Problem structuring using the Iceberg model
The foodborne disease workshop for stakeholders identified 33 elements and underlying behaviors influencing the current system of FBD-surveillance in Africa. The Iceberg model was employed to evaluate the interplay of these elements within the existing system of FBD surveillance in Africa. At each level, these factors were categorized based on their primary influence in society, the healthcare system, or the food and agricultural sector (Fig. 2). “Events” (i.e., outcomes that are visible and above the waterline in the iceberg) included those related to foodborne disease infections, severity of illnesses, including hospital admissions, and access to food resources. These outcomes were consistently reflected in both the stakeholder workshop and the scoping review findings (Supplementary Data 2). Additional literature identified outside the scoping review further supported the relevance of these observed outcomes2,5,17,41,42. “Patterns/trends” included the increasing demand for food and the inadequate prioritization of surveillance for locally consumed food sold in the informal sector, ill-equipped diagnostic facilities, proximity of medical facilities, limited data sharing and reporting, climate change, and food fraud. These patterns were frequently cited by workshop participants and were further substantiated through the scoping review and complementary literature sources13,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60. The “Underlying structures” were identified as limited education, human- and animal interactions (i.e., free-roaming animals, proximity of livestock and crops (mixed farming), close contact between farmers, their families, and animals). Other foundational challenges identified were antimicrobial resistance and usage, cost of treatment, access to sanitation, unhygienic conditions in slaughterhouses and transportation of farm produce, and disposal of untreated agricultural waste. These structural drivers were consistently reflected in both the participant discussions and the broader body of literature consulted2,5,12,13,17,21,54,55,56,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82. Finally, the ‘Mental models’ were described as primarily driven by stigma, political- or societal knowledge, and inadequate prioritization and preservation of age-long agricultural practices, e.g., free-roaming animals, which was supported by the scoping review and the complementary literature sources5,21,67,83,84,85. Participants in the workshop described the increasing population as well as limited information as central focus for interventions. They emphasized that, as the increasing population intensifies the demand on resources, there is a need for sustainable solutions that will be able to work long-term.
Additionally, color shading is used to indicate the primary domain in which each element has the greatest impact: broader society (orange), the healthcare system (blue), and the food and agricultural sector (green). The literature search conducted as part of the validation of the results is referenced in Supplementary Data 2.
Scoping review results
The purpose of this review was to assess whether the challenges raised by participants were reflected in existing research and to identify any additional elements or overlooked issues that were not mentioned during the workshop. The search strings, number of hits, and corresponding curated publications are summarized in Supplementary Data 2. Publications deemed relevant were recorded and are presented in the Supplementary References.
Causal loop diagrams and archetype definitions
The elements identified in the Iceberg model were classified as internal, external, or outside elements and were displayed in an interconnected map (Fig. 3). To limit the system to elements that had a direct impact on the present FBD surveillance system, outside elements were simplified when necessary. This was the case for the elements “Climate change,” “Malnutrition,” “Sewage systems,” “Economic instability,” “Increasing demand for food,”, “Disposal of agricultural waste,” and “Proximity of medical facilities,” which were identified as important drivers outside the system.
Gray circles depicts elements outside of the system. Arrows indicate directional relationships between elements: red arrows signify negative or decreasing effects, while blue arrows indicate positive or increasing effects.
The CLDs were structured by the primary categories of influence on FBD surveillance: societal factors, healthcare system factors, and food and agriculture factors. Each CLD represents a set of interconnected feedback loops within these categories, revealing how specific variables reinforce or counterbalance challenges within the surveillance system. For example, reinforcing loops may perpetuate existing issues by amplifying factors like underreporting and limited diagnostic capacity, while balancing loops can contribute to system stability by counteracting such changes through public pressure or policy responses.
The key variables included “Underreporting of disease,” “Limited education and awareness,” “Resource allocation,” “Mistrust in public health and regulatory systems,” and “Antimicrobial resistance”. These were chosen because they represent critical points of influence in the system across all three categories, which either directly or indirectly impacted the shape of the system’s response to FBD and their ability to detect them. System dynamics show that these elements are connected to each other through several reinforcing or balancing feedback loops. The CLDs for each individual loop are available in the Supplementary Information (Supplementary Figs. 1–8).
The loops named “Public trust and compliance with surveillance” and “Public pressure and system response” are both balancing loops that illustrates how the public perception can both hinder and harbor the implementation of FBD surveillance systems; rising FBDs can lead to mistrust in health systems, reduced compliance and surveillance reporting, which then exacerbates the underreporting of disease and hides the true extent of FBDs, thus perpetuating the issue. On the other hand, increased incidence of FBDs can also drive public pressure on health authorities to improve surveillance, which can lead to reduced FBD incidence and restored trust in the future. Similarly, the loop named “Public demand and agricultural surveillance” shows how underreporting initially limits awareness and investment in agricultural surveillance but can eventually drive public demand for safer practices and surveillance improvements, as the FBD rates increase.
