Exposure behaviour to Escherichia coli among households in Imvepi refugee settlement, Terego district Uganda

Abstract

Introduction

Exposure to Escherichia coli (E. coli) is a risk factor for diarrhoeal diseases, which pose a significant problem in refugee settlements. Refugee populations are exposed to faecal microorganisms through multiple pathways including sub-optimal sanitary facilities, contaminated drinking water, produce and food, flood water, bathing water, and soil among others. While these pathways are well-documented, specific exposure behaviours remain underexplored. We assessed exposure behaviour to E. coli among households in Imvepi refugee settlement, Uganda, and provided evidence-based recommendations for the design of interventions to reduce excreta-related disease in refugee settlements.

Methods

Guided by the Sanitation Safety Planning approach, we surveyed 426 households in Imvepi refugee settlement, Uganda, using a digitized questionnaire and an observation checklist. We collected data on the background characteristics and exposure behaviour of women and emancipated girls (minors living on their own, having borne a child, married, or pregnant). The outcome variable, E. coli exposure behaviour, was measured using a five-point Likert scale, assessing behaviours that increase the risk of exposure. Data were cleaned in Microsoft Excel and analyzed in Stata version 17. Descriptive statistics were performed to summarize the data. We used modified Poisson regression to determine the factors associated with the outcome.

Results

Over 59.4% (253) exhibited high-risk exposure behaviour. Residing in compound homes (Adjusted Prevalence Ratio (APR) = 0.72, 95% Confidence interval (CI): 0.58–0.90), being aged 35–49 years (APR = 0.76, 95% CI: 0.60–0.97), having household heads with post-primary education (APR = 0.54, 95% CI: 0.38–0.77), high knowledge (APR = 0.69, 95% CI: 0.59–0.80), and high-risk perceptions regarding exposure to E. coli (APR = 0.75, 95% CI: 0.64–0.88) were associated with a lower prevalence of high-risk E. coli exposure behaviours. Conversely, having sanitary facilities with excreta overflowing from the squat hole (APR = 1.26, 95% CI: 1.08–1.48) was associated with a higher prevalence of high-risk exposure behaviours.

Conclusion

The study indicates a substantial prevalence of high-risk E. coli exposure behaviours in the refugee settlement.. There’s a need to implement behaviour change interventions targeted at preventing or minimizing exposure, especially among households whose heads have low education attainment, those with young caretakers and those with limited knowledge and low-risk perceptions regarding exposure to E. coli.

Introduction

Significant progress has been made in improving access to Water, Sanitation and Hygiene (WASH). However, households remain underserved. An estimated 2.2 billion people still lack access to safe drinking water, 3 billion do not have hand washing facilities with soap and 3.6 billion still rely on shared or unimproved facilities [1]. Refugee settlements are no exception to the global WASH crisis [2, 3] since households still fall short of the post-emergency target of at least 85% sanitation coverage and 95% of households with access to soap [1]. Recent evidence from a study conducted in 2022 in 21 refugee camps in sub-Saharan Africa (SSA) indicated that only 24% of households had access to basic hand hygiene, 30% had access to basic sanitation and 5% practised open defecation [3]. In Uganda, the household latrine coverage in refugee settlements stands at only 69% [4].

The inadequacies of WASH in refugee settlements are exacerbated by limited resources, poor infrastructure and unplanned rapid population growth [5,6,7]. The rapid population growth in refugee settlements is often not commensurate with the provision of social services and appropriate infrastructure including housing, solid waste management, water supply, and excreta and wastewater treatment among others [7, 8]. The limited access to WASH services exposes refugee populations to infectious microorganisms from faecal matter, including pathogenic strains of Escherichia coli (E. coli). E. coli is associated with gastrointestinal infections, which are known causes of morbidity and mortality, especially in tropical regions [9,10,11]. Although data specific to refugee settings are lacking, exposure to E.coli through the use of unimproved sanitation facilities is estimated to cause over 564,000 diarrhoeal disease-related deaths globally [12]. The incidence of diarrhoeal diseases is also known to be higher in refugee settings, where it causes 40% and 80% of deaths among all refugees and children under two years respectively [6]. Irrespective of the setting, diarrhoeal diseases are also a major reason for outpatient visits and hospitalizations in the global south [13,14,15].

