Quantitative microbial risk assessment associated with the use of container-based toilets in Haiti

Although access to sanitation is now considered a fundamental human right, about 2.3 billion people worldwide still lack access to basic sanitation facilities (Dickin et al. 2020). This problem is particularly common in low-income countries and promotes inadequate sanitation practices such as open defecation and/or dumping untreated feces into the environment (Jean et al. 2017; Ufomba et al. 2021). These practices pose a threat to human health (Saleem et al. 2019; Ufomba et al. 2021). Pathogens present in feces can contaminate the environment and, subsequently, cause infectious diseases in humans (Feachem et al. 1983; Mara 2004), most of which are contagious (Cloeckaert & Kuchler 2020; Zhang 2022). In Haiti, due to fecal contamination of the Artibonite River in 2010 (Guimier 2011), cholera was responsible for nearly 9,800 deaths and more than 820,000 suspected cases from 2010 to 2019 (Griffiths et al. 2021).

Quantitative Microbial Risk Assessment (QMRA) is an assessment method developed in the 1970s by the United States National Research Council, which is inspired by the chemical risk assessment method (De Giudici et al. 2011). It consists of four main steps: hazard identification, exposure assessment, hazard characterization (often reduced to dose-response assessment), and risk characterization (Haas et al. 1999; U.S. EPA 2012; WHO 2022). Studies relating to the QMRA have been carried out on composting toilets, but they have generally focused on the health risks associated with either the use of compost resulting from the recovery of feces (Nakagawa et al. 2006; Schonning et al. 2007; Darimani et al. 2015; Kumwenda et al. 2017) or spreading feces on gardens or agricultural fields (WHO 2006). The conclusions were divided in regard to the results of these studies. According to Nakagawa et al. (2006) and Schonning et al. (2007), the risk of infection is generally below the acceptable level of risk, whereas it is above the acceptable level of risk according to Kumwenda et al. (2017) and Darimani et al. (2015). The acceptable level of risk corresponds to 10⁻⁴ per person per year (Nakagawa et al. 2006; Schonning et al. 2007; Darimani et al. 2015). This disparity is mainly due to (i) the pathogens targeted, which differ between studies and/or regions, (ii) the different exposure scenarios developed by the authors, and (iii) the types of toilets considered. In addition, a semi-QMRA associated with the use of CBTs was carried out by Mackinnon et al. (2018) where Escherichia coli was considered as the target pathogen. The study revealed a high level of fecal contamination on toilet surfaces and a high risk of infection, through hand-to-mouth contact, in users and operators.

Unlike the aforementioned previous studies, the present paper aims to quantitatively assess the microbial risks faced by operators. It is not interested in the risk associated with the use of compost or feces as a fertilizer, as that subject has already been extensively studied in previous research. Furthermore, this study takes into consideration pathogens that have not been considered by other studies, such as Campylobacter spp., poliovirus, Shigella spp., and Vibrio cