Advertisement

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Rainfall can reduce risk of enteric pathogen exposure and child GI infection: Evidence from a QMRA analysis in Kibera, Kenya

  • Syed Anjerul Islam,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Validation, Writing – original draft, Writing – review & editing

    Affiliation Department of Environmental Sciences and Engineering, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States of America

    ⨯
  • Sarah Lebu,

    Roles Conceptualization, Data curation, Methodology, Resources, Validation, Writing – review & editing

    Affiliation Department of Environmental Sciences and Engineering, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States of America

    ⨯
  • Joe Brown

    Roles Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing – review & editing

    joebrown@unc.edu

    Affiliation Department of Environmental Sciences and Engineering, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States of America

    ⨯

Abstract

In low- and middle-income countries (LMICs), enteric infections and diarrheal diseases among children are widespread due to insufficient water, sanitation, and hygiene (WASH) infrastructure. Children may be exposed to enteric pathogens by ingesting soil contaminated with feces. Although previous research has detected elevated levels of fecal indicator bacteria and enteric pathogen genes in soils, no studies have yet used quantitative microbial risk assessment (QMRA) to assess the potential effect of rainfall on child enteric infections and the potential disease burden via this pathway. We collected a total of 144 soil samples between May and December 2024 in Kibera, Kenya, both before and within an hour after rainfall events from household entrances, compounds, and toilet entrances. Over 70% of soil samples were E. coli positive, with mean concentrations ranging from 51.5 MPN/g of dry soil (SD: 46.1) pre-rainfall to 39.3 MPN/g of dry soil (SD: 45.7) post-rainfall. The reduction was marginally significant (p = 0.058, Kruskal-Wallis test). A one-way ANOVA revealed no statistically significant difference in E. coli levels across sampling locations (p = 0.943) and across latrine types (urine diverting, hanging, pit latrine, and septic tank) (p = 0.46). Using QMRA with soil ingestion parameters specific to children under five years in LMIC settings and locally derived indicator-pathogen ratios for soils, we estimated annual infections and diarrheal disease burden in disability-adjusted life years (DALYs) per person per year. We estimated an annual infection risk of 100% for adenovirus, irrespective of rainfall events. Following rainfall, median annual infection risks attributable to soil ingestion were an estimated 1 in 2 for Campylobacter jejuni, 1 in 31 for Shigella spp., and 1 in 33 for Vibrio cholerae. These findings highlight the potential importance of soil ingestion as an exposure pathway, and point to the need for sanitation improvements to limit the migration of enteric pathogens to soils near living environments.

Citation: Islam SA, Lebu S, Brown J (2026) Rainfall can reduce risk of enteric pathogen exposure and child GI infection: Evidence from a QMRA analysis in Kibera, Kenya. PLOS Water 5(2): e0000455. https://doi.org/10.1371/journal.pwat.0000455

Editor: Vikram Kapoor, University of Texas at San Antonio, UNITED STATES OF AMERICA

Received: September 30, 2025; Accepted: January 29, 2026; Published: February 19, 2026

Copyright: © 2026 Islam 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 and code are publicly accessible at https://osf.io/ztmy2/

Funding: The author(s) received no specific funding for this work.

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

Introduction

Children living in low- and middle-income countries (LMICs) and particularly in informal settlements face an increased risk of gastrointestinal (GI) infections due to widespread environmental fecal contamination. Informal settlements often suffer from overcrowding, poor sanitation, and inadequate waste disposal systems, exacerbating the risk of fecal-oral disease transmission [1]. Sanitation practices in these communities have been known to contribute to environmental contamination, leading to the release of enteric pathogens in living environments. Open defecation remains prevalent in areas where latrines are either absent or dysfunctional, though is generally lower in urban areas compared with rural settings. Particularly in informal settlements, inefficient fecal sludge emptying practices can contribute to pathogen releases [2]. Even where sanitation infrastructure is moderately effective in containing waste, facilities in informal settings often fail due to flooding, erosion, and other factors [3]. Poor management of child and animal feces further contributes to the presence of enteric pathogens in soil [4,5].

Persistence of these pathogens in the environment depends on meteorological parameters, including temperature, rainfall, ultraviolet (UV) radiation, and other factors. Climate change has introduced variability in rainfall patterns, leading to more intense and unpredictable precipitation events that can significantly impact the transport and distribution of fecal pathogens in the environment [6]. Studies have found that rainfall could influence enteric disease patterns in a variety of ways [7,8]. Rainfall can mobilize microbiological contaminants, which can reduce pathogens where they are concentrated and increase them elsewhere [9].

Children in LMICs are frequently exposed to contaminated soil while playing in household yards. In LMICs, both in rural and urban settings, ingestion of fecally contaminated soil has been shown to be an important route of exposure for young children [10–12] including via geophagy [10,13–15] and dirty hands. This exposure pathway has been linked to a range of adverse health outcomes, including GI infections, impaired cognitive development, and stunting [16–18]

Despite the known risks, significant gaps remain in understanding the quantitative risk associated with soil ingestion in settings with poor sanitation. Quantitative Microbial Risk Assessment (QMRA) is a valuable tool for estimating infection risks and disease burden based on exposure to microbial hazards [19,20]. While previous studies have applied QMRA to assess risks of soil ingestion [21–23], to our knowledge, no previous QMRA studies have specifically examined this exposure pathway in the context of rainfall, despite its potential importance in terms of overall pathogen exposure in early childhood [10,11]. To capture both the immediate and long-term impacts of disease, we used disability-adjusted life years (DALYs), which combine years of life lost due to premature mortality with years lived with disability [24]. DALYs are now widely applied in disease burden research and serve as a key metric for evaluating interventions and prioritizing actions in water, sanitation, and hygiene (WASH) studies [24,25].

This study aims to (1) investigate precipitation-driven variations in fecal indicator bacteria (E. coli) concentrations across household micro-environments (latrine entryways, courtyards/compound areas where child plays, and household entrance) in informal settlements, (2) employ a QMRA framework to estimate annual diarrheal DALYs among children from soil ingestion pre- and post-rainfall events across four enteric pathogens compared with the World Health Organization (WHO) normative guideline for DALYs, and (3) conduct sensitivity analyses to determine the relative contributions of key QMRA input parameters to predicted infection risks. We hypothesized that rainfall-associated changes to fecal contamination of soils would result in differences in exposure and predicted health risks among children in an informal settlement with poor sanitation.

