity: “The problem I see from the district downwards, which we call the community-level; capacity is low. So, in other words, we have capacity upwards, but we don’t have capacity downwards because the farmers do not know the disease, they can’t report it.” (KII-ACAD-1).
  This lack of recognition was so profound that “All the RVF cases that we have confirmed have been submitted as suspects of Ebola, not Rift Valley fever.” (KII-ACAD-2). This indicates that communities cannot report what they do not recognise, leading to misidentification of diseases.
2. ii) Confusion of RVF with Other Diseases: A related and frequently mentioned barrier (14 references) was the confusion of RVF with other animal and human diseases. Farmers often struggled to differentiate RVF symptoms from those of other common animal diseases. Similarly, health workers sometimes confuse RVF symptoms in humans with malaria. This confusion was exacerbated by the absence of a local name for RVF, as highlighted by a key FAO informant. “Most people don’t know; the disease doesn’t have a local name. When you say RVF, it sounds new, and it can be confused with fever or malaria. So, there is still confusion” (KII-FAO-1). This linguistic and diagnostic ambiguity contributes to underreporting and misreporting.
3. iii) Lack of Skills to Distinguish Normal from Abnormal Abortions: Participants expressed difficulty in differentiating between normal, sporadic abortions and those that might indicate a disease outbreak. FGD participants repeatedly stated, “We don’t know if every abortion needs to be reported” (FGD-LCL-3). This suggests a gap in practical skills for discerning reportable events from routine occurrences.
4. iv) Insufficient Training on Reporting Procedures: Community members and some local-level personnel reported insufficient training on the specific procedures to report livestock abortions. This includes a lack of clarity on who to report to, what information to provide, and the expected follow-up process. Poor reporting mechanisms (5 references) were also noted as a barrier.

Physical Capability Barriers
1. i) Few Staff, Big District: The scarcity of veterinary and surveillance personnel, particularly in relation to the large geographical areas they cover, was a significant barrier to physical capability (18 references). Participants frequently highlighted that veterinary doctors often cover four or more sub-counties, leading to long response times. A participant in a Kitagati FGD explained: “A veterinary officer looks after like 4 sub-counties. For example, like now here, I can spend like 5 months without seeing the doctor. Now that person, when will you report to him that cows are sick?” (FGD-LO-2). The issue was compounded by the lack of replacements for retired personnel. “We have this disease, but we do not have a doctor in our sub-county of Kashumba. We need a doctor to come and educate us. That is a major challenge that we have. The doctor who was there retired.” (FGD-LO-5). This shortage of personnel directly affects the ability to respond quickly and investigate.
2. ii) Limited Diagnostic Expertise at District Level: Diagnostic expertise was reported to be limited at the district level. This means that local veterinarians or health personnel may not have the specialised knowledge or equipment required to accurately diagnose RVF or other causes of abortion without relying on central laboratories.
3. iii) Underpayment and Poor Facilitation of Staff: Staff motivation was undermined by inadequate facilitation (6 references) and low pay (1 reference). A KII participant stated: “Some laboratory personnel, sometimes they can get a suspect or probable case and fail to work on samples or screening… because these people are not facilitated; they say why should I go ahead to get these samples when I am not facilitated” (KII-DTO-18). A FGD participant observed: “The challenges we face in fighting the RVF are that the veterinary doctors are not well paid, which means they are not motivated to work (FGD-LO-4, Rwetango). This creates a cycle where demotivated staff provide poor services, further discouraging community reporting.
4. iv) Weak Policies and Regulations: Weaknesses in policies and regulations (8 references) were identified as a capability barrier, indicating a lack of clear frameworks or enforcement mechanisms to support effective surveillance.
5. v) Farmers’ Self-Diagnosis and Management: The common practice of farmers who engage in self-diagnosis and the management of animal diseases (7 references) was noted as a barrier, as it bypasses formal reporting channels and professional veterinary care.

b) Opportunity Barriers.

Physical Opportunity Barriers
1. i) Inadequate Transport Infrastructure: The lack of adequate transport infrastructure was identified as a critical barrier. A district technical officer stated, “Our surveillance personnel lack transport means... This specific vehicle may be needed for other surveillance activities” (KII-DTO-10). This indicates that the personnel responsible for follow-up investigations and sample collection often lack reliable means of transportation, which hinders timely responses. District officials also noted that “we share one vehicle between human and animal health, and it is often broken” (KII-DTO-12). This issue of limited transportation means was frequently mentioned (13 references), with vehicles often being called back mid-surveillance, making the process inefficient.
2. ii) Limited District Laboratory Capacity (Physical Resource Gap): Diagnostic capacity was reported to be severely limited at the district level. A MoH Technical Officer noted: “...not having access to a state-of-the-art lab that can easily conduct abortion screening is also a barrier...” (KII-MOH-8). This indicates a lack of physical laboratory infrastructure and equipment at the local level, making it difficult to conduct timely diagnostic tests for livestock abortions. This often results in the reliance on distant national laboratories, leading to delays. Other physical resource gaps included limited storage capacity (4 references) and general limited infrastructure (4 references).
3. iii) Geographic Remoteness and Challenging Terrain: The geographic remoteness of certain communities was reported as a barrier. This physical distance from veterinary services, health facilities and transport routes makes it difficult for community members to report cases and for response teams to reach affected areas promptly. Furthermore, rough terrain (9 references) was frequently cited as an additional physical obstacle, making access to remote areas challenging for surveillance teams. The sheer size of the districts (5 references) also contributed to the difficulty in covering all areas effectively.
4. iv) Large District Size for Surveillance Coverage: The sheer size of the districts (5 references) contributed to the difficulty in covering all areas effectively. A district official noted: “Isingiro is a very big district, very many roads are full of mud and distant, by the time you reach there, you find that even this person you are going to take samples from, he has already been told by people that we are coming, and sometimes they run away (KII-DTO-29). This geographical challenge compounds staffing shortages.
5. v) Limited Cold Chain and Storage Capacity: Limited storage capacity (4 references) and general limited infrastructure (4 references) were not cited as physical resource gaps. A FGD participant from Kikagati noted: “We were looking at other facilities for keeping drugs. We would fail to get them to keep up with vaccines. We did not have them. Even storage channels. Even when there is an outbreak, they use health centre III refrigerators to keep the vaccines” (FGD-LCO-Kikagati). A district official confirmed: “We have a challenge of storage facilities for keeping our drugs and vaccines as a department of veterinary” (KII-DTO-9)
Social Opportunity Barriers
1. i) Poor Coordination Between One Health Sectors at Sub-National Level: Poor coordination between human and animal health sectors, particularly at the sub-national (district) level, was a pervasive barrier. A MoH Technical Officer observed: “...the most challenging part where we have not strengthened is the ‘one health approach’...” (KII-MOH-5). This indicates a lack of operational collaboration and resource sharing between the two sectors, leading to fragmented efforts in surveillance and response.
2. ii) Political and Public Opinion Factors: Political considerations and public opinion (7 references) were identified as social opportunity barriers. These factors can influence decision-making and resource allocation, potentially hindering surveillance efforts. The fear among traders of market disruption (5 references) due to RVF reports also created a disincentive for transparent reporting. Additionally, the perception that RVF was not sufficiently prioritised (3 references) and the involvement of two ministries in charge (2 references), leading to potential bureaucratic hurdles, were noted as structural barriers.
3. iii) No Local Name for RVF (Linguistic/Cultural Barrier): The absence of a local name for RVF was identified as a significant barrier to recognition and reporting. A policymaker stated that “Most people do not know the disease [RVF], as it does not have a local name...” (KII-PM-1). This linguistic gap prevents the disease from entering common community discourse and shared knowledge, making it difficult for community members to communicate about or identify the disease.
4. iv) Economic Sensitivity of RVF Reporting (Fear of Trade/Tourism Impacts): The economic sensitivity surrounding RVF reporting was identified as a major barrier. A district official candidly stated, “When you talk about the Rift Valley fever, you spoil someone’s market for products… as a country” (KII-DO-3). This indicates a fear among officials and potential community members that reporting RVF could lead to negative economic consequences, such as trade restrictions or impacts on tourism, creating disincentives for transparent reporting.
5. v) RVF not prioritised in practice: Despite policy-level prioritisation, the perception that RVF was not sufficiently prioritised in practice (3 references) was noted as a structural barrier. A KII participant from the Science, Technology and Innovation Secretariat – Office of the President (STI-OP) noted: “The opportunities for RVF, of course, as other diseases exist; the only unfortunate bit is that the country has not prioritised it” (KII-NTO-12). This gap between policy rhetoric and resource allocation undermines surveillance efforts.
6. vi) Two Ministries in Charge: The involvement of two ministries (MAAIF and MoH) (2 references) was noted as a potential source of bureaucratic hurdles and coordination gaps. A KII participant explained: “The challenge we have as a country with RVF is that it falls into two Ministries; we have MAAIF, MoH, because the disease starts in animals and small ruminants; it is detected late in most cases, a spill over (KII-ACAD-2). This structural fragmentation created ambiguity about leadership and accountability.
7. vii) Limited Research about RVF: Limited research (1 reference) was noted as a barrier to understanding disease drivers and designing evidence-based interventions. A KII participant noted: “There is also limited research because we don’t know the drivers of these outbreaks, the exact drivers explaining why the outbreaks have come up somewhere are lacking” (KII-NTO-2).

Motivation Barriers

Reflective Motivation Barriers.

i. Absence of Formal Feedback Mechanisms: The lack of formal mechanisms for providing feedback to reporters was consistently cited as a barrier. This means that people who report cattle abortions often do not receive information on the outcome of their report, such as diagnostic results or actions taken. The lack of feedback after making a report was the most cited demotivating factor. A policymaker observed that “It has not been easy because when someone reports, they should get action and feedback... most technical people have been demotivated” (KII-PM-2). Participants in various FGD also stated that “we report, but we never hear what happened” (FGD-LO-3, FGD-LCL-7, FGD-LO-11). This indicates that when reporting efforts do not produce visible results or communication, individuals perceive their actions as futile, leading to reduced motivation to report in the future.
ii. Distrust in Government Response: Distrust in the government’s ability or willingness to respond effectively and fairly was identified as a motivational barrier. When community members lack confidence in authorities, they are less likely to participate in formal reporting systems.
iii) Preference for Self-Treatment Practices Over Formal Reporting: The common practice of self-treatment for animal illnesses has been reported to be a significant barrier to formal reporting. A livestock owner explained, “What stops people from reporting… People were used to buying drugs in stores, so they decided to buy from stores and treat their animals themselves” (FGD-LO-6). This indicates that community members often opt for readily available, informal solutions, bypassing the formal surveillance system.
iv) Fear of Quarantine and Movement Restrictions: A significant motivational barrier was the fear that quarantine and movement restrictions would be imposed after an RVF report. This fear stems from the potential economic disruption and social inconvenience associated with such measures, leading people to hesitate to report.
v) Low Perceived Risk of RVF During Inter-Epidemic Periods: During periods between RVF outbreaks (inter-epidemic periods), the perceived risk of the disease was reported to be low. This reduced perception of threat leads to decreased vigilance and motivation to report potential cases, as the urgency and immediate threat are not felt.

Automatic motivation barriers.

i) Economic Sensitivity of RVF Reporting (Fear of Trade/Tourism Impacts): As noted in Opportunity Barriers, the economic sensitivity of RVF reporting also functions as a strong motivational barrier. Fear of negative economic repercussions, such as trade bans or reduced tourism, directly discourages individuals and even officials from reporting suspected cases, as they anticipate adverse outcomes for themselves or the country.
ii. Despair from repeated Unanswered Reports: Beyond the reflective assessment that reporting is futile, repeated experiences of unanswered reports create an automatic emotional response of hopelessness and resignation. Participants who had reported multiple times without any feedback or action described a sense of learned helplessness – they simply stopped reporting because experience had taught them it made no difference.
iii. Cultural Shame associated with Disease Outbreaks: Implicit in several responses was a sense of cultural shame or stigma associated with having a disease in one’s herd. This emotional response can automatically suppress reporting, as farmers may fear being seen as having “unclean” animals or being responsible for bringing disease into the community.