To address the negative consequences of WASH on refugee health, humanitarian actors developed the Sphere minimum standards for humanitarian response. According to the Sphere standards, there should be a minimum of one shared toilet per 20 people, located within 50 m of each household to reduce open defecation [16]. Aside, each refugee household must have two water containers (10–20 L), and soap and water at a hand washing station [16]. Through the United Nations (UN) plan of action for people, planet and prosperity, the UN member states committed to work towards universal coverage of WASH services (SDG 6), good health and wellbeing (SDG 3), and reducing inequalities (SDG 10) [17, 18]. Despite these commitments and the availability of sphere standards, sanitation services in refugee settings remain suboptimal. The ratio of toilets to users in humanitarian settings remains higher than that prescribed by the Sphere standards [19].

While the majority of refugee populations rely on the use of shared sanitary facilities, which represents a step on the sanitation ladder, this arrangement still presents challenges and increases the potential exposure to E. coli due to the suboptimal maintenance and hygiene conditions of these shared facilities [20]. Refugee populations are also exposed to faecal microorganisms through the ingestion or contact with contaminated drinking water [21], bathing water, surface water, open drain water, flood water, contaminated produce and street food [22,23,24]. Although faecal exposure pathways at the household level are documented, our understanding of the specific exposure behaviours remains limited. Existing household-level studies in refugee settlements primarily focus on drinking water as an exposure pathway [21] and have not adequately explored exposure behaviours beyond fecal-oral transmission [21, 25]. This study utilized the Sanitation Safety Planning (SSP) tool to assess exposure behaviour to E. coli in households in Imvepi refugee settlement, focusing specifically on containment within the sanitation chain. E. coli is an indicator organism for fecal contamination [26, 27], and is a widely accepted proxy for assessing microbial exposure pathways and behaviours [28, 29].

The SSP tool is a risk-based management tool for the systematic identification and management of health risks and exposure pathways along the sanitation chain [30]. This tool helps in identifying and prioritizing risk mitigation measures to improve the safety of sanitation systems and protect public health [30]. The SSP has previously been applied in assessing exposure behaviours among sanitation workers [31,32,33]. While the SSP tool can assess risks across the entire sanitation chain [30, 34], its application at containment at household level remains limited. Nonetheless, assessment of behaviours at containment can generate evidence of exposure risks and intervention points that are critical for reducing the microbial transmission of pathogens within a humanitarian setting. Additionally, containment impacts subsequent steps including emptying, transport, treatment, and final disposal or productive use, that would otherwise become much more challenging for sanitation workers, the general public and the environment [31,32,33, 35, 36]. For instance, poorly designed containment systems may require manual cleaning or specialized equipment, leading to higher costs and inefficiency [31,32,33, 35, 36]. Furthermore, household behaviours, such as disposing of non-biodegradable materials and using harmful additives, can obstruct and damage emptying equipment, disrupt biological treatment processes, and introduce foreign materials that complicate waste treatment and disposal [31,32,33, 35, 36]. Evidence generated by this study can be used to inform household-level behaviour change interventions.

Materials and methods

Study design and setting

A cross-sectional study was conducted in September-October 2023 in Imvepi refugee settlement located in Terego district in the Northern region of Uganda. Imvepi refugee settlement was established in February 2017 to give room for South Sudanese asylum seekers fleeing the war in their country. The settlement is located in Odupi sub-county and is divided into four zones. Zone 1 has 12 villages, zone 2 has 20 villages, zone 3 has 9 villages, and zone 4 has one village. A village in this context refers to a defined administrative unit within the settlement, typically comprising a cluster of households organized under the leadership of a village chairperson. The settlement stretches over around 53 km² and currently hosts 70,200 refugees and 20,178 households, a vast majority of which are South Sudanese. The non-refugee population of Odupi sub-county is estimated at 45,300 [37]. Imvepi refugee and host communities are rural communities, with no major urban centers. With a sanitation coverage of 78% [37], the entire population in the Imvepi refugee settlement largely depends on traditional unlined pit latrines that are abandoned when they are full [38]. Despite this relatively high sanitation coverage, the prevalence of sanitation-related diseases remains high in Odupi sub-county, where Imvepi refugee settlement lies [39]. This study was conducted in Zones 1,2 and 3. Zone 4 was initially included in the pretest, and thus excluded in the main data collection phase of our study.

Study population and eligibility criteria

The study population included women in households with access to a sanitation facility. Girls aged between 15 and 17 years in child-headed households with sanitation facilities were also included since they were considered emancipated minors. Only households with access to sanitation facilities (private or shared) were selected because the study aimed at assessing the exposure behaviours associated with the containment of faecal matter. Women and girls were purposively selected due to their elevated risk of microbial exposure from the different household environmental pathways since they spend a considerable amount of time at home, and bear the primary responsibility of fetching water and cleaning and maintenance of household compounds, indoor environments and sanitary facilities [40, 41]. Women who were absent and those who were sick at the time of the study were excluded.

Sample size estimation

The Kish Leslie formula was used to estimate the sample size [42] as shown below;

$$\:\text{n=}\left[\frac{{Z}^{2\:}\text{p}\left(\text{1-p}\right)}{{\delta}^{\text{2}}}\right]$$

The prevalence (p) of high-risk E. coli exposure behaviour was estimated at 50% due to limited literature in similar settings [43]. A two-sided Z statistic corresponding to a 95% confidence interval (1.96), and a precision of 5% yielded a sample size of 385 respondents. Considering a 10% non-response rate, a sample size of 428 households were considered. The sample was distributed proportionate to size (Table 1).

Table 1 Sample size allocation by zone

Sampling procedure

A three-stage cluster sampling method was used (Fig. 2). During the first stage, we purposively selected zones 1–3 to ensure geographic representation of the settlement. The second stage involved the random selection of 3 villages from each cluster. To determine the number of households to select per zone, probability proportionate to population size of the zone was used (Table 1). During the third stage, a Microsoft excel randomizer was used to generate a random sample from a list of households (obtained from Office of the Prime Minister) per village. At each selected household, women or emancipated girls (main caretakers) at home were interviewed (Fig. 2).

Fig. 1

Map showing Imvepi refugee settlement

Fig. 2

A flow chart depicting the sampling technique

Data collection techniques, and quality assurance and control

A structured questionnaire preloaded on the Kobocollect mobile application was administered through face-to-face interviews. The tool obtained data on respondent demographics, household sanitation infrastructure, and knowledge and perceptions of the health risks associated with sanitation facilities. Data were also collected on the frequency of behaviour related to exposure to E. coli. A recall period of two weeks was used to reduce recall bias. During interviews, only one individual was interviewed per selected household, and the sanitary facility used by the household was inspected. In cases where multiple households were sharing the same facility, only the individuals from the selected households were interviewed, but the facility itself was still inspected. The questionnaire used in this study was newly developed based on the exposure routes recommended for consideration in sanitation safety planning [30, 44]. To ensure its validity, the questionnaire underwent review by a panel of five experts in environmental health and sanitation-related decision-making tools. The experts focused on assessing the relevance of each question to the study objectives, clarity, comprehensiveness, cultural appropriateness and redundancy. Furthermore, the tool was pilot-tested in zone 4 of Imvepi refugee settlement (not part of the study). An English-language version of the complete questionnaire has been provided as a supplementary file (Supplementary File 1).

An observation checklist was used to quantify the potential risk of exposure to faecal matter. Development of the checklist was based on the WHO SSP tool which proposes simple sanitary inspection as one of the approaches for assessing hazardous events in the sanitation system [30]. Research assistants observed the availability and functionality of a hand hygiene facility, cleanliness of the sanitary facility, status/condition of the facility, presence of anal cleansing materials, flies, odour, spillages potential for insect entry, and signs of excreta overflow from the latrine pit. The use of observation methods also minimized social desirability bias [45].

Data were collected by research assistants with a background in environmental health science research and with a minimum of three years of experience. Research assistants underwent a three-day training to familiarize themselves with the study protocol, standard operating procedures, and ethical considerations. Data collection tools were translated (into Juba Arabic) and rigorous supervision of the data collection process was conducted to ensure data quality. The data collection tools were also designed with skips and restrictions to minimize errors, and daily data reviews were carried out to detect and correct any missing or inconsistent values.