Methods

Study site

Our study site was the informal settlement of Kibera in Nairobi, Kenya. We chose this area based on previous research experience, variable but overall limited sanitation coverage, flooding history, and a strong relationship with local partners. Kibera is a 2.5 km2 area with an estimated population of 1.2 million people [26]. In Kibera and the surrounding Nairobi region, rainfall is characterized by a bimodal distribution, featuring two distinct rainy seasons, exhibiting high inter-annual and intra-seasonal variability. This variability often results in alternating periods of drought and floods, significantly impacting the local environment and community. The long rains occur from March to May, usually peaking in April. This season generally receives more total rainfall (around 310 mm). The short rains occur from October to December, peaking in November. This season typically records less rain than the long rains (around 200 mm) [27]. This study was conducted during the long rainy season and at the onset of an intense flooding period.

Ethics

Our study protocol (24–1333) was reviewed and approved by the Institutional Review Board of the University of North Carolina at Chapel Hill, as well as by the Kenya National Commission for Science, Technology, and Innovation for in-country approvals. We obtained written consent from users, owners, and operators of the sanitation facilities at least 24 hours prior to sampling and observations. In cases of nonresponse and unwillingness to participate, we replaced the sanitation facility by using determined sampling protocols. Our study team carried and displayed flyers with study information at sampling sites to address potential community concerns and provide additional details.

Soil sampling

Between May and December 2024, we systematically collected soil samples from three designated outdoor household locations, including latrine entrances, household entrances, and key compound areas where children frequently play. We selected these locations because they have previously exhibited high levels of fecal contamination in domestic soils [28]. For example, latrine entrances can become contaminated if latrines do not effectively contain and isolate waste [22]. We did not consider potential upstream and downstream positioning of sampling locations in relation to wastewater or sludges, as “upstream” and “downstream” definitions can be inconsistent and variable given the ubiquity of poor sanitation in these dense living environments. We gathered samples both before and within one hour after rainfall. We chose a one-hour delay to conduct sampling after rain as a common practice in environmental monitoring, having the advantage of being repeatable between sampling efforts. Prior research studies have used this strategy to capture the initial, most concentrated “first flush” of pollutants in runoff [29]. At each sampling point, we placed a sterilized 20 × 20 cm stencil and used a sterilized metal scoopula to excavate soil up to a depth of 2 cm. We mixed the soil for 30 seconds with the scoopula before sampling to homogenize the sample. Using a sterile plastic spoon, we transferred 1 g of soil into a 100 mL Whirl-Pack bag, filled it with deionized water up to the 100 mL mark, and carefully added growth media before sealing the bag. We then gently swirled and squeezed the bag to fully dissolve the medium. The mixture was transferred to an Aquagenx Compartment Bag and incubated at an ambient temperature (varies between 25–30°C) for 48 hours before recording most probable number (MPN) readings. We confirmed the presence of E. coli via visual color changes in the compartments and quantified MPN concentrations by comparing the color sequence of the five compartments to a manufacturer-provided reference table. The Aquagenx Compartment Bag Test (CBT) is a well-established method for quantifying E. coli and total coliform in drinking water and wastewater [30–32]. It is particularly suitable for low-resource settings, as it does not require extensive laboratory equipment, incubators, cold chains, or advanced expertise, yet consistently produces reliable results when compared with membrane filtration [33] and the IDEXX Colilert method [34] for E. coli enumeration.

Quantitative Microbial Risk Assessment (QMRA)

Hazard ID

We obtained pathogen-specific data from Bauza et al, [35] who conducted molecular analysis in Kibera under similar conditions in 2020; the pathogen data we used was presented in Table 1. We established a pathogen-to-indicator ratio based on these data. In brief, we selected four enteric pathogens to estimate enteric infection risks in the QMRA model via the pathway of interest: Shigella spp./EIEC, adenovirus, Vibrio cholerae, and Campylobacter jejuni. Pathogen detection and quantification estimates were derived via measurement of specific genetic markers: ipaH for Shigella spp., hexon for adenovirus, ciaB for Campylobacter jejuni, and ctxA for Vibrio cholerae [35]. These markers are highly specific and sensitive in identifying target pathogens via molecular methods.

thumbnail
Download:
Table 1. Model Input parameters.

https://doi.org/10.1371/journal.pwat.0000455.t001

Exposure assessment

We developed QMRA models based on child soil ingestion scenarios for low-resource settings similar to our study site [11]. In low-resource settings, children are frequently exposed to soil environments due to limited indoor play areas and cultural norms that support outdoor play. Increased contact with contaminated soil can result in higher rates of soil ingestion. Poor sanitation infrastructure and the prevalence of unimproved or dirt flooring in households further contribute to exposures via this pathway, elevating the potential risks to children. Additionally, children’s age plays a significant role in overall exposure, as younger children often engage in more frequent hand-to-mouth behaviors. To calculate the pathogen dose for each sample, we applied the following equations (equation:1 to equation:3), assuming 100% pathogen viability (viability factor = 1) as a means of comparing total pathogen-specific potential risk:

(1)(2)(3)

Dose-response and risk characterization

Following dose calculations, we determined the dose-response models from literature [40–43] for each pathogen of interest and estimated the model and parameters from previous studies. Each pathogen dose-response was characterized by a Beta Poisson model (equation 4) as the best fit option to estimate daily infection risk. The dose-response optimized parameters for each pathogen were collected and listed in Table 1.

(4)

where Pinf,event is the probability of infection resulting from a single exposure dose D. In the Beta-Poisson models, N50 denotes the dose at which 50% of the population is expected to be affected and α is the slope parameter used to model pathogen survival probabilities.

The annual risk of infection, Pinf,an was used to characterize the health risk impact using equation 5.

(5)

Burden of disease

We estimated the loss of DALYs per person per year (pppy) among children under five years of age resulting from the ingestion of fecally contaminated soil using equation (6) for each pathogen, assuming 100% of the population was susceptible to estimate total potential risk. From the literature, we estimated the probability of diarrheal illness given infection for Shigella spp. as 36% [44], for Campylobacter jejuni as 32% [45], for Vibrio cholerae as 45% [46], and as 15% for adenovirus [47]. Additionally, we applied global foodborne DALY estimates per case for Shigella spp., Vibrio cholerae and adenovirus using median DALY weights of 0.02, 2.0, and 0.02 DALYs per case, respectively [48]. For Campylobacter jejuni,we used a DALYs estimate of 4.6 x 10-3 per case based on values derived from the drinking water exposure [49].

