Community Resilience to Cholera Epidemics in Uganda: A Realist-informed review of Vulnerability and Intervention mechanisms.

Conceptual framework

This paper employs a realist evaluation conceptual framework by Pawson and Tilley (1997). The framework applies four main linked concepts for explaining and understanding programs or interventions: “Context”, “Mechanism”, “Outcome pattern” and “context- mechanism-outcome pattern configurations”. The concepts enable us to explain and understand what interventions work, for which communities, in what circumstances/respects and why¹². According to Pawson and Tilley [1997], “Context” describe the features of the conditions in which programs are introduced, that are relevant to the operation of the program mechanism. “Mechanisms” describe what it is about programs or interventions that can bring about any effect^12,13. “Outcome patterns” comprise the intended and unintended consequences of programs or interventions resulting from the activation of the different mechanisms in different contexts. Lastly, “context-mechanism-outcome pattern configurations” describe how different components of a program need to be harmonised to bring about alterations in behavioural/ event/ state regularities in different contexts¹².

This paper focuses on cholera outbreaks in Uganda, the different contexts in which cholera emerges, various cholera related interventions and their intended or unintended effects thereof. Communities where cholera is recurrent are exposed to repetitive effects arising from outbreaks. International efforts have recently targeted cholera elimination through deployment of the cholera vaccine¹⁴ but the recurrency of outbreaks has remained high especially among communities that have received the vaccination⁶. How these communities cope with the shocks and the extent resilience is developed is the subject of national health security and a public health concern.

Furthermore, we utilise a broader theoretical framework by Blanchet et al. [2017] to help us interpret Outcome patterns that lead to broader resilience of communities¹⁵. Blanchet et al. [2017] draw attention to four community level capabilities/ mechanisms that drive community resilience to shocks. These include mechanisms that; 1) generate knowledge capacity – combine and integrate different knowledge systems; 2) generate capacity for interdependency – handle multiple and dynamic engagements; 3) build legitimacy – capacity to develop socially acceptable norms and institutions; and 4) deal with uncertainty – capacity to anticipate and cope with unplanned events. These concepts are applied to the literature on cholera outbreak management and control in Uganda. Blanchet et al. [2017] conclude that the mentioned capabilities or mechanisms can lead to three main categories of resilience to shocks – absorptive, adaptive, and transformative capacities.

Patient and Public Involvement(PPI)

This was a review article where the main source of data was secondary documents, patients and the public were not involved in any way.

Information sources and search

Electronic searches were done across three electronic databases (PubMed, EBSCO, and Scopus). Searches were also done on the Ministry of Health(MoH) website. Searches were first conducted in 2022 and then updated in January 2024. Purposive search for reports about cholera and resilience were also done using Google search engine. In our search the following search syntax was used:

Furthermore, we identified kinship literature (N=18) to help us understand the broader concepts related to community resilience. The kinship papers were identified based on program theories from Non-government Organizations (NGOs) and government Ministries, Departments and Agencies (MDAS) that informed the programing and evaluations about Uganda and beyond. Types of kinship literature included grey literature like social action and interventions reports and/or program evaluations.

Inclusion and exclusion criteria

The initial search was not restricted by year of publication but was limited to title and abstract only in each database to ascertain eligibility. After screening the titles and abstracts, we conducted a secondary search to retrieve full paper/articles. JN, AAK and HS read and extracted data from the eligible/included studies.

The inclusion criteria for the papers to review were that they are; 1) original studies, 2) peer reviewed, 3) accessible without cost, 4) in English language, 5) included information about cholera, community engagement, health systems and resilience and 6) in Uganda and East Africa. However, for kinship literature, documents reviewed were beyond Uganda but perceived rich in conceptual explanations for broader resilience.

The search results were exported to a reference manager (Endnote X9) to check for duplicates. After removing duplicates, we identified a total of 123 publications for screening after removing duplicates. Initially, JN and AAK screened all titles and abstracts, and a random sample (10%) was checked by FS. The full text of the eligible studies was then retrieved and reviewed by JN, AAK, HS, RT and FS. Where there was disagreement, the team discussed agreeing on a consensus. After screening 31 studies met the inclusion criteria for data extraction and synthesis. Excluded studies were due to several reasons including lack of access to full text, not including resilience and community engagement, does not include Uganda or East Africa. Figure 1 illustrates the study identification, screening and inclusion.