Discussion

This study provides a theoretically grounded analysis of the behavioural determinants shaping community-based livestock abortion reporting for RVF surveillance in Uganda. By applying the COM-B framework and systematically comparing perspectives across participant groups, we have generated three novel insights that extend current understanding of zoonotic disease surveillance in resource-limited settings. First, the barriers and facilitators are not merely multiple, but they are interconnected; deficits in one domain amplify deficits in others.

Secondly, the same phenomenon is understood fundamentally differently by policymakers and communities, with implications for intervention design that addresses both systemic structures and lived experiences. Thirdly, the ranking of barriers by frequency and emphasis reveals that the most powerful demotivator was the absence of feedback, which was different from the most cited barrier – knowledge deficits. This suggests that intervention priorities must be guided by impact rather than prevalence alone.

The most frequently cited barrier, poor community knowledge about RVF (55 references), represents a foundational deficit that cascades through the entire surveillance system. Communities cannot report what they cannot recognise, especially a disease that lacks a local name. The finding that all confirmed RVF cases were initially submitted as Ebola suspects illustrates the profound consequences of this cognitive gap. This aligns with Tumusiime et al [34], who documented similar knowledge deficits in Ugandan pastoral communities, and we extend their work by quantifying the magnitude of the gap and revealing its linguistic dimension. Although Tumusiime et al found that pastoralists in Napak district had a local name for RVF (Lonyang – a term symbolising jaundice, high fever, abortions in pregnant cows and sudden deaths in calves, this name was not reported in Isingiro district. Their study sampled from four districts of Napak, Butebo, Isingiro and Lyantonde and reported that livestock keepers in all four districts had knowledge of RVF and even had local names or descriptions of it; the actual published findings do not specify the local name for Masha (sub-county), one of their study sites. The persistence of knowledge gaps is corroborated by multiple studies across East Africa. Chiuya et al (2023) demonstrated that in Baringo South, Kenya, only 9.6% attained at least half of the total knowledge score on clinical signs, transmissions, and prevention [35]. Odinoh et al (2025) in their multi-country study in Uganda, the Democratic Republic of Congo and Kenya, found that only 20.5% of 4,806 participants had basic knowledge of RVF [36]. The poor community knowledge of RVF and the absence of a local name for RVF is not merely a translation problem but a structural constraint. A participatory process that enables communities to name and conceptualise the RVF threat in their own terms would contribute to addressing this gap.

The second-ranked barrier, severe understaffing of veterinary professionals (18 references), represents a systemic gap. The finding that veterinary officers and paraveterinary staff cover up to four or more sub-counties and that the district administrators are unaware of true staffing levels reveals gaps in human resource governance that no amount of community sensitisation can overcome. This aligns with Vudriko et al [37] and Kebede et al [38], who documented identical challenges across East Africa, but our findings add a crucial dimension: the staffing gap is uniformly distributed but concentrates its impact in remote communities, where farmers report waiting “5 months without a veterinary doctor”. Indeed, this was reported in our broader study, where we engaged closely with the private veterinary practitioners in Isingiro district, which facilitated reporting and investigation of livestock abortions [4]. This geographical inequity means that surveillance operates as a two-tier system: Response in accessible areas is present, but is absent in hard-to-reach locations, where ecological conditions may heighten disease risk. Addressing this requires not simply increasing the absolute numbers of staff, but targeted deployment to underserved areas with appropriate retention incentives [39].

The third-ranked barrier was the absence of feedback (12 references), which was perhaps the most striking finding, revealing that motivational factors can outweigh capability deficits in determining reporting behaviour. Despite being cited less frequently than knowledge gaps, the absence of feedback emerged as the most powerful demotivator across all participant groups. The repeated refrain “we report, but we never hear what happened” represents not merely a request for information but a fundamental breakdown in the social contract between communities and the surveillance system. When reporting yields no visible response, farmers rationally conclude that their effort is futile and revert to self-treatment practices. This finding resonates with Carlfjord et al. [40] and Chilundo et al [41], who documented similar disengagement in health incident reporting systems when feedback was absent. In contrast, the issue of reporting needs to be viewed as a potential barrier. How? If farmers report and the feedback from the government is to institute geographical quarantine or livestock animal movement restrictions, this is likely to hinder future reports [42,43]. This is because farmers rely on livestock for their livelihoods, and quarantining livestock movements means limited or no trade, likely shutting their businesses. One possible intervention for this is coming up with clear and stronger compensation mechanisms for the farmers. However, this compensation addresses the economic needs of the farmers, but livestock to farmers is beyond money. There are strong social ties to this, and it needs to be considered as well in the compensation plans/mechanisms. Future work needs to dismantle this if reporting for surveillance is to be scaled up effectively. Crucially, our comparative analysis reveals that the absence of feedback operates differently across levels: policymakers experience it as professional demotivation, whereas communities experience it as personal disillusionment. Effective feedback mechanisms must therefore operate at both levels, providing technical case updates to officers while offering communities acknowledgement and explanation [44].

The systematic comparison between KII and FGD perspectives reveals that barriers are not simply present or absent but are understood and experienced differently across levels of the system. This divergence is not a limitation to be resolved but an analytical asset that illuminates the multi-level nature of surveillance failure. Policymakers frame economic concerns at the macroeconomic level (threats to international trade, national tourism) while communities experience them at the household level (inability to pay school fees, loss of dowry cattle). Policymakers discuss One Health coordination failures as institutional problems; communities experience their effects as inaccessible veterinarians and unresponsive systems. This implies that top-down solutions developed from policy perspectives alone will miss the experiential realities that determine whether communities use surveillance systems or not. Bottom-up solutions addressing immediate community frustrations will fail if they do not tackle the structural causes that perpetuate those frustrations. Effective intervention requires integrating both perspectives, addressing understaffing while simultaneously ensuring remaining staff are accessible, establishing feedback systems while ensuring they provide meaningful acknowledgement [12,44].

As demonstrated in our interconnectedness analysis (Section 3.1), the absence of feedback could create a vicious cycle of disengagement, while fear of economic loss is likely to provide only a fragile compensatory mechanism that collapses when reporters encounter an unresponsive system. This could also explain why training intervention alone consistently fails to produce sustained behaviour change [1]. Knowledge without a responsive system to receive reports and provide feedback can not take root. Interventions must therefore address all domains simultaneously.

With these findings, we suggest four priority areas for interventions: First, address the foundational knowledge deficit through participatory processes that co-develop local RVF terminology, integrate disease recognition into existing community structures (LC1 meetings, VHT/CHWs networks), and leverage mobile phone penetration for ongoing reinforcement. Second, resolve the human staffing gaps through targeted recruitment for underserved areas, retention incentives, and engagement of private veterinary practitioners who are already serving communities but lack formal surveillance links. Third, establish simple, reliable feedback mechanisms that operate at both professional and community levels. For example, SMS notifications confirming report receipt, community meetings updating on investigation outcomes, and regular communications from district veterinary offices. Fourth, mitigate economic disincentives through compensation schemes, subsidised veterinary services for reported cases, or micro-insurance that transforms reporting from a financial threat to tangible benefits.

Crucially, these interventions must be designed and implemented through structures that already function, such as LC1 meetings, VHT networks, and WhatsApp groups, rather than creating parallel systems that communities must learn anew. The facilitators our study identified were: trained community health workers, mobile phone penetration, community governance structures, and national laboratory capacity, which provide ready-made infrastructure. This challenge is not a lack of capacity but a fragmentation of capacity across sectors and levels. Strengthening surveillance requires not building new systems but connecting and reinforcing existing ones.

This study possesses several strengths. The application of COM-B provides a systematic, theory-driven analysis that identifies modifiable behavioural determinants. The large, diverse sample and achievement of data saturation ensured comprehensive capture of relevant perspectives. The rigorous analytical approach, including independent coding verification and member checking, enhanced trustworthiness. The comparative analysis between participant groups adds methodological depth, revealing how the same phenomenon is understood differently across levels of the system. However, some limitations warrant consideration. As a qualitative study in a single district, findings may not be generalisable, though theoretical transferability is enhanced by the COM-B framework and consistency with regional literature. Social desirability bias may have influenced responses, particularly in FGD settings. The ranking of barriers and facilitators, while grounded in frequency and thematic emphasis, involves subjective judgment. The study captured processes at a single time point following a training intervention; longitudinal research would be needed to assess how barriers evolve. Potential selection bias as presented earlier in the methods section remain an important area for future research using alternative recruitment approaches such as snowball sampling or direct community engagement without gatekeeper mediation. Finally, the study focused on barriers and facilitators to reporting and understanding the full surveillance cascade, which requires additional research.

Supporting information

S1 Appendix. The Training Material used for sensitising different stakeholders about Rift Valley fever disease and the need for abortion surveillance.

https://doi.org/10.1371/journal.pone.0359872.s001

(DOCX)

S1 Table. A List of participants trained as part of capacity-building efforts.

https://doi.org/10.1371/journal.pone.0359872.s002

(DOCX)

S2 Appendix. The FGD guide used during Focus Group Discussions.

https://doi.org/10.1371/journal.pone.0359872.s003

(DOCX)

S3 Appendix. The Key Informant Guide used for Key Informant Interviews.

https://doi.org/10.1371/journal.pone.0359872.s004

(DOCX)

S4 Appendix. Comparative Analysis and Ranking of Facilitators and Barriers: KII vs FGD Perspectives.

https://doi.org/10.1371/journal.pone.0359872.s005

(DOCX)

S5 Appendix. Ranking of Facilitators and Barriers: Ranking of Barriers (Highest to Lowest Priority).

https://doi.org/10.1371/journal.pone.0359872.s006

(DOCX)

Acknowledgments

We are grateful to the technical staff at the Ministry of Agriculture, Animal Industry and Fisheries (Drs. Robert Mwebe and Dan Tumusiime) and the Isingiro District Local Government. We thank Conservation and Ecosystem Health Alliance (CEHA) for administrative and technical support during the implementation of the research. Special appreciation to Mellon Ainembabazi, who was very helpful in conducting the FGDs in the local language and later transcribing the transcripts. Lastly, we are grateful to the different leaders and farmers who agreed to participate in our FGDs and KIIs.