(6)

There are no applicable normative standards or guidelines for tolerable risk due to soil ingestion as a specific pathway of interest. Therefore, we compared the estimated annual burden of diarrheal disease to the WHO’s recommended threshold for drinking water, set at 10 ⁻ ⁶ DALYs pppy [50].

Data analysis

We used statistical software R version 4.3.0 for descriptive analysis, QMRA model generation, and statistical analysis. We calculated descriptive statistics, including mean and standard deviation, for E. coli before and after rainfall events and across sampling locations. We used box plots to visualize the distribution of E. coli across different sampling locations (toilet entrance, compound locations and household entrance) and rainfall conditions. To assess the impact of rainfall on microbial contamination, we first performed a non-parametric Kruskal-Wallis test across location and toilet type comparing E. coli levels between pre- and post-rainfall periods. To evaluate whether E. coli concentrations varied significantly by sampling location and toilet type, we conducted a one-way ANOVA. We assumed a significance level at p < 0.05 to aid in interpretation of statistical hypothesis testing.

We estimated the pathogen dose for each sample observation by using equations (1) through (3) and input parameters from Table 1 [22]. From our dose–response model (equation 4) and selected distributions, we estimated daily infection risks for each pathogen from each sample. For each soil E. coli sample, other input values were randomly sampled from their respective probability distributions (Table 1). We started with a seed value of 31 for reproducibility. We used equations (5) and (6) to estimate annual infection risks and disease burden for each pathogen, respectively. For the log-transformed plots, we addressed the zero values by replacing them with 1 × 10 ⁻ ⁵, which yielded interpretable and consistent results. We performed sensitivity analysis using Spearman’s rank correlation to evaluate the influence of input parameters on the risk of infection within the QMRA framework.

Results

Descriptive results

A total of 144 soil samples were collected, with an equal split between pre- and post-rainfall events (72 samples each). Overall, 70.8% of soil samples tested positive for E. coli, with approximately 79% of samples showing contamination before rainfall, compared with 62.5% after rainfall. The mean daily rainfall recorded at the rain gauge station during the sampling period was 2.1 mm/day. Within each event, sampling was evenly distributed across three locations: compound, household entrance, and toilet entrance. On average, E. coli levels were higher before rainfall (mean: 51.5 MPN/g of dry soil, SD: 46.1 MPN/g of dry soil) compared to after rainfall (mean: 39.3 MPN/g of dry soil, SD: 45.7 MPN/g of dry soil) (Table A in S1 Text). We observed the highest mean concentration (61.6 MPN/g of dry soil) at toilet entrances before rainfall, which decreased by 46% following rainfall (mean: 33.4 MPN/g of dry soil). Visual assessment using a boxplot (Fig 1) supported this trend, showing consistently higher E. coli counts across all sampling locations before rainfall. We also tracked 4 types of latrines (Fresh life/urine diverting dry toilets (UDDT), hanging toilets, pit latrine, and septic tank) used by the households equally distributed (18 for each toilet type) across the samples. We found pit latrines had the highest E. coli levels in adjacent soils (mean: 75.2 MPN/g of dry soil, SD: 42.4 MPN/g of dry soil) before rainfall, and septic tanks had the lowest associated soil E. coli levels (mean: 36.2 MPN/g of dry soil, SD: 39.9 MPN/g of dry soil) before rainfall events (Table B in S1 Text). A one-way ANOVA revealed no statistically meaningful differences in E. coli levels overall by sampling location (p = 0.94) and latrine type (p = 0.46).

thumbnail
Download:
Fig 1. E. coli count (MPN/g of dry soil) across soil sampling locations and rainfall events.

https://doi.org/10.1371/journal.pwat.0000455.g001

Two of the locations (compound and household entrance) showed no significant difference in E. coli concentrations before and after rainfall (p = 0.50 and p = 0.74, respectively). However, samples collected from the entrance of toilets showed a statistically significant reduction in E. coli levels before versus after rainfall (p = 0.020). Similarly, after running Kruskal-Wallis test across toilet types, we did not find any significant association between before and after rainfall events except for pit latrines (p = 0.005).

QMRA model output

Risk of infection

The daily and annual risk of infection were quantified for Shigella spp., Vibrio cholerae, adenovirus, and Campylobacter jejuni based on soil ingestion from low-resource settings and presented using a box plot (Fig 2). The results indicated that the daily median risk of infection was highest for adenovirus, followed by Campylobacter jejuni, Shigella spp., and Vibrio cholerae, and they were 1 in 7 persons, 1 in 350 persons, 1 in 11,058 persons, and 1 in 11,910 persons, respectively after rainfall events. Both daily and annual infection risks decreased following rainfall events.

thumbnail
Download:
Fig 2. Daily and annual infection risks by Shigella spp., Vibrio cholerae, adenovirus and Campylobacter jejuni across rainfall events.

https://doi.org/10.1371/journal.pwat.0000455.g002

For Shigella spp., the median daily risk declined from 5.93 x 10-4 (95% CI: 9.52 x 10-5 - 1.15 x 10-3) before rainfall to 9.04 x 10-5 (95% CI: 1.44 x 10-5 – 7.24 x 10-4) after rainfall, an 84% reduction (Table C in S1 Text). Daily risks were similarly reduced for Vibrio cholerae (80%), adenovirus (76%), and Campylobacter jejuni (80%) post rainfall event (Table C in S1 Text). Likewise, the annual risk of infection from Shigella spp. declined from 0.19(95% CI: 0.03-0.34) before rainfall to 0.03 (95% CI: 0.01-0.23) after rainfall (Table D in S1 Text). A similar trend was evident for Vibrio cholerae showing notable declines (~79%), whereas Campylobacter jejuni demonstrated only a 35% reduction. The median annual infection risks from Vibrio cholerae and Campylobacter jejuni were 1 in 33 and 1 in 2 after rainfall, respectively. Adenovirus demonstrated 100% infection regardless of rainfall events, indicating all children were at risk of infection annually (Fig 2).

Disease burden

The burden of enteric disease associated with selected pathogens is shown in Fig 3 and Table E in S1 Text. Generally, the disease burden (DALYs), decreased following rainfall events across all reference pathogens, though the pattern of reduction varied.

thumbnail
Download:
Fig 3. Annual DALYs by Shigella spp., Vibrio cholerae, adenovirus and Campylobacter jejuni across rainfall events.