Figure 1. Flow Diagram of Study Identification, Screening, and Inclusion in the Realist-Informed Review.

This figure outlines the PRISMA-style process used to identify, screen, and select studies for inclusion in the review. It details the number of records retrieved, screened, excluded, and finally synthesized, highlighting the systematic and transparent methodology employed.

Data extraction and synthesis

We conducted data extraction and analysis following a six-stage process (Figure 2tiff). We were guided by the realist evaluation framework and related concepts of “Context”, “Mechanisms”, “Outcomes and configurations to extract into a matrix, organize and present findings from the literature as well as synthesize and make logic of interventions and related outcomes.

Figure 2. Six-Step Process for Data Extraction and Synthesis.

This diagram illustrates the realist-informed six-step approach adopted for extracting, synthesizing, and interpreting evidence. The process includes: identifying sources, coding, theme generation, configuring context-mechanism-outcome patterns, refining theory, and validating findings. Abbreviations: CMO – Context, Mechanism, Outcome; CMOC – Context–Mechanism–Outcome Configuration.

At the first stage, the team developed an excel matrix to collect the following information: Author, Year Published, Title, Country Focus, Health/Environmental focus, Methods, Information on Context, Information on Mechanisms, and Information on Outcomes (CMO). Thereafter the team extracted broader information around the attributes above, with a lot more detail relating to Context, Mechanisms, and Outcome (CMO) in relation to Cholera (see annex 1). A detailed excel sheet is attached as supplementary information.

Table 1 summarizes what was done at the second and third stage. The second stage involved a summary of stage-1 information, with specific focus to synthesis of data related to CMOs. For each of the knowledge products under review, we synthesized contextual or setting issues, mechanisms in form of vulnerabilities and interventions that worsened or reduced the risk to cholera. At the third stage, we collated data about similar CMOs emerging across all the literature in the excel file and created a fourth column for references with similar or related information. Mechanisms leading to vulnerabilities (negative outcome) were indicated with a “V”- i.e. M1(V) and mechanisms leading to resilience (positive outcome) were indicated as “R” - i.e. M4(R).

The fourth and fifth stages involved a synthesis of all Context, Mechanisms, Outcomes, Configurations (CMOCs) and development of a narrative for the general midrange theory of change along the CMOCs (Table 2). The CMOCs were then shared with overall consortium during a literature review workshop to generate additional comments. The sixth stage involved bundling of CMOCs into resilience and vulnerability mechanisms. Lastly, we held a meeting to discuss and agree on the final CMOCs and then developed a general Cholera resilience theory of change/ causal framework (Figure 3). Figure 3 indicates plausible pathways to various forms of resilience for the main mechanisms that we extracted from the literature.

Figure 3. Pathways to Community Resilience Against Cholera Epidemics in Uganda.

This conceptual model depicts how various contextual factors, mechanisms, and outcomes interact to enhance or constrain community resilience to cholera outbreaks. It synthesizes findings from the review and maps triggers such as social cohesion, local leadership, and community health systems. Mechanisms are categorized into community-led actions and externally supported interventions. Dotted lines represent areas where further evidence is needed to establish direct links between mechanisms and broader resilience capabilities. The framework is guided by Blanchet et al. (2017)’s typology of resilience and Pawson and Tilley’s realist evaluation model. Abbreviations: CMOC – Context–Mechanism–Outcome Configuration.

Vulnerability mechanisms to cholera epidemics in Uganda

We found a sizable body of epidemiological literature that was focused on the vulnerability mechanisms – factors responsible for the spread or transmission of Cholera^2,16–22. Maps developed by Bwire et al. (2023) indicate that Uganda has several districts that have been confirmed as hotspots for cholera epidemics between 2011–2021⁶. According to the Ministry of Health, districts that have experienced cholera outbreaks include; Kasese, Rukungiri, Bulisa, Nebbi, Mbale, Hoima, Arua, Zombo, Bududa, Butaleja, Sironko, Manafwa, Ntoroko, Busia, Namayingo, Moyo, Bulambuli and Kapchorwa^5,19.