References

1. Hartman A. Rift Valley Fever. Clin Lab Med. 2017 Jun;37(2):285–301. pmid:28457351
- View Article
- PubMed/NCBI
- Google Scholar
2. Ali H, Ali AA, Atta MS, Cepica A. Common, emerging, vector-borne and infrequent abortogenic virus infections of cattle. Transboundary and Emerging Diseases. 2012;59(1):11–25.
- View Article
- Google Scholar
3. Thomas KM, Kibona T, Claxton JR, de Glanville WA, Lankester F, Amani N. Prospective cohort study reveals unexpected aetiologies of livestock abortion in northern Tanzania. Scientific Reports. 2022;12(1):11669.
- View Article
- Google Scholar
4. Walekhwa AW, Conlan AJK, Atim SA, Ademun AR, Hasahya E, Wood JLN. Exploring community-based reporting of livestock abortions for rift valley fever and brucellosis surveillance in Uganda: A pilot study. Sci Rep. 2025;15(1):42641.
- View Article
- Google Scholar
5. Thurmond MC, Picanso JP. A surveillance system for bovine abortion. Preventive Veterinary Medicine. 1990;8(1):41–53.
- View Article
- Google Scholar
6. Giorgio M, Makumbi F, Kibira SPS, Shiferaw S, Seme A, Bell SO, et al. Self-reported abortion experiences in Ethiopia and Uganda, new evidence from cross-sectional community-based surveys. PLOS Glob Public Health. 2023;3(9):e0002340. pmid:37682781
- View Article
- PubMed/NCBI
- Google Scholar
7. Gachohi J, Njoki P, Mogoa E, Otieno F, Muturi M, Mwatondo A, et al. Higher livestock abortion burden in arid and semi-arid lands, Kenya, 2019–2020. PLoS One. 2024;19(2):e0297274. pmid:38386647
- View Article
- PubMed/NCBI
- Google Scholar
8. Gachohi J, Njoki P, Mogoa E, Otieno F, Muturi M, Mwatondo A. Higher livestock abortion burden in arid and semi-arid lands, Kenya, 2019–2020. PLOS ONE. 2024;19(2):e0297274.
- View Article
- Google Scholar
9. Worsley-Tonks KEL, Angwenyi S, Carlson C, Cissé G, Deem SL, Ferguson AW, et al. A framework for managing infectious diseases in rural areas in low- and middle-income countries in the face of climate change-East Africa as a case study. PLOS Glob Public Health. 2025;5(1):e0003892. pmid:39883787
- View Article
- PubMed/NCBI
- Google Scholar
10. Lankester F, Kibona T, Allan KJ, Glanville WA de, Buza JJ, Katzer F, et al. The value of livestock abortion surveillance in Tanzania: identifying disease priorities and informing interventions. bioRxiv. 2024. 2024.01.07.574517. https://doi.org/10.1101/2024.01.07.574517
11. Semango G, Yoder J, Kibona T, Claxton JR, Buza J, Mmbaga BT, et al. Economic Burden of Livestock Abortions in Northern Tanzania. J Agric Appl Econ. 2024;56(2):195–215.
- View Article
- Google Scholar
12. Michie S, van Stralen MM, West R. The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implement Sci. 2011;6:42. pmid:21513547
- View Article
- PubMed/NCBI
- Google Scholar
13. Agaba C, Ferguson L, Emoto S, Haumba M, Kizito D, Kolchina M, et al. Gender and exposure pathways to zoonotic infections in communities at the interface of wildlife conservation areas of Uganda: A qualitative study. Glob Public Health. 2025;20(1):2503858. pmid:40358123
- View Article
- PubMed/NCBI
- Google Scholar
14. Bintoora KKA. The cost of poor land use practices in Lake Nakivale Wetland in Isingiro District, Uganda. Afr J Environ Sci Technol. 2013;7(6):448–56.
- View Article
- Google Scholar
15. Bwengye E, Nagawa GM, Tumwesigye W. Diversity of the On-Farm Crop Dry Spell Adaptation Technologies in Isingiro Town Council, Isingiro District, Uganda. Afr J Clim Chang Resour Sustain. 2023;2(1):102–16.
- View Article
- Google Scholar
16. Mubiru S, Marshall K, Lukuyu BA, Oba P, Ahumuza R, Ouma EA. Beef value chain situation analysis for Uganda. ILRI. 2023. https://hdl.handle.net/10568/138120
17. Kweyu RM, Asokan SM, Ndesanjo RB, Obando JA, Tumbo MH. Climate governance in Eastern Africa: The challenges and prospects of climate change adaptation policies. In: Onyango G, editor. State politics and public policy in Eastern Africa: A comparative perspective. Cham: Springer International Publishing. 2023. p. 347–69. https://doi.org/10.1007/978-3-031-13490-6_16
18. Nagasha JI, Mugisha L, Kaase-Bwanga E, Onyuth H, Ocaido M. Effect of climate change on gender roles among communities surrounding Lake Mburo National Park, Uganda. Emerald Open Research. 2019;1(6).
- View Article
- Google Scholar
19. Taremwa IM, Ashaba S, Ayebazibwe C, Kemeza I, Adrama HO, Omoding D, et al. Mind the gap: scaling up the utilization of insecticide treated mosquito nets using a knowledge translation model in Isingiro district, rural south western Uganda. Health Psychol Behav Med. 2020;8(1):383–97. pmid:34040878
- View Article
- PubMed/NCBI
- Google Scholar
20. Mpeshe SC, Luboobi LS, Nkansah-Gyekye Y. Modeling the impact of climate change on the dynamics of Rift Valley Fever. Comput Math Methods Med. 2014;2014:627586. pmid:24795775
- View Article
- PubMed/NCBI
- Google Scholar
21. Bauer MS, Kirchner J. Implementation science: What is it and why should I care?. Psychiatry Research. 2020;VSI:Implementation Science283:112376.
- View Article
- Google Scholar
22. Cecilia H, Drouin A, Métras R, Balenghien T, Durand B, Chevalier V, et al. Mechanistic models of Rift Valley fever virus transmission: A systematic review. PLoS Negl Trop Dis. 2022;16(11):e0010339. pmid:36399500
- View Article
- PubMed/NCBI
- Google Scholar
23. Connolly SL, Hogan TP, Shimada SL, Miller CJ. Leveraging Implementation Science to Understand Factors Influencing Sustained Use of Mental Health Apps: A narrative review. J Technol Behav Sci. 2021;6(2):184–96.
- View Article
- Google Scholar
24. O.Nyumba T, Wilson K, Derrick CJ, Mukherjee N. The use of focus group discussion methodology: Insights from two decades of application in conservation. Methods Ecol Evol. 2018;9(1):20–32.
- View Article
- Google Scholar
25. Tumusiime D, Nijhof AM, Groschup MH, Lutwama J, Roesel K, Bett B. Participatory survey of risk factors and pathways for Rift Valley fever in pastoral and agropastoral communities of Uganda. Prev Vet Med. 2023;221:106071. pmid:37984160
- View Article
- PubMed/NCBI
- Google Scholar
26. Guest G, Namey E, Chen M. A simple method to assess and report thematic saturation in qualitative research. PLoS One. 2020;15(5):e0232076. pmid:32369511
- View Article
- PubMed/NCBI
- Google Scholar
27. Braun V, Clarke V. Using thematic analysis in psychology. Qualitative Res Psychol. 2006;3(2):77–101.
- View Article
- Google Scholar
28. Azungah T. Qualitative research: deductive and inductive approaches to data analysis. QRJ. 2018;18(4):383–400.
- View Article
- Google Scholar
29. Edhlund B, McDougall A. NVivo 12 Essentials. Lulu.com. 2018.
30. Reynolds J, Kizito J, Ezumah N, Mangesho P, Allen E, Chandler C. Quality assurance of qualitative research: a review of the discourse. Health Res Policy Syst. 2011;9:43. pmid:22182674
- View Article
- PubMed/NCBI
- Google Scholar
31. Birt L, Scott S, Cavers D, Campbell C, Walter F. Member checking: A tool to enhance trustworthiness or merely a nod to validation?. Qual Health Res. 2016;26(13):1802–11.
- View Article
- Google Scholar
32. Candela A. Exploring the function of member checking. TQR. 2019.
- View Article
- Google Scholar
33. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups [Internet]. 2007. p. 349–57. Available from: https://www.equator-network.org/reporting-guidelines/coreq/
34. Tumusiime D, Nijhof AM, Groschup MH, Lutwama J, Roesel K, Bett B. Participatory survey of risk factors and pathways for Rift Valley fever in pastoral and agropastoral communities of Uganda. Prev Vet Med. 2023;221:106071. pmid:37984160
- View Article
- PubMed/NCBI
- Google Scholar
35. PLOS Global Public Health Staff. Correction: Understanding knowledge, attitude and perception of Rift Valley fever in Baringo South, Kenya: A cross-sectional study. PLOS Glob Public Health. 2024;4(6):e0003400. pmid:38870158
- View Article
- PubMed/NCBI
- Google Scholar
36. Odinoh R, Dawa J, Situma S, Nyakarahuka L, Lepore L, Vanlerberghe V, et al. Have you heard of Rift Valley fever? Findings from a multi-country study in East and Central Africa. PLoS One. 2025;20(7):e0327398. pmid:40601695
- View Article
- PubMed/NCBI
- Google Scholar
37. Vudriko P, Ekiri AB, Endacott I, Williams S, Gityamwi N, Byaruhanga J, et al. A survey of priority livestock diseases and laboratory diagnostic needs of animal health professionals and farmers in Uganda. Front Vet Sci. 2021;8:721800. pmid:34631853
- View Article
- PubMed/NCBI
- Google Scholar
38. Kebede G, Moje N, Shafi A, Kassahun R, Merera O, Leta S, et al. Investigation of veterinary service delivery and service providers on dairy cattle farms in central Ethiopia. BMC Vet Res. 2025;21(1):125. pmid:40025467
- View Article
- PubMed/NCBI
- Google Scholar
39. Driscoll DC. Credentialed veterinary technician intrinsic and extrinsic rewards: a narrative review. J Am Vet Med Assoc. 2022;260(9):1069–75. pmid:35417418
- View Article
- PubMed/NCBI
- Google Scholar
40. Carlfjord S, Öhrn A, Gunnarsson A. Experiences from ten years of incident reporting in health care: a qualitative study among department managers and coordinators. BMC Health Serv Res. 2018;18(1):113. pmid:29444680
- View Article
- PubMed/NCBI
- Google Scholar
41. Chilundo B, Sundby J, Aanestad M. Analysing the quality of routine malaria data in Mozambique. Malar J. 2004;3:3. pmid:14998435
- View Article
- PubMed/NCBI
- Google Scholar
42. Ilukor J, Joyce A, Okiror SP. The impacts of COVID-19 on cattle traders and their response in agro-pastoral and pastoral regions in Uganda: A case of Karamoja and Teso cattle traders. Pastoralism. 2022;12(1):18. pmid:35437469
- View Article
- PubMed/NCBI
- Google Scholar
43. Mubiru S, Marshall K, Lukuyu BA, Oba P, Ahumuza R, Ouma EA. Beef value chain situation analysis for Uganda. International Livestock Research Institute. 2023. https://hdl.handle.net/10568/138120
44. Mogg J. A synthesized approach to explaining the expansion of surveillance: risk perceptions and surveillance as a positive feedback loop. Crossings: An Undergraduate Arts J. 2022;2(1).
- View Article
- Google Scholar