The dotted line corresponds to the recommended drinking water risk in terms of annual DALYs (10-6).

https://doi.org/10.1371/journal.pwat.0000455.g003

Vibrio cholerae exhibited the highest annual DALYs, with a median of 0.03 DALYs (95% CI: 1.71 x10-3-0.11) post-rainfall, compared to 0.13 DALYs (95% CI: 0.04-0.25) pre-rainfall, indicating a significant shift in burden from annual infection risk. A DALY of 0.03 per person per year indicated that, on average, an individual would be predicted to lose approximately 0.04 years of healthy life—equivalent to about 10 days of life annually—due to this diarrheal disease. Adenovirus contributed 0.30% to the annual DALYs, with no notable change in this proportion before and after rainfall events. For Shigella spp. and Campylobacter jejuni, the annual DALYs were 2.33 x 10-4 (95% CI: 3.79 x 10-5 -1.67 x 10-3) and 1.07 x 10-3 (95% CI: 5.66 x 10-5 -1.62 x 10-3), respectively, post-rainfall. When compared to the WHO drinking water safety thresholds for each pathogen, the DALYs for Vibrio cholerae, adenovirus, Campylobacter jejuni, and Shigella spp. after rainfall events exceeded the standards by factors of 27,100; 3,000; 1,100; and 234, respectively.

Sensitivity analysis

Sensitivity analysis using Spearman rank correlations revealed distinct tiers of association with infection risk. E. coli concentrations in soil were the strongest contributors to infection risk estimates across all pathogens, with a correlation coefficient (ρ) exceeding 0.8 (Fig 4). In contrast, correlations with the pathogen indicator ratio, gene target abundance, and soil ingestion rates were notably weaker (ρ ≤ 0.30), suggesting these parameters contributed to variability in risk but were secondary drivers compared to E. coli. While most pathogenic gene targets demonstrated moderate associations (ρ ≈ 0.2) with infection risk in Monte Carlo simulations, Vibrio cholerae (gene target: ctxA) was an exception, showing a weaker correlation (ρ < 0.07). Overall, these findings indicated that variability in daily infection risk was primarily driven by E. coli levels, while other exposure and pathogen-specific factors played a more modest role.

thumbnail
Download:
Fig 4. Sensitivity analysis of input variables influencing the risk of GI infection from enteric pathogens of interest.

https://doi.org/10.1371/journal.pwat.0000455.g004

Discussion

Our study revealed a decrease in E. coli levels following rainfall, corresponding to lower estimated pathogen doses and modeled diarrheal disease burdens across attributable to soil ingestion. In both pre- and post-rainfall events, the estimated disease burden exceeded the WHO’s acceptable threshold for drinking water, generally by several orders of magnitude. These findings highlight the potential importance of soils in contributing to the burden of enteric disease in settings like the one we studied, where sanitation was limited.

Our findings were consistent with previous literature suggesting that rainfall could reduce diarrheal burden under some conditions. In a study in Ecuador, rainfall was associated with a reduction in diarrheal disease incidence (incidence rate ratio [IRR]: 0.74, 95% CI: 0.59–0.92), indicating that diarrheal cases were approximately 26% lower following wet periods [7]. This reduction was driven by lower exposure doses in a QMRA framework, which directly reflected the observed decline in E. coli concentrations in soil after rainfall. A study in Bangladesh reported that heavy rainfall was associated with significantly lower E. coli counts in soil (E.coli count ratio = 0.36, 95% CI: 0.24–0.53, p < 0.0005) compared to no rainfall [51]. This pattern can be explained by the persistence of E. coli in surface soils under favorable conditions, followed by mobilization during periods of intense precipitation. Extreme rainfall can generate infiltration-excess runoff, facilitating the dispersal of pathogens into the broader environment [52], possibly reducing near-source concentrations. Studies have found reduced contamination in courtyard soil after rainfall, but increased contamination of drinking water and food [51]. This redistribution does not necessarily eliminate pathogens but rather shifts them to adjacent soil or water bodies, thereby broadening potential exposure pathways and complicating public health interventions.

In contrast, many studies have reported an opposite trend—an increase in fecal contamination and diarrheal incidence with heavier rainfall [7,8,52–55]. In the same study in Ecuador, following dry periods, observed heavy rainfall was associated with an increased incidence of diarrheal disease (IRR: 1.39, 95% CI: 1.03–1.87), suggesting approximately a 39% higher risk compared to non-rainy conditions [7]. Similarly, a meta-analysis found increased diarrheal incidence (IRR: 1.26, 95% CI: 1.05-1.51) following extreme rainfall which followed by dry period as well [56]. The “first flush” phenomenon, whereby pathogens accumulate during dry periods and are suddenly released following intense rainfall, has been linked to increased diarrheal incidence and outbreaks [51,56,57]. In other words, the relationship between rainfall and diarrheal disease is complex rather than strictly linear, as pathogen survival in the environment is highly context-specific and varies across different organisms. Clearly, rainfall can either increase or decrease fecal contamination in localized environments as contaminants are mobilized and redistributed. Rainfall’s effect on exposure to enteric pathogens is probably not generalizable, therefore.

Among the pathogens assessed, V. cholerae accounted for the highest DALYs pppy while lowest daily and annual infections followed rainfall. The disproportionate DALY contribution of V. cholerae reflects the high DALY weight and severity associated with cholera outcomes, which elevated its overall disease burden compared to the other reference pathogens we considered. Previous work has demonstrated that settings with inadequate WASH infrastructure are susceptible to cholera outbreaks [58,59]. Approximately 565 cholera outbreaks were reported annually between 2011 and 2020 in India, for example [59]. Transmission of V. cholerae can also occur through person-to-person contact, and because infections are frequently asymptomatic, the true incidence of cholera is often underestimated [58,59].

We estimated a high daily and annual infection risk for adenovirus, reaching 100% across all scenarios-regardless of rainfall. This suggested a persistent and robust environmental presence, possibly attributable to adenovirus’s moderate to high environmental stability compared with other enveloped viruses, relative resistance to degradation, and relatively low median infectious dose [42].

The estimated disease burden of C. jejuni exceeded the WHO drinking water threshold by more than 1,000-fold, irrespective of rainfall events. C. jejuni also has a low median infectious dose, with as few as 1–10 CFU sufficient to initiate illness [60]. Campylobacter is common among children under five years of age as an etiology diarrheal disease [61].