Variable measurement

Outcome variable

The outcome variable was exposure behaviour to E. coli. This was assessed by using self-reports on behaviours that predispose individuals to E. coli. These included: (1) Hand washing after visiting the toilet, (2) cleaning of sanitary facilities, (3) contact with overflowing/leaking contents from sanitary facilities, (4) contact with facility floors without shoes, (5) contact with flies and other vectors, and (6) consumption of contaminated water and consumption of contaminated produce i.e., plants on land irrigated or fertilized with a sanitation product. The exposure-related behaviours were based on the exposure routes recommended for consideration in sanitation safety planning [30, 44].

The outcome variable was measured as a composite informed by scores of a five-point Likert scale ranging from 0 for ‘Never’, to 4 for ‘Always’. We summed up individual scores to generate total scores. Since this study aimed at classifying exposure behaviour as low or high, a median score was used as a cut-off. Respondents who scored above or equal to the median were classified as exhibiting “High-Risk Exposure Behaviour” (assigned a value of 1), indicating more frequent engagement in behaviours associated with higher exposure risk. Those who scored below the median were categorized as having “Low-Risk Exposure Behaviour” (assigned a value of 0), indicating less frequent engagement in these behaviours. Median has previously been used as a cut-off in measuring WASH-related behaviour [46].

Independent variables

Based on available literature [47,48,49,50,51,52,53,54], the independent variables included; (1) individual factors such as age, highest education attainment, socioeconomic status, presence of children in a household, household size, risk perception, and individual knowledge of potential risks associated with sanitation facilities; (2) sanitation facility-related factors such as type of facility, design of the facility, cleanliness of the facility, structural integrity, number of users, and presence of anal cleansing materials; and (3) environmental factors such as access to safe water.

Respondents’ risk perceptions were assessed based on constructs of the Health Belief Model i.e. perceived susceptibility, perceived severity, perceived benefits, perceived barriers, self-efficacy, and cues to action. A 5-point Likert scale (Very unlikely = 1, Unlikely = 2, Moderate = 3, Likely = 4, Very likely = 5) was used for rating individual risk perceptions. During analysis, these ratings were collapsed into two categories i.e. ‘1’ for desirable responses and ‘0’ for negative responses. Ratings of “Very unlikely” (1) or “Unlikely” (2) were considered negative responses and coded as 0 whereas, ratings of “Moderate” (3), “Likely” (4), or “Very likely” (5) were considered desirable responses and coded as 1. Total scores were calculated, and the median score was used as a cutoff to categorize risk perception into Low and High. Respondents with scores below the median were considered to have a low-risk perception, while those with scores above the median were considered to have a high-risk perception. Previous studies have used the median score as a cut-off in assessing attitudes [55].

We employed Principal Component Analysis (PCA) in assessing the household socioeconomic status. This statistical technique is widely used to derive a single index from multiple correlated variables, such as household amenities and possessions (e.g., electricity, mobile phones, televisions, etc.) [56, 57]. PCA helps in reducing dimensionality while retaining most of the variation present in the data. In our analysis, each listed item (electricity, working radio, television, mobile phone, bicycle, motorcycle or motor scooter, and plot of land) was coded as a binary variable (1 for possession and 0 for non-possession). We retained the first eight principal components based on the Kaiser-Guttman rule, which states that components with eigenvalues greater than 1 should be retained [58, 59]. Additionally, we used the Scree test to visually assess the point at which the curve starts to flatten, which indicated the optimal number of components to retain [58, 59]. The Scree plot further supported the retention of the first 8 components as the eigenvalues dropped sharply after the eighth component. These components were used to form the SES index. The index was then divided into tertiles, representing ‘Low’, ‘Medium’, and ‘High’ SES categories.

Regarding knowledge, a set of 7 questions were designed to assess the respondent’s understanding of the risks associated with sanitation facilities and the potential health consequences of exposure to microbial hazards. Each question was coded as 1 if the answer was correct and 0 if it was wrong. We summed the scores from all seven questions for each respondent, and higher scores (equal to or greater than the median) indicated a higher knowledge level. Respondents who scored below the median were considered to have a low knowledge level. better understanding and awareness.