Communities that are at great risk of cholera include fishing communities, peri-urban slum dwellers, border communities, migratory farm workers, landslide and flood victims, refugees, prisoners and the mentally ill^3,16. This implies that the location of communities and districts increases the likelihood of transmission and cholera outbreak. The highly at-risk communities are characterized by weak WASH infrastructure such as lack of latrines and existence of open water sources. Specific to communities in land slide and flood prone communities namely Butaleja district, there are poor sanitation conditions with latrines filling up with water or inability to construct latrines due to the high-water tables and consequently faecal matter spilling over to the open water sources^3,16,17,19.

Community poor sanitation practices such as open defecation, use of leaves and sand as anal cleansing material and lack of hand washing was found to increase the susceptibility to cholera transmission and spread^7,20. Furthermore, the vulnerability of such communities is aggravated by the limited capacity of Uganda’s health care system to implement early warning systems to enhance detection of cholera outbreaks¹⁸. There are also limited capacities among community health workers. Despite being the first responders and focal contact persons for existing community level early warning systems, they have not been sufficiently trained in effective cholera epidemic surveillance, investigation, response and reporting²⁰.

Preventive resilience mechanisms against Cholera

Training of formal health workers

In the context of weak health system capacity, where early detection and response was perceived as weak, low numbers of health workers was a main factor for communities to experience high morbidity mortality from cholera and diarrheal diseases^5,19. Studies that sought to build coping capabilities against cholera outbreaks implicitly and explicitly tried to improve surveillance, lessen the time needed to effectively respond to the outbreak^3,5,20,23. For both these outcomes, health workers were increased in numbers, skills or provided with tools as the central mechanisms. Training upstream health workers (epidemiologist and laboratory technicians) in outbreak detection and response and epidemic management^3,24, were also important components for this mechanism.

Training of Community Health Workers (CHWs)

In the context of weak health system capacity and few formally trained health workers, substitutes in the form of CHWs were engaged to assist in the roles of disease surveillance and response^3,18,24–26. CHW’s embeddedness in the community was seen as helpful and more culturally sensitive to mobilise and support health promotion in communities during disease outbreaks^9,16,25,27. For this reason, CHWs were seen as a helpful mechanism in mitigating the shocks arising from cholera and other epidemics.

Community knowledge

Specifically, residual knowledge from prior experience about cholera was reported as a vital asset for adaptive resilience^7,18,27. This asset, if well managed can help communities to take appropriate actions in time to mitigate the epidemics. This was linked to transformative resilience in the typology above. Lived experiences and the recall and re-enactment of the WASH behaviours and practices to ameliorating disease outbreaks such as cholera was a major mechanism in the community resilience pathway²².

Health promotion of WASH practices

WASH practices such as washing hands with soap before and after food, boiling or chlorinating drinking water, treatment of open water sources and good toilet practices especially among groups at high risk – refugees, fishing communities, IDPs and slum settlements – is likely to improve WASH practices and reduce outbreaks/recurrence^{3,5,7,16,17,22,27,28}. Community action towards WASH interventions was expected to reduce vulnerability to transmission as well as outbreaks/recurrence. Peer-to-peer engagement, community solidarity/bonding was reported in a few studies as vital for engineering desirable or positive norms and practices such as personal and homestead hygiene^7,9.

Community cohesion

Community cohesion was also found to have a high likelihood of improving adaptive and transformative means to reduce vulnerability to cholera effects^3,5,7,21. Two mechanisms related to community cohesion seem to operate through engagements, bonding and linking types of social actions – broadly terms as social capital^29–32. Under the bonding type, internally the community can form close solidarity to work together⁷ and (ii), the linking type enables external engagement, where community can partner/connect with external entities to get support to build infrastructure and other WASH facilities^22,27,28. As Observed by Mayhew et al. (2021), construction of infrastructure (e.g. Toilets, hand washing facilities, water source protection) through community engagement and peer-to-peer support is likely to reduce outbreak risk⁹. Furthermore, formation of peer-to-peer groups was identified as a mechanism of improve WASH infrastructure and practices that required joint actions such as protecting water sources, building toilets, and creating awareness about sanitation and safe disposal of waste. Peer- to- peer groups were more likely to create or enhance adaptive measures that help reduce risk of outbreaks especially among refugee settlements and areas prone to landslides^7,9,19,22. Working with community health workers also hinged on their social embeddedness to the communities they serve.