In comparison, the median DALY estimate for Shigella spp. was more than 200 times higher than the WHO normative drinking water-attributable burden threshold, though risks were relatively lower than those of the other reference pathogens assessed. Shigella spp. is less environmentally stable and can be rapidly inactivated under unfavorable environmental conditions [62]. Shigella spp. is characterized by a low median infectious dose (10–100 organisms to cause disease) [44] and is among the more important enteric infections in children in LMICs [63].

Beyond pathogen-specific differences in attributable risk estimates, location can also play an important role in shaping exposure patterns among children. Notably, although no significant differences were observed between pre- and post-rainfall events across most sites (household entrance, compound, and toilet entrance), soil at toilet entrances exhibited substantially higher E. coli concentrations before rainfall. This pattern could be attributed to multiple factors, though our study was not intended to examine them. For example, ineffective fecal sludge management practices that result in spills, poor structural integrity of latrines resulting in leaks, or possibly behaviors (e.g., fecal matter on footwear, inadequate containment of child or animal feces). A study in Mozambique, also found E.coli in 87% of soil entrance samples with a mean concentration of 3.0 log10 CFU/g of dry soil (SD: 1.1) [22]. In contrast, A study in Tanzania reported higher E. coli pathotype gene on household floors (83%) compared to toilet entrances (33%), likely due to children defecating in the household yard [64].

The type of sanitation facility was a key determinant of soil contamination. Our findings indicated that pit latrines were associated with particularly high E. coli levels in the soil prior to rainfall events. Similarly, a recent study in peri-urban Africa found that pit latrines were associated with soil contamination [65]. It could be due to the practice of manual and unhygienic pit emptying in low-resource settings [66] or the other factors we have noted. While previous studies have suggested that pit latrines may offer greater adaptability, our findings indicated that septic systems and dry toilets may be more effective in containing waste, thereby reducing potential pathogen exposure via soils near latrines. Well-functioning septic systems can minimize pathogen leakage into soil via sequestration of waste in a subsurface tank, while dry toilets can reduce microbes via desiccation. These systems should be considered as viable options in rainfall-affected areas [3], where runoff can exacerbate or mobilize contamination in soil, resulting in downstream exposures.

Our sensitivity analysis using Spearman rank correlations identified E. coli concentrations in soil as the most influential parameter across all studied pathogens, with strong positive correlations regardless of rainfall events. E. coli is a widely accepted but imperfect [67] fecal indicator organism. A previous study also found strong relationships (Spearman correlation coefficient = 0.70, P < 0.01) between soil E.coli and diarrheal cases between March 2018 and March 2019 in Burkina Faso [68]. Its moderately strong correlation with infection and disease risks [69–71] across pathogens suggests utility as a robust proxy for fecal contamination severity, even in soils. Higher E. coli levels likely reflect greater fecal loading, amplifying pathogen exposure risks [22].

Limitations

This study has several limitations that should be acknowledged. First, the cross-sectional nature of the data collection provided only a snapshot of E. coli contamination across different locations and time points. Given the known variability in microbial concentrations in soil due to temporal and environmental fluctuations, the results may not capture the full range of contamination dynamics, especially under different rainfall conditions. Longitudinal data would be more informative to understand the persistence and changes in microbial contamination over time.

Second, rainfall intensity and seasonality were not explicitly incorporated into the exposure assessment. Consequently, we estimated annual infection risk by assuming daily exposure over 365 days and aggregating daily risks, without accounting for temporal variability in rainfall patterns or differences in exposure frequency associated with wet and dry seasons. A seasonally specific QMRA framework, informed by high-resolution rainfall data and behavioral observations would give more realistic risk estimates.

Third, while we used previously published pathogen-specific data to estimate pathogen-to-indicator ratios from Kenya, these ratios were derived from different studies and may not represent the microbial ecology of our study area, introducing a degree of uncertainty into the risk estimates. Ideally, performing molecular (PCR-based) detection for the target pathogens directly on the same environmental samples would have allowed for more accurate quantification and risk estimation. Additionally, our study did not include parasitic or helminthic pathogens, which may play a significant role in soil-based transmission, particularly among children. The reliance on literature-based conversions, while necessary, may thus affect the precision of the QMRA model outputs.

Fourth, we did not collect any data on soil characteristics like soil type, which might be an important factor for pathogen transport after rainfall. Soils vary in their permeability and therefore potential to contain or mobilize microbes in the localized environment [51].

Fifth, the generalizability of our findings to other contexts may be limited. The study was conducted in a specific low-resource setting with unique infrastructural, environmental, and behavioral characteristics. Like any study conducted in a specific time and place, we stress that the results we report are generalizable only to the extent that the study conditions hold in other settings. As such, the patterns of contamination, exposure, and infection risks observed here may not reflect those in other geographic or socio-economic contexts, particularly where sanitation and hygiene practices differ significantly.

Conclusion

This study provided insights into enteric infections related to precipitation among children under five and was among the first to apply QMRA to compare pre- and post-rainfall events. Rainfall appeared to reduce the predicted diarrheal disease burden in this setting, though evidence from other studies suggests the relationship between rainfall and exposures can be complex and context-specific. This study underscored the critical role of environmental contamination, particularly in low-resource settings, in driving enteric infection risks among vulnerable populations. Our findings suggest that pit latrines and toilet-adjacent soils were key microbial hotspots, with significantly higher E. coli levels observed before rainfall events, suggesting the accumulation of pathogens that may later be dispersed by rain. Although rainfall appeared to reduce surface contamination, the marginal significance points to the need for further investigation using larger sample sizes and refined methodologies. The QMRA model revealed that annual disease burden for all four pathogens was substantially higher than the WHO annual drinking water threshold for disease burden regardless of rainfall, suggesting risk of disease via this exposure pathway. The greatest estimated DALYs were attributable to Vibrio cholerae, followed by adenovirus, Campylobacter jejuni and Shigella spp. and reduced after rainfall across all. Sensitivity analysis further identified E. coli concentration as the most influential parameter in predicted infection risk, reinforcing the need for focused sanitation interventions and behavioral strategies to limit children’s exposure near toilets. These findings provided evidence to inform public health policies aimed at reducing pathogen exposure and improving child health outcomes in high-risk, informal communities.

Supporting information

S1 Text. Table A in S1 Text. E. coli count by location and rainfall event (all units in MPN/g of dry soil).