Data analysis and management

Data were entered and uploaded to the server using the KoboCollect mobile application. Afterwards, data were downloaded into Microsoft Excel, cleaned, and analyzed using STATA version 17.0. Descriptive statistics such as mean and standard deviation (SD) were used to present age, duration as a refugee, and household size, while frequencies and proportions were used to present categorical data. A “Modified” Poisson Regression model was run to establish the risk factors associated with exposure behaviour to E. coli. Variables with a p-value < 0.2 at bivariate analysis were selected for inclusion in the model. The multicollinearity effect was assessed with a cutoff point of variation inflation factor (VIF) of greater than 10 [60, 61]. Thereafter, model building was done by using backward elimination. A p-value < 0.05 was considered statistically significant.

Results

Individual characteristics of the respondents

A total of 426 respondents were interviewed representing a response rate of 99.5%. The mean duration as a refugee was 5.9 years (SD ± 1.2). The average age of respondents was 35.1 ± 13.3 years. About a third, 46.9% (200/426) belonged to the 35 years and above age group. The average household size was 8.0 ± 3.5 members. More than half, 54.5% (232/426) of the household heads had completed primary education, while 61.7% (263/426) of the participants had completed primary education. The primary source of water was public tap stands for 98.1% (418/426) of the households. Overall, 55.6% (236/426) had a high level of knowledge while 53.3% (227/426) had high-risk perceptions of exposure to E. coli. (Table 2).

Table 2 Individual characteristics of the respondents in Imvepi Refugee settlement

Sanitary facility-related factors

Over half, 54.0% (230/426) of households shared their sanitary facility with other households. Based on observations, the majority, 81.9% (349/426) used pit latrines with slabs, and 3.3% (14/426) used Ventilated Improved Pit Latrines. Nearly half, 42.5% (181/426) had sanitary facilities with absent or incomplete superstructures. Approximately, 58.5% (249/426) had anal cleansing material absent or inappropriate for the technology. Majority, 70.2%, (299/426) lacked handwashing facilities inside or next to the toilet (Table 3).

Table 3 Sanitary facility characteristics in Imvepi Refugee settlement, Terego district

Exposure behaviour to E. Coli in households in Imvepi Refugee settlement

Overall, more than half, 59.4% (95% CI [54.6–64.1]) of the participants exhibited high-risk E. coli exposure behaviour. Regarding direct contact with faecal matter, more than a tenth, 15.5% (66/426) reported touching faecal matter with their bare hands almost every day in the past two weeks. About half, 50.0%, (213/426) reported that faecal matter had splashed on their skin and 13.6% (58/426) always walked barefoot or touched soil near their sanitary facility within the past two weeks. More than a third, 36.6% (156/426) reported that they always consumed unboiled/untreated water from a water source within 30 m of a sanitary facility (Table 4).

Table 4 Exposure behaviour to E. Coli among households in Imvepi Refugee settlement

Factors associated with E. Coli exposure behaviours among households

At multivariate analysis, respondents living in compound homes had a 28% lower prevalence of high-risk exposure behaviour (Adjusted Prevalence Ratio (APR) = 0.72, 95% Confidence interval (CI): 0.58–0.90) compared to those in single-family homes. Those aged 35–49 years had a 24% lower prevalenceof high-risk exposure behaviour (APR = 0.76, 95% CI: 0.60–0.97) compared to those aged 15–19 years. The prevalence of high-risk exposure behaviour was 46% lower among households led by a head with post-primary education (APR = 0.54, 95% CI: 0.38–0.77) when compared to households where the head had no formal education. Using surface water as the main source was associated with an 84% higher prevalence of high-risk exposure behaviour (APR = 1.84, 95% CI: 1.34–2.54) compared to using tap water. Respondents with high knowledge of exposure to E. coli had a 31% lower prevalence high-risk exposure behaviour (APR = 0.69, 95% CI: 0.59–0.80) compared to those with lower knowledge. Those with high-risk perceptions had a 25% lower prevalence of high-risk exposure behaviour (APR = 0.75, 95% CI: 0.64–0.88) compared to those with low-risk perceptions. Having a sanitary facility with overflowing excreta from the squat hole was associated with a 26% higher prevalence of high-risk exposure behaviour compared to their counterparts (APR = 1.26, 95% CI: 1.08–1.48) (Table 5).