[ref1] 1. Hartman A. Rift Valley Fever. Clin Lab Med. 2017 Jun;37(2):285–301. pmid:28457351
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Ali H, Ali AA, Atta MS, Cepica A. Common, emerging, vector-borne and infrequent abortogenic virus infections of cattle. Transboundary and Emerging Diseases. 2012;59(1):11–25.
View Article
Google Scholar

[6] View Article

[7] Google Scholar

[ref3] 3. Thomas KM, Kibona T, Claxton JR, de Glanville WA, Lankester F, Amani N. Prospective cohort study reveals unexpected aetiologies of livestock abortion in northern Tanzania. Scientific Reports. 2022;12(1):11669.
View Article
Google Scholar

[9] View Article

[10] Google Scholar

[ref4] 4. Walekhwa AW, Conlan AJK, Atim SA, Ademun AR, Hasahya E, Wood JLN. Exploring community-based reporting of livestock abortions for rift valley fever and brucellosis surveillance in Uganda: A pilot study. Sci Rep. 2025;15(1):42641.
View Article
Google Scholar

[12] View Article

[13] Google Scholar

[ref5] 5. Thurmond MC, Picanso JP. A surveillance system for bovine abortion. Preventive Veterinary Medicine. 1990;8(1):41–53.
View Article
Google Scholar

[15] View Article

[16] Google Scholar

[ref6] 6. Giorgio M, Makumbi F, Kibira SPS, Shiferaw S, Seme A, Bell SO, et al. Self-reported abortion experiences in Ethiopia and Uganda, new evidence from cross-sectional community-based surveys. PLOS Glob Public Health. 2023;3(9):e0002340. pmid:37682781
View Article
PubMed/NCBI
Google Scholar

[18] View Article

[19] PubMed/NCBI

[20] Google Scholar

[ref7] 7. Gachohi J, Njoki P, Mogoa E, Otieno F, Muturi M, Mwatondo A, et al. Higher livestock abortion burden in arid and semi-arid lands, Kenya, 2019–2020. PLoS One. 2024;19(2):e0297274. pmid:38386647
View Article
PubMed/NCBI
Google Scholar

[22] View Article

[23] PubMed/NCBI

[24] Google Scholar

[ref8] 8. Gachohi J, Njoki P, Mogoa E, Otieno F, Muturi M, Mwatondo A. Higher livestock abortion burden in arid and semi-arid lands, Kenya, 2019–2020. PLOS ONE. 2024;19(2):e0297274.
View Article
Google Scholar

[26] View Article

[27] Google Scholar

[ref9] 9. Worsley-Tonks KEL, Angwenyi S, Carlson C, Cissé G, Deem SL, Ferguson AW, et al. A framework for managing infectious diseases in rural areas in low- and middle-income countries in the face of climate change-East Africa as a case study. PLOS Glob Public Health. 2025;5(1):e0003892. pmid:39883787
View Article
PubMed/NCBI
Google Scholar

[29] View Article

[30] PubMed/NCBI

[31] Google Scholar

[ref10] 10. Lankester F, Kibona T, Allan KJ, Glanville WA de, Buza JJ, Katzer F, et al. The value of livestock abortion surveillance in Tanzania: identifying disease priorities and informing interventions. bioRxiv. 2024. 2024.01.07.574517. https://doi.org/10.1101/2024.01.07.574517