Table B in S1 Text. E. coli count by toilet type and rainfall (all units in MPN/g of dry soil). Table C in S1 Text. Summary statistics of daily risk of infection for pathogens across rainfall events. Table D in S1 Text. Summary statistics of annual risk of infection for pathogens across rainfall events. Table E in S1 Text. Summary statistics of annual DALYs for pathogens before and after rainfall events. And comparison with WHO drinking water threshold of risk (10-6 DALYs).

https://doi.org/10.1371/journal.pwat.0000455.s001

(DOCX)

S1 Checklist. Inclusivity in global research.

https://doi.org/10.1371/journal.pwat.0000455.s002

(DOCX)

Acknowledgments

We want to acknowledge CFK Africa for their valuable partnership in supporting this research. Special thanks to the dedicated field staff who contributed to data collection: Wally, Mariam, Andrew, Elizabeth, and Ben. Also, we would like to thank Dr Megan Lott for organizing and leading the QMRA working group, which provided valuable insights for this work.

References

  1. 1. Centers for Disease Control and Prevention CDC. Global WASH Fast Facts. https://www.cdc.gov/healthywater/global/wash_statistics.html 2022. 2023 November 13.
    • 2. Conaway K, Lebu S, Heilferty K, Salzberg A, Manga M. On-site sanitation system emptying practices and influential factors in Asian low- and middle-income countries: A systematic review. Hygiene and Environmental Health Advances. 2023;6:100050.
    • 3. Lebu S, Gyimah R, Nandoya E, Brown J, Salzberg A, Manga M. Assessment of sanitation infrastructure resilience to extreme rainfall and flooding: Evidence from an informal settlement in Kenya. J Environ Manage. 2024;354:120264. pmid:38354609
    • 4. Beardsley R, Lebu S, Anthonj C, Manga M. Child feces disposal practices in humanitarian and non-humanitarian settings across 34 low- and middle-income countries. Sci Total Environ. 2024;940:173547.
    • 5. Berendes DM, Kirby AE, Clennon JA, Agbemabiese C, Ampofo JA, Armah GE, et al. Urban sanitation coverage and environmental fecal contamination: Links between the household and public environments of Accra, Ghana. PLoS One. 2018;13(7):e0199304. pmid:29969466
    • 6. Shahid S. Rainfall variability and the trends of wet and dry periods in Bangladesh. Int J Climatol. 2010;30(15):2299–313.
    • 7. Carlton EJ, Eisenberg JNS, Goldstick J, Cevallos W, Trostle J, Levy K. Heavy rainfall events and diarrhea incidence: the role of social and environmental factors. Am J Epidemiol. 2014;179(3):344–52. pmid:24256618
    • 8. Levy K, Woster AP, Goldstein RS, Carlton EJ. Untangling the impacts of climate change on waterborne diseases: a systematic review of relationships between diarrheal diseases and temperature, rainfall, flooding, and drought. Environ Sci Technol. 2016;50(10):4905–22.
    • 9. Cregger MA, Schadt CW, McDowell NG, Pockman WT, Classen AT. Response of the Soil Microbial Community to Changes in Precipitation in a Semiarid Ecosystem. Applied and Environmental Microbiology. 2012;78(24):8587–94.
    • 10. Bauza V, Byrne DM, Trimmer JT, Lardizabal A, Atiim P, Asigbee MAK, et al. Child soil ingestion in rural Ghana - frequency, caregiver perceptions, relationship with household floor material and associations with child diarrhoea. Trop Med Int Health. 2018;23(5):558–69. pmid:29537690
    • 11. Kwong LH, Ercumen A, Pickering AJ, Unicomb L, Davis J, Leckie JO, et al. Soil ingestion among young children in rural Bangladesh. J Expo Sci Environ Epidemiol. 2021;31(1):82–93. pmid:31673039
    • 12. Ngure FM, Humphrey JH, Mbuya MNN, Majo F, Mutasa K, Govha M, et al. Formative research on hygiene behaviors and geophagy among infants and young children and implications of exposure to fecal bacteria. Am J Trop Med Hyg. 2013;89(4):709–16. pmid:24002485
    • 13. Geissler PW, Mwaniki DL, Thiong’o F, Friis H. Geophagy among school children in western Kenya. Trop Med Int Health. 1997;2(7):624–30. pmid:9270730
    • 14. Kwong LH, Ercumen A, Pickering AJ, Unicomb L, Davis J, Leckie JO, et al. Soil ingestion among young children in rural Bangladesh. J Expo Sci Environ Epidemiol. 2021;31(1):82–93. pmid:31673039
    • 15. Vermeer DE, Frate DA. Geophagia in rural Mississippi: environmental and cultural contexts and nutritional implications. Am J Clin Nutr. 1979;32(10):2129–35. pmid:484531
    • 16. Faruque A, Alam B, Nahar B, Parvin I, Barman AK, Khan SH. Water, sanitation, and hygiene (WASH) practices and outreach services in settlements for Rohingya population in Cox’s Bazar, Bangladesh, 2018-2021. Int J Environ Res Public Health. 2022;19(15):9635.
    • 17. Kotloff KL, Nataro JP, Blackwelder WC, Nasrin D, Farag TH, Panchalingam S, et al. Burden and aetiology of diarrhoeal disease in infants and young children in developing countries (the Global Enteric Multicenter Study, GEMS): a prospective, case-control study. Lancet. 2013;382(9888):209–22. pmid:23680352
    • 18. Null C, Stewart CP, Pickering AJ, Dentz HN, Arnold BF, Arnold CD, et al. Effects of water quality, sanitation, handwashing, and nutritional interventions on diarrhoea and child growth in rural Kenya: a cluster-randomised controlled trial. Lancet Glob Health. 2018;6(3):e316–29. pmid:29396219
    • 19. Haas CN. Microbial dose response modeling: past, present, and future. Environ Sci Technol. 2015;49(3):1245–59. pmid:25545032
    • 20. Haas CN. Quantitative Microbial Risk Assessment and Molecular Biology: Paths to Integration. Environmental Science & Technology. 2020;54(14):8539–46.