Table 5 Factors associated with E. Coli exposure behaviours among households in Imvepi

Discussion

This study assessed exposure behaviour to E. coli and associated factors in households in Imvepi refugee settlement. We found that more than half of the participants exhibited high-risk E. coli exposure behaviour. High-risk exposure behaviours were significantly associated with the type of home the respondent was living in, the age of the caretaker, the highest education attainment of the household head, the main source of water, knowledge of exposure to E. coli, risk perceptions of exposure, and having sanitary facilities with excreta overflowing from the squat hole. These findings highlight a need for targeted interventions aimed at reducing high-risk exposure behaviours in refugee settlements.

The high-risk exposure behaviours observed in this study can largely be attributed to several factors including knowledge of exposure to E. coli and risk perceptions of exposure, which were associated with the outcome. Our study findings suggest that household members are at a high risk of exposure to E. coli due to the high-risk exposure behaviours exhibited by their primary caretakers. Thus, there is a need for interventions focused on health educating caretakers about exposure pathways and behaviours, so as to minimize risk. These findings corroborate those reported in previous studies that indicated a high prevalence of behaviours that increased the risk of exposure to sanitation-related microorganisms in resource-constrained settings, like refugee settlements [3, 25]. For example, a study conducted among female heads of households in three refugee camps in Malawi revealed that 54% did not observe hand washing with soap after toilet use [25]. However, comparing these findings with studies from similar settings is challenging because existing research at household level in refugee settlements has not consistently assessed exposure routes beyond faecal-oral transmission. This study’s assessment of various exposure routes, including dermal contact and contact with leaking sanitary facility contents, makes it valuable.

The study revealed that respondents living in compound homes had a lower likelihood of exhibiting high-risk exposure behaviour to E. coli compared to those in single-family homes. Compound homes involve multiple households or families living in close proximity. This social arrangement could create a sense of community and shared responsibility. In such settings, individuals may be more likely to collectively mobilize for and address WASH issues, which could reduce high-risk exposure behaviours. Additionally, individuals within such homes may be influenced by the behaviour of their peers living in the same compound, which motivates them to adopt good practices. The findings concur with those reported in previous studies which have shown peer influences to be significant drivers of household WASH behaviours [62, 63].

However, these findings are somewhat surprising, given that communal settings are often associated with mismanagement and higher risks of contamination, as noted in previous studies [64, 65]. The discrepancy might be explained by the nature of compound homes in our study, which typically involved households that shared familial or neighbourly bonds cultivated from staying in close proximity to one another as refugees for a long time. These bonds might differ from those in former studies where neighbors are often transient tenants. Our findings highlight the potential for hinging on the social dynamics within compound homes to improve exposure behaviours in refugee settlements.

The study also found that households led by a head with post-primary education had a lower likelihood of high-risk exposure behaviour when compared to households where the head had no formal education. This could be because household heads with above primary education have a better understanding of the risks associated with exposure to E. coli, which translates into good practices, including among the rest of the household members. Additionally, households where the head has higher education levels might also have better access to WASH-related resources, thereby impacting their behaviour. Existing studies have consistently shown that individuals with higher levels of education are more likely to practice better WASH-related behaviours [47,48,49, 51]. The study findings imply that education is an important determinant of WASH behaviours, and thus interventions should particularly target those with low education levels.

This study also found that respondents with a high knowledge level had a lower likelihood of exhibiting high-risk exposure behaviour compared to those with a low level of knowledge. Respondents with a high knowledge level are likely to be well-informed about the risks associated with E. coli exposure and are thus more inclined to adopt WASH practices. For example, they may choose to use treated or protected water sources, use footwear when using sanitation facilities, maintain proper sanitation facilities, and practice regular handwashing, among others. Moreover, those with a high knowledge level may be more likely to perceive themselves at risk of exposure, and thus take the necessary precautions. These findings corroborate those reported by Assefa and Kumie [66] in a cross-sectional study in Ethiopia which revealed that knowledge influenced hygiene behaviour status. These findings highlight the importance of health education initiatives aimed at improving awareness of E. coli-related health risks.