[ref11] 11. Semango G, Yoder J, Kibona T, Claxton JR, Buza J, Mmbaga BT, et al. Economic Burden of Livestock Abortions in Northern Tanzania. J Agric Appl Econ. 2024;56(2):195–215.
View Article
Google Scholar

[34] View Article

[35] Google Scholar

[ref12] 12. Michie S, van Stralen MM, West R. The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implement Sci. 2011;6:42. pmid:21513547
View Article
PubMed/NCBI
Google Scholar

[37] View Article

[38] PubMed/NCBI

[39] Google Scholar

[ref13] 13. Agaba C, Ferguson L, Emoto S, Haumba M, Kizito D, Kolchina M, et al. Gender and exposure pathways to zoonotic infections in communities at the interface of wildlife conservation areas of Uganda: A qualitative study. Glob Public Health. 2025;20(1):2503858. pmid:40358123
View Article
PubMed/NCBI
Google Scholar

[41] View Article

[42] PubMed/NCBI

[43] Google Scholar

[ref14] 14. Bintoora KKA. The cost of poor land use practices in Lake Nakivale Wetland in Isingiro District, Uganda. Afr J Environ Sci Technol. 2013;7(6):448–56.
View Article
Google Scholar

[45] View Article

[46] Google Scholar

[ref15] 15. Bwengye E, Nagawa GM, Tumwesigye W. Diversity of the On-Farm Crop Dry Spell Adaptation Technologies in Isingiro Town Council, Isingiro District, Uganda. Afr J Clim Chang Resour Sustain. 2023;2(1):102–16.
View Article
Google Scholar

[48] View Article

[49] Google Scholar

[ref16] 16. Mubiru S, Marshall K, Lukuyu BA, Oba P, Ahumuza R, Ouma EA. Beef value chain situation analysis for Uganda. ILRI. 2023. https://hdl.handle.net/10568/138120

[ref17] 17. Kweyu RM, Asokan SM, Ndesanjo RB, Obando JA, Tumbo MH. Climate governance in Eastern Africa: The challenges and prospects of climate change adaptation policies. In: Onyango G, editor. State politics and public policy in Eastern Africa: A comparative perspective. Cham: Springer International Publishing. 2023. p. 347–69. https://doi.org/10.1007/978-3-031-13490-6_16

[ref18] 18. Nagasha JI, Mugisha L, Kaase-Bwanga E, Onyuth H, Ocaido M. Effect of climate change on gender roles among communities surrounding Lake Mburo National Park, Uganda. Emerald Open Research. 2019;1(6).
View Article
Google Scholar

[53] View Article

[54] Google Scholar

[ref19] 19. Taremwa IM, Ashaba S, Ayebazibwe C, Kemeza I, Adrama HO, Omoding D, et al. Mind the gap: scaling up the utilization of insecticide treated mosquito nets using a knowledge translation model in Isingiro district, rural south western Uganda. Health Psychol Behav Med. 2020;8(1):383–97. pmid:34040878
View Article
PubMed/NCBI
Google Scholar

[56] View Article

[57] PubMed/NCBI

[58] Google Scholar

[ref20] 20. Mpeshe SC, Luboobi LS, Nkansah-Gyekye Y. Modeling the impact of climate change on the dynamics of Rift Valley Fever. Comput Math Methods Med. 2014;2014:627586. pmid:24795775
View Article
PubMed/NCBI
Google Scholar

[60] View Article

[61] PubMed/NCBI

[62] Google Scholar

[ref21] 21. Bauer MS, Kirchner J. Implementation science: What is it and why should I care?. Psychiatry Research. 2020;VSI:Implementation Science283:112376.
View Article
Google Scholar

[64] View Article

[65] Google Scholar

[ref22] 22. Cecilia H, Drouin A, Métras R, Balenghien T, Durand B, Chevalier V, et al. Mechanistic models of Rift Valley fever virus transmission: A systematic review. PLoS Negl Trop Dis. 2022;16(11):e0010339. pmid:36399500
View Article
PubMed/NCBI
Google Scholar

[67] View Article

[68] PubMed/NCBI

[69] Google Scholar

[ref23] 23. Connolly SL, Hogan TP, Shimada SL, Miller CJ. Leveraging Implementation Science to Understand Factors Influencing Sustained Use of Mental Health Apps: A narrative review. J Technol Behav Sci. 2021;6(2):184–96.
View Article
Google Scholar

[71] View Article

[72] Google Scholar

[ref24] 24. O.Nyumba T, Wilson K, Derrick CJ, Mukherjee N. The use of focus group discussion methodology: Insights from two decades of application in conservation. Methods Ecol Evol. 2018;9(1):20–32.
View Article
Google Scholar

[74] View Article

[75] Google Scholar

[ref25] 25. Tumusiime D, Nijhof AM, Groschup MH, Lutwama J, Roesel K, Bett B. Participatory survey of risk factors and pathways for Rift Valley fever in pastoral and agropastoral communities of Uganda. Prev Vet Med. 2023;221:106071. pmid:37984160
View Article
PubMed/NCBI
Google Scholar

[77] View Article

[78] PubMed/NCBI

[79] Google Scholar

[ref26] 26. Guest G, Namey E, Chen M. A simple method to assess and report thematic saturation in qualitative research. PLoS One. 2020;15(5):e0232076. pmid:32369511
View Article
PubMed/NCBI
Google Scholar

[81] View Article

[82] PubMed/NCBI

[83] Google Scholar

[ref27] 27. Braun V, Clarke V. Using thematic analysis in psychology. Qualitative Res Psychol. 2006;3(2):77–101.
View Article
Google Scholar

[85] View Article

[86] Google Scholar

[ref28] 28. Azungah T. Qualitative research: deductive and inductive approaches to data analysis. QRJ. 2018;18(4):383–400.
View Article
Google Scholar

[88] View Article

[89] Google Scholar

[ref29] 29. Edhlund B, McDougall A. NVivo 12 Essentials. Lulu.com. 2018.