    • 21. Balderrama-Carmona AP, Gortáres-Moroyoqui P, Álvarez-Valencia LH, Castro-Espinoza L, Mondaca-Fernández I, Balderas-Cortés J de J, et al. Occurrence and quantitative microbial risk assessment of Cryptosporidium and Giardia in soil and air samples. Int J Infect Dis. 2014;26:123–7. pmid:25043458
    • 22. Capone D, Bivins A, Knee J, Cumming O, Nalá R, Brown J. Quantitative Microbial Risk Assessment of Pediatric Infections Attributable to Ingestion of Fecally Contaminated Domestic Soils in Low-Income Urban Maputo, Mozambique. Environ Sci Technol. 2021 Feb 2;55(3):1941–52.
    • 23. Katukiza AY, Ronteltap M, van der Steen P, Foppen JWA, Lens PNL. Quantification of microbial risks to human health caused by waterborne viruses and bacteria in an urban slum. J Appl Microbiol. 2014;116(2):447–63. pmid:24127653
    • 24. Troeger C, Colombara DV, Rao PC, Khalil IA, Brown A, Brewer TG, et al. Global disability-adjusted life-year estimates of long-term health burden and undernutrition attributable to diarrhoeal diseases in children younger than 5 years. Lancet Glob Health. 2018;6(3):e255–69. pmid:29433665
    • 25. Bivins AW, Sumner T, Kumpel E, Howard G, Cumming O, Ross I, et al. Estimating Infection Risks and the Global Burden of Diarrheal Disease Attributable to Intermittent Water Supply Using QMRA. Environ Sci Technol. 2017;51(13):7542–51. pmid:28582618
    • 26. Gallaher CM, Mwaniki D, Njenga M, Karanja NK, WinklerPrins AMGA. Real or perceived: the environmental health risks of urban sack gardening in Kibera slums of Nairobi, Kenya. Ecohealth. 2013;10(1):9–20. pmid:23512752
    • 27. Kotikot SM, Smithwick EAH, Greatrex H. Observations of enhanced rainfall variability in Kenya, East Africa. Sci Rep. 2024;14(1):12915. pmid:38839907
    • 28. Holcomb DA, Knee J, Sumner T, Adriano Z, de Bruijn E, Nalá R, et al. Human fecal contamination of water, soil, and surfaces in households sharing poor-quality sanitation facilities in Maputo, Mozambique. Int J Hyg Environ Health. 2020;226:113496. pmid:32135507
    • 29. Deletic A. The first flush load of urban surface runoff. Water Res. 1998;32(8):2462–70.
    • 30. Gronewold A, Sobsey M, McMahan L. The compartment bag test (CBT) for enumerating fecal indicator bacteria: basis for design and interpretation of results. Sci Total Environ. 2017;587.
    • 31. Stauber C, Miller C, Cantrell B, Kroell K. Evaluation of the compartment bag test for the detection of Escherichia coli in water. J Microbiol Methods. 2014;99:66–70. pmid:24566129
    • 32. Wang A, McMahan L, Rutstein S, Stauber C, Reyes J, Sobsey MD. Household microbial water quality testing in a Peruvian demographic and health survey: evaluation of the compartment bag test for Escherichia coli. Am J Trop Med Hyg. 2017;96(4):970–5.
    • 33. Mates A, Shaffer M. Membrane filtration differentiation of E. coli from coliforms in the examination of water. J Appl Bacteriol. 1989;67(3):343–6. pmid:2693426
    • 34. Kinzelman JL, Singh A, Ng C, Pond KR, Bagley RC, Gradus S. Use of IDEXX Colilert-18® and Quanti-Tray/2000 as a Rapid and Simple Enumeration Method for the Implementation of Recreational Water Monitoring and Notification Programs. Lake and Reservoir Management. 2005;21(1):73–7.
    • 35. Bauza V, Madadi V, Ocharo R, Nguyen TH, Guest JS. Enteric pathogens from water, hands, surface, soil, drainage ditch, and stream exposure points in a low-income neighborhood of Nairobi, Kenya. Sci Total Environ. 2020;709:135344.
    • 36. Lin WS, Cheng CM, Van KT. A Quantitative PCR Assay for Rapid Detection of Shigella Species in Fresh Produce. J Food Prot. 2010;73(2):221–33.
    • 37. Luo Y, Ye J, Payne M, Hu D, Jiang J, Lan R. Genomic Epidemiology of Vibrio cholerae O139, Zhejiang Province, China, 1994-2018. Emerg Infect Dis. 2022;28(11):2253–60. pmid:36285907
    • 38. Ebner K, Pinsker W, Lion T. Comparative sequence analysis of the hexon gene in the entire spectrum of human adenovirus serotypes: phylogenetic, taxonomic, and clinical implications. J Virol. 2005;79(20):12635–42. pmid:16188965
    • 39. Malik-Kale P, Parker CT, Konkel ME. Culture of Campylobacter jejuni with sodium deoxycholate induces virulence gene expression. J Bacteriol. 2008;190(7):2286–97. pmid:18223090
    • 40. DuPont HL, Levine MM, Hornick RB, Formal SB. Inoculum size in shigellosis and implications for expected mode of transmission. J Infect Dis. 1989;159(6):1126–8. pmid:2656880
    • 41. Diringer H, Roehmel J, Beekes M. Effect of repeated oral infection of hamsters with scrapie. J Gen Virol. 1998;79(3):609–12.
    • 42. Teunis P, Schijven J, Rutjes S. A generalized dose-response relationship for adenovirus infection and illness by exposure pathway. Epidemiol Infect. 2016;144(16):3461–73. pmid:27571926
    • 43. Black RE, Levine MM, Clements ML, Hughes TP, Blaser MJ. Experimental Campylobacter jejuni infection in humans. J Infect Dis. 1988;157(3):472–9. pmid:3343522
    • 44. Zaidi MB, Estrada-García T. Shigella: A Highly Virulent and Elusive Pathogen. Curr Trop Med Rep. 2014 June 1;1(2):81–7.
    • 45. Nielsen HL, Engberg J, Ejlertsen T, Bücker R, Nielsen H. Short-term and medium-term clinical outcomes of Campylobacter concisus infection. Clin Microbiol Infect. 2012;18(11):E459-65. pmid:22882347
    • 46. Midani FS, Weil AA, Chowdhury F, Begum YA, Khan AI, Debela MD, et al. Human Gut Microbiota Predicts Susceptibility to Vibrio cholerae Infection. J Infect Dis. 2018;218(4):645–53. pmid:29659916
    • 47. Shieh W-J. Human adenovirus infections in pediatric population - An update on clinico-pathologic correlation. Biomed J. 2022;45(1):38–49. pmid:34506970