Additionally, the study found that women with high-risk perceptions had a lower likelihood of exhibiting high-risk exposure behaviour compared to those with low-risk perceptions. Individuals who have high-risk perceptions may be more likely to understand their susceptibility to infection and the potential severity of excreta-related diseases. Furthermore, they may recognize the benefits of practices, be more responsive to cues for action such as health education and have a higher perceived efficacy, which influences their likelihood of uptake of good practices. These findings concur with those reported in a similar cross-sectional study among mothers of under-fives which revealed that attitudes influenced hand hygiene practice [53]. Another study conducted in Rohingya camps in Bangladesh also found similar findings [52]. The study’s findings regarding the influence of risk perception on behaviour highlight the importance of not only raising awareness about health risks but also shaping individuals’ perceptions of these risks. Health promotion efforts should raise awareness with a target of increasing an individual’s perceived susceptibility, severity, benefits, and efficacy while reducing perceived benefits.

This study also indicated that having a sanitary facility whose excreta was overflowing from the squat hole was associated with a higher likelihood of exhibiting high-risk exposure behaviour compared to their counterparts. Prior research has highlighted how the condition of sanitation facilities can influence exposure behaviour. When facilities are poorly maintained or overflowing, individuals are more likely to have poor practices, including inadequate handwashing or improper waste disposal. In contrast, sanitary facilities that are well-maintained and functioning properly create an environment conducive for healthy practices. A study conducted in a rural setting in Ethiopia found an association between the presence of faeces in the compound and hand hygiene practices. Additionally, the study indicated that when using unimproved latrines, respondents’ post-defecation handwashing behaviour with soap was significantly reduced [54].

Study strengths and limitations

This study may be the first of its kind to utilize the SSP tool to assess exposure behaviour to E. coli in households within refugee settlements. This novelty is important because it fills a crucial knowledge gap in understanding microbial exposure pathways in such settings and can potentially inform appropriate interventions aimed at improving the health and well-being of refugees and similar communities. The study recruited a relatively large sample of households (426) which enhanced the generalizability of the study findings to similar populations. The study also used a combination of methods, e.g., behavioural assessments and observation which provided a better understanding of exposure risks. However, this study is prone to recall bias since exposure behaviour was assessed retrospectively through self-reports. Nevertheless, we trained research assistants in interview techniques to ensure that they could adequately elicit information on exposure behaviour. Respondents in the study may have responded to questions on exposure behaviour in ways that appear socially acceptable (social desirability bias). This was reduced by assuring respondents that their data would be anonymized and kept confidential and that the findings of the study would be used for this study only. Furthermore, indirect questions were designed to assess exposure behaviour. Moreover, this was a cross-sectional study design, and thus we cannot infer causality. The study also focused on assessing three exposure pathways in the private domain, potentially overlooking other relevant pathways. In light of these limitations, future research should assess other pathways and explore in-depth the factors influencing exposure behaviours. In addition, qualitative research can uncover the underlying reasons for certain behaviours, consequently tailoring interventions more effectively. Researchers should employ more robust study designs to assess E. coli exposure behaviours and associated factors. In addition, researchers may consider directly observing behaviour to reduce the biases linked to relying on self-reported data.

Conclusion

This study highlights a concerning situation in Imvepi refugee settlement, where there was a substantial prevalence of high-risk exposure behaviour among households, pointing to a significant risk of exposure to excreta-related microorganisms in this community. The study found that living in single-family homes, being aged 15–19 years, having household heads with a low education level, having low-risk perceptions and a low knowledge of exposure, and having sanitary facilities with excreta overflowing from the squat hole, were significantly associated with high-risk exposure behaviour to E. coli. These results highlight the importance of interventions that include behaviour change communication and improved sanitation infrastructure. These interventions should target households whose heads have a low education level, those with young caretakers (15–19 years), those with poor sanitation infrastructure, individuals with limited knowledge of exposure to E. coli and those with low-risk perceptions of exposure.