[ref30] 30. Reynolds J, Kizito J, Ezumah N, Mangesho P, Allen E, Chandler C. Quality assurance of qualitative research: a review of the discourse. Health Res Policy Syst. 2011;9:43. pmid:22182674
View Article
PubMed/NCBI
Google Scholar

[92] View Article

[93] PubMed/NCBI

[94] Google Scholar

[ref31] 31. Birt L, Scott S, Cavers D, Campbell C, Walter F. Member checking: A tool to enhance trustworthiness or merely a nod to validation?. Qual Health Res. 2016;26(13):1802–11.
View Article
Google Scholar

[96] View Article

[97] Google Scholar

[ref32] 32. Candela A. Exploring the function of member checking. TQR. 2019.
View Article
Google Scholar

[99] View Article

[100] Google Scholar

[ref33] 33. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups [Internet]. 2007. p. 349–57. Available from: https://www.equator-network.org/reporting-guidelines/coreq/

[ref34] 34. Tumusiime D, Nijhof AM, Groschup MH, Lutwama J, Roesel K, Bett B. Participatory survey of risk factors and pathways for Rift Valley fever in pastoral and agropastoral communities of Uganda. Prev Vet Med. 2023;221:106071. pmid:37984160
View Article
PubMed/NCBI
Google Scholar

[103] View Article

[104] PubMed/NCBI

[105] Google Scholar

[ref35] 35. PLOS Global Public Health Staff. Correction: Understanding knowledge, attitude and perception of Rift Valley fever in Baringo South, Kenya: A cross-sectional study. PLOS Glob Public Health. 2024;4(6):e0003400. pmid:38870158
View Article
PubMed/NCBI
Google Scholar

[107] View Article

[108] PubMed/NCBI

[109] Google Scholar

[ref36] 36. Odinoh R, Dawa J, Situma S, Nyakarahuka L, Lepore L, Vanlerberghe V, et al. Have you heard of Rift Valley fever? Findings from a multi-country study in East and Central Africa. PLoS One. 2025;20(7):e0327398. pmid:40601695
View Article
PubMed/NCBI
Google Scholar

[111] View Article

[112] PubMed/NCBI

[113] Google Scholar

[ref37] 37. Vudriko P, Ekiri AB, Endacott I, Williams S, Gityamwi N, Byaruhanga J, et al. A survey of priority livestock diseases and laboratory diagnostic needs of animal health professionals and farmers in Uganda. Front Vet Sci. 2021;8:721800. pmid:34631853
View Article
PubMed/NCBI
Google Scholar

[115] View Article

[116] PubMed/NCBI

[117] Google Scholar

[ref38] 38. Kebede G, Moje N, Shafi A, Kassahun R, Merera O, Leta S, et al. Investigation of veterinary service delivery and service providers on dairy cattle farms in central Ethiopia. BMC Vet Res. 2025;21(1):125. pmid:40025467
View Article
PubMed/NCBI
Google Scholar

[119] View Article

[120] PubMed/NCBI

[121] Google Scholar

[ref39] 39. Driscoll DC. Credentialed veterinary technician intrinsic and extrinsic rewards: a narrative review. J Am Vet Med Assoc. 2022;260(9):1069–75. pmid:35417418
View Article
PubMed/NCBI
Google Scholar

[123] View Article

[124] PubMed/NCBI

[125] Google Scholar

[ref40] 40. Carlfjord S, Öhrn A, Gunnarsson A. Experiences from ten years of incident reporting in health care: a qualitative study among department managers and coordinators. BMC Health Serv Res. 2018;18(1):113. pmid:29444680
View Article
PubMed/NCBI
Google Scholar

[127] View Article

[128] PubMed/NCBI

[129] Google Scholar

[ref41] 41. Chilundo B, Sundby J, Aanestad M. Analysing the quality of routine malaria data in Mozambique. Malar J. 2004;3:3. pmid:14998435
View Article
PubMed/NCBI
Google Scholar

[131] View Article

[132] PubMed/NCBI

[133] Google Scholar

[ref42] 42. Ilukor J, Joyce A, Okiror SP. The impacts of COVID-19 on cattle traders and their response in agro-pastoral and pastoral regions in Uganda: A case of Karamoja and Teso cattle traders. Pastoralism. 2022;12(1):18. pmid:35437469
View Article
PubMed/NCBI
Google Scholar

[135] View Article

[136] PubMed/NCBI

[137] Google Scholar

[ref43] 43. Mubiru S, Marshall K, Lukuyu BA, Oba P, Ahumuza R, Ouma EA. Beef value chain situation analysis for Uganda. International Livestock Research Institute. 2023. https://hdl.handle.net/10568/138120

[ref44] 44. Mogg J. A synthesized approach to explaining the expansion of surveillance: risk perceptions and surveillance as a positive feedback loop. Crossings: An Undergraduate Arts J. 2022;2(1).
View Article
Google Scholar

[140] View Article

[141] Google Scholar

Figures

Abstract

Introduction

Methods

Ethics Statement

Study design

Data Collection Period

Description of the study sites

Theoretical framework of this study

Intervention (Training)

Data collection

Participants and sampling.

Procedure of focus group discussions

Procedure of key informant interviews

Interviewer characteristics and team composition

Data processing and quality assurance

Validation process

Outcome of member checking

Independent validation by research assistants

Reflexivity and researcher positionality

Mitigation strategies

Reporting of findings

Results

Comparative perspectives across participant groups

Facilitators

a) Capability Facilitators.

b) Opportunity Facilitators.

c) Motivation Facilitators.

Barriers

a) Capability Barriers.

b) Opportunity Barriers.

Motivation Barriers

Reflective Motivation Barriers.

Automatic motivation barriers.

Discussion

Supporting information

S1 Appendix. The Training Material used for sensitising different stakeholders about Rift Valley fever disease and the need for abortion surveillance.

S1 Table. A List of participants trained as part of capacity-building efforts.

S2 Appendix. The FGD guide used during Focus Group Discussions.

S3 Appendix. The Key Informant Guide used for Key Informant Interviews.

S4 Appendix. Comparative Analysis and Ranking of Facilitators and Barriers: KII vs FGD Perspectives.

S5 Appendix. Ranking of Facilitators and Barriers: Ranking of Barriers (Highest to Lowest Priority).

Acknowledgments

References