    • 48. Havelaar AH, Kirk MD, Torgerson PR, Gibb HJ, Hald T, Lake RJ, et al. World Health Organization Global Estimates and Regional Comparisons of the Burden of Foodborne Disease in 2010. PLoS Med. 2015;12(12):e1001923.
    • 49. World Health Organization. Quantitative microbial risk assessment: application for water safety management. Geneva: World Health Organization. 2016. https://iris.who.int/handle/10665/246195
      • 50. World Health Organization. Guidelines for drinking-water quality, 4th edition, incorporating the 1st addendum. 2017. https://www.who.int/publications/i/item/9789241549950
        • 51. Niven C, Islam M, Nguyen A, Mertens A, Pickering AJ, Kwong LH, et al. Effects of weather extremes on fecal contamination along pathogen transmission pathways in rural Bangladeshi households. medRxiv. 2023. 2023.12.27.23300582. https://www.medrxiv.org/content/10.1101/2023.12.27.23300582v1
          • 52. Liu X, Zuo C, Guan J, Ma Y, Liu Y, Zhao G, et al. Extreme rainfall disproportionately impacts E. coli concentrations in Texas recreational waterbodies. Sci Total Environ. 2025;958:178062. pmid:39674162
          • 53. Mudadu AG, Spanu C, Salza S, Piras G, Uda MT, Giagnoni L, et al. Association between rainfall and Escherichia coli in live bivalve molluscs harvested in Sardinia, Italy. Food Res Int. 2023;174(Pt 1):113563. pmid:37986518
          • 54. Poulin C, Peletz R, Ercumen A, Pickering AJ, Marshall K, Boehm AB, et al. What Environmental Factors Influence the Concentration of Fecal Indicator Bacteria in Groundwater? Insights from Explanatory Modeling in Uganda and Bangladesh. Environ Sci Technol. 2020;54(21):13566–78. pmid:32975935
          • 55. Powers JE, Mureithi M, Mboya J, Campolo J, Swarthout JM, Pajka J, et al. Effects of High Temperature and Heavy Precipitation on Drinking Water Quality and Child Hand Contamination Levels in Rural Kenya. Environ Sci Technol. 2023;57(17):6975–88. pmid:37071701
          • 56. Kraay ANM, Man O, Levy MC, Levy K, Ionides E, Eisenberg JNS. Understanding the Impact of Rainfall on Diarrhea: Testing the Concentration-Dilution Hypothesis Using a Systematic Review and Meta-Analysis. Environ Health Perspect. 2020;128(12):126001. pmid:33284047
          • 57. Nichols G, Lane C, Asgari N, Verlander NQ, Charlett A. Rainfall and outbreaks of drinking water related disease and in England and Wales. J Water Health. 2008;7(1):1–8.
          • 58. Montero DA, Vidal RM, Velasco J, George S, Lucero Y, Gómez LA, et al. Vibrio cholerae, classification, pathogenesis, immune response, and trends in vaccine development. Front Med (Lausanne). 2023;10:1155751. pmid:37215733
          • 59. Muzembo BA, Kitahara K, Debnath A, Ohno A, Okamoto K, Miyoshi SI. Cholera Outbreaks in India, 2011-2020: A Systematic Review. Int J Environ Res Public Health. 2022;19(9):5738.
          • 60. Abe H, Takeoka K, Fuchisawa Y, Koyama K, Koseki S. A New Dose-Response Model for Estimating the Infection Probability of Campylobacter jejuni Based on the Key Events Dose-Response Framework. Appl Environ Microbiol. 2021;87(20):e0129921. pmid:34347512
          • 61. Mason J, Iturriza-Gomara M, O’Brien SJ, Ngwira BM, Dove W, Maiden MCJ, et al. Campylobacter infection in children in Malawi is common and is frequently associated with enteric virus co-infections. PLoS One. 2013;8(3):e59663. pmid:23555739
          • 62. von Seidlein L, Kim DR, Ali M, Lee H, Wang X, Thiem VD, et al. A multicentre study of Shigella diarrhoea in six Asian countries: disease burden, clinical manifestations, and microbiology. PLoS Med. 2006;3(9):e353. pmid:16968124
          • 63. Muzembo BA, Kitahara K, Mitra D, Ohno A, Khatiwada J, Dutta S, et al. Burden of Shigella in South Asia: a systematic review and meta-analysis. J Travel Med. 2023;30(1):taac132. pmid:36331282
          • 64. Pickering AJ, Julian TR, Marks SJ, Mattioli MC, Boehm AB, Schwab KJ, et al. Fecal contamination and diarrheal pathogens on surfaces and in soils among Tanzanian households with and without improved sanitation. Environ Sci Technol. 2012;46(11):5736–43. pmid:22545817
          • 65. Otunola BO, Zhou L. The Impacts of Septic Tanks and Pits Latrines on Soil and Water in Peri-Urban Areas of Africa. IJSDP. 2024;19(7):2779–87.
          • 66. Capone D, Buxton H, Cumming O, Dreibelbis R, Knee J, Nalá R, et al. Impact of an intervention to improve pit latrine emptying practices in low income urban neighborhoods of Maputo, Mozambique. Int J Hyg Environ Health. 2020;226:113480. pmid:32086016
          • 67. Oh S, Buddenborg S, Yoder-Himes DR, Tiedje JM, Konstantinidis KT. Genomic diversity of Escherichia isolates from diverse habitats. PLoS One. 2012;7(10):e47005. pmid:23056556
          • 68. Robert E, Grippa M, Nikiema DE, Kergoat L, Koudougou H, Auda Y, et al. Environmental determinants of E. coli, link with the diarrheal diseases, and indication of vulnerability criteria in tropical West Africa (Kapore, Burkina Faso). PLoS Negl Trop Dis. 2021;15(8):e0009634. pmid:34403418
          • 69. Brown JM, Proum S, Sobsey MD. Escherichia coli in household drinking water and diarrheal disease risk: evidence from Cambodia. Water Sci Technol. 2008;58(4):757–63. pmid:18776609
          • 70. Moe CL, Sobsey MD, Samsa GP, Mesolo V. Bacterial indicators of risk of diarrhoeal disease from drinking-water in the Philippines. Bull World Health Organ. 1991;69(3):305–17. pmid:1893505
          • 71. Gruber JS, Ercumen A, Colford JM Jr. Coliform bacteria as indicators of diarrheal risk in household drinking water: systematic review and meta-analysis. PLoS One. 2014;9(9):e107429. pmid:25250662