The loop named “Diagnostic capacity and disease reporting” shows that underfunded diagnostic facilities contribute to underreporting, which leads to decreased awareness of FBDs and consequently less resource allocation, further worsening diagnostic capacity.
These loops are all balancing loops that keep the system in place by counteracting change and resist drastic changes.
Reinforcing loops included “Education and cost-benefit perception of food safety” and “Cross-contamination cycle in agriculture”, which showed how the limited education reduces the understanding of food safety, which normalizes FBD illnesses and lowers the public demand for food safety measures, thereby reinforcing low investment in FBD surveillance. Similarly, improper handling of food leads to contamination, which increases FBDs and decrease public trust, which in turn reduces compliance and further perpetuating unsafe food practices.
Other reinforcing loops included “Severity of illness and resource strain,” where the severity of illnesses increase the hospital admissions and puts increased strain on resources, which diverts funding away from surveillance and prevention, and leads to increased FBD incidence, and the “Food production and safety compliance” loop, which show how economic pressure can encourage antimicrobial use, leading to resistance and worsening of FBD diseases, and further economic pressure on food production systems. These loops lead to gradual worsening of FBD incidence or severity because of resource strain, low compliance, and economic pressures compound issues.
Archetypes of the system were identified within each loop to help clarify how certain behaviors were perpetuated and highlight how common structural challenges could hinder the effectiveness of a FBD surveillance system. Several recurrent patterns emerged within the FBD surveillance system (Supplementary Data 3). The “Fixes that fail” archetype highlighted how short-term responses, such as temporary increases in diagnostic resources during outbreaks, fail to create sustainable diagnostic capacity improvements, leading to repeated gaps in disease detection. This was also the case for the “Escalation” archetype, where the increased severity of FBD leads to healthcare resources becoming more strained and creates a cycle of escalating demand on an already strained system. Similarly, the “Drifting goals” archetype showed how inconsistent resource allocation and low public trust have contributed to reduced compliance with food safety standards, normalizing high levels of foodborne disease incidence.
The “Limits to success” archetype showed how growth in diagnostic capacity may initially improve, since resources are allocated to the problem, but as the system evolves, financial constraints, limited skilled personnel, or outdated infrastructure begin to slow the progress further. This prevents the diagnostic capacity from expanding, thus hindering the system’s ability to effectively monitor and respond to FBDs. This archetype goes together with the “Growth and underinvestment” archetype, which describes how a system’s growth is limited due to insufficient investment in critical infrastructure. If FBD risks are underreported, the public’s awareness associated with food is reduced, which removes the pressure to prioritize food safety measures in agriculture. Without the pressure for legislators to invest in FBD surveillance, the system continues to miss effectively addressing the problem, which leads to delayed responses and limited surveillance capabilities. Same with the “Shifting the burden” archetype, where the resources are often allocated to immediate, symptomatic responses to outbreaks rather than a comprehensive education in food safety, thus reinforcing the perception that food safety practices are costly and not worth prioritizing.
“Tragedy of the Commons” showed how the inadequate regulation of informal food vendors can lead to increased incidence of FBDs when vendors decide to bypass safety measures, ultimately harming both consumers and the sellers.
These archetypes emphasize the underlying issues that prevent the effectiveness of surveillance efforts, suggesting areas where targeted, long-term interventions could disrupt these patterns.
Leverage points of the FBD systems map
Building on these archetypes, we identified several leverage points where targeted interventions could effectively disrupt the reinforcing loops that hinder the implementation of effective FBD surveillance systems in Africa. Deep leverage points included “Public trust in health interventions,” “Compliance with food safety practices,” and “Data sharing across sectors”. These points were classified as deep leverage points because they were determined to create important structural and behavioral changes in the FBD surveillance system.
Public trust is central to the feedback loops, and they influence compliance, reporting, and overall public cooperation, so changing the level of public trust would impact the public’s willingness to participate in surveillance initiatives and follow food safety guidelines. Similarly, compliance affects the spread of pathogens and the overall safety of food production, which means that targeting this point will affect the core behaviors around food handling and safety practices.
Improved data sharing will support coordination among healthcare workers, the public sector, and the agricultural sector, thereby enhancing a system-wide response. This point restructures the information flow across sectors, enabling collaboration between different sectors. These points affect the core paradigm around health interventions and public engagements.
These components suggest that improving the public understanding of safe food practices, acknowledging cultural norms and knowledge on sustainable agricultural practices, is crucial to ensure a system-wide change. In many African communities, traditional beliefs and practices often influence health behaviors, like reliance on traditional healers, who are sometimes trusted over modern medical practices67,68. This belief can lead to seeking remedies from traditional healers instead of professional healthcare providers, which complicates efforts to report, manage, and control FBD infections effectively. Additionally, there may be resistance to visiting hospitals due to mistrust in modern medicine, which is often influenced by past negative experiences, or the stigma associated with certain illnesses67. This emphasizes how the “limited education” element indicates a profound need for paradigmatic shift in how food safety knowledge is utilized and communicated. If the mindset is shifted towards awareness and prevention, it could lead to more effective implementation of food safety practices. This is essential for long-term sustainability and effectiveness, because they address the root causes of the FBD surveillance system and point to public health vulnerabilities.
Shallow leverage points included “Diagnostic capacity”, “Resource allocation for surveillance”, “Antimicrobial resistance control”, “Severity of foodborne disease”, and “FBD incidence”. These points can all help to provide important improvements within the existing system, but do not fundamentally alter the system’s goals, structures, or paradigms. They point to regulatory changes that need to happen to enhance facility equipment and optimize public health logistics, which are crucial for improving accessibility to medical services. While allocating funding is a straightforward starting point, it is essential to integrate other elements within the communities to ensure sustained efforts. Engaging local communities in recognizing FBDs can facilitate the establishment of surveillance strategies that capture general trends across diverse communities and help prioritize resources to areas most in need.
Scenario planning and implementation
Building on the leverage point analysis, we approached the scenario planning process through a series of steps, including the identification of key drivers of change, definition of scope, timeframe, and boundaries, as well as assessment of each scenario against performance measures. To clarify, key drivers were identified based on insights from the CLDs and the leverage point analyses. The following drivers were selected based on their influence on the FBD incidence, effectiveness of the system, and public engagement:
Public trust influences the public’s willingness to comply with food safety guidelines, report cases of FBDs, and engage with public health interventions. Trust levels fluctuate based on the transparency in health communication, visibility of legislation efforts such as rules or guidelines, and public perception of health authorities’ effectiveness. Scenarios were designed to test variations in the public trust to understand how changes in this driver could affect compliance and data sharing.
Compliance with food safety practices is directly related to the effectiveness of a successful FBD surveillance system, which is why compliance was identified as a key driver on the risk of contamination and, subsequently, the incidence of FBDs. Scenarios were designed to test the effects of high, medium, and low compliance across various settings, particularly in informal settings where legislation is often limited.
Effective data and information sharing practices between different sectors, such as public health, agriculture, and food, are very important for coordinating outbreak detection and response. It was identified as a driver since data sharing enables faster response times and collaboration. Scenarios were designed to explore different levels of investment in digital infrastructure and reporting to assess how improvements in data sharing could enhance system resilience.
Each scenario was constructed to simulate varying conditions based on the identified key drivers of change, within the scope, timeframe and boundaries defined as scope, timeframe and boundaries.
The scope focused on the three components identified in the leverage analysis, due to their direct impact on FBD incidence and system responsiveness. The scenarios simulated how shifts in each of these areas would influence the effectiveness of FBD surveillance.
The timeline was determined to be 10 years in order to capture both immediate and long-term effects. This timeframe should allow for observation of short-term responses like changes in FBD incidence following, e.g., a change in legislation, and long-term effects, like sustained improvements due to improved compliance and data sharing.
The boundaries were set to focus on policy-relevant factors affecting FBD surveillance, excluding anything beyond the control of the system (e.g., global warming and economic crisis). These boundaries ensured that the scenarios remained grounded in actionable policies within the FBD surveillance system.
Each scenario was then evaluated against effectiveness, adaptability, and resilience. An overview of the different scenarios and their impact on the three key drivers of change, as well as their effectiveness, adaptability, and resilience, is available in Table 1.
Through the evaluation of each scenario, scenario 5 was determined to be the most robust option for enhancing FBD surveillance in Africa. Scenario 5 combines high levels of public trust, compliance with food safety practices and data sharing across sectors, achieving optimal results across both the effectiveness, adaptability and resilience criteria. Although this scenario represents ideal conditions, it also presents some challenges, especially in African contexts where limited IT infrastructure and resources may restrict capabilities. Addressing these constraints through targeted investments in infrastructure and capacity building would further strengthen the implementation of scenario 5, making it feasible for a long-term impact.