Mechanisms from literature related to broader resilience as a concept

As earlier noted, kinship literature was added to enhance our understanding of broader resilience as well as supporting how to make logic of the outcomes relating to interventions in relation to Blanchet et al. (2017)’s resilience capacities; absorptive, adoptive and transformative. Majority of the literature focused on several interventions including; livelihoods and economic empowerment, community resilience to infectious diseases and shocks, pandemic preparedness of community health workers, governance and capacity for managing health system resilience, realistic evaluation and guiding frameworks for chorea interventions in Uganda.

Below, we summarize take away messages from kinship literature.

Social safety nets³³ in form of; financial or asset transfers like Send-A-Cow (SAC) project, Social Assistance Grant for Empowerment (SAGE) programme, Northern Uganda Social Action Fund (NUSAF) especially in the context of good community social capital and group efficacy, can improve resilience of vulnerable communities at risk of cholera and other shocks¹¹. In this literature, resilience was defined as the ability to bounce back or thrive in case of occurrence of disasters^10,11,15. However, corruption, poor financial management, mistrust, delays of disbursement of financial resources, and the lack of social acceptability, the mechanism may not be sustained. In other cases, social safety nets were also said to encourage dependency on external hand-outs instead of sovereign community sustenance^34,35.

Coordination, cooperation, joint community actions and peer monitoring are positive social mechanisms that arise from social capital and social networks. These can help to instill positive social norms and trust^31,32. This literature is more about the difficulty of dealing with epidemics including cholera in situations where communities are consumed by rivalries, mistrust, and conflicts. The literature also observe that the mission at the core of the social network, if not positive (e.g. tribal and ethnic rivalries) can weaken resilience and bring about more vulnerability²⁸.

Health system transformation such as decentralization can affect the performance of frontline health workers in their preparedness for epidemic control³⁶. The granular mechanisms relate to recruitment and retention of health workers in decentralized settings.

Prior mechanisms are necessary to build resilience overtime and these are, planning with broader goals including Sustainable Development Goals (SDGs)³⁷, and everyday planning i.e. change management for health systems resilience³⁶.

Having the knowledge and technology to detect, plan and respond to crises/stressors calls for flexibility in institutional norms and practices. That is, ability to change quickly as and when it is necessary³⁶.

The multi-dimensionality of crisis requires many capabilities/whole of society action/co-creation/multi-stakeholder action/coalitions. There is need for smart dependency/up and downstream linkages and support^3,5,22,38. Again, this aligns with the importance of bridging types of social capital to achieve transformative resilience.

Plausible pathways to the various forms of Resilience to cholera epidemics in Uganda

Drawing from the findings indicated in 3.1 and 3.2, the team developed a general Cholera resilience conceptual model indicating plausible pathways to various forms of resilience (Figure 3). This was guided by capacities for resilience management¹⁵, that are absorptive capacity, adaptive capacity, and transformative capacity as elaborated by Blanchet, Nam et al. (2017). This typology of vulnerability is also acknowledged by studies related to health systems resilience to disease outbreaks⁹. Overall, the realist-informed review found five mechanisms that enhance resilience against cholera outbreaks in Uganda, and these are enabled by either external support or community action. There were two absorptive mechanisms against vulnerability to cholera, one adaptive mechanism and one transformative mechanism. However, there was also one mechanism that consisted of both adaptive and transformative features. The relationships between the various mechanisms are very interlinked. This implies that to achieve a particular resilience outcome, a combination of mechanisms is necessary. The last columns are the ultimate resilience outcomes and capacities as highlighted by Blanchet et al. (2017). The relationship between the intermediate outcomes and ultimate resilience was not possible to demonstrate since most retrieved papers were less inclined to link to these ultimate resilience measurements but had custom-made intermediate outcomes that we found useful in depicting the pathways for change. The mechanisms we identified were grouped into two main categories, 1) external support – arising outside of the communities and 2) community action – endogenously derived from the community.

Study limitations

The exclusion of literature on interventions and surveys that are not in Uganda and not focused on cholera represents a limitation as this could have enriched or diversified our synthesis of published literature on the subject matter. However, the strength in narrowing the focus and context was to ensure in-depth review of the issue in the country and provide some preliminary lessons as well as insights on broader community resilience in the event of health shocks, using cholera as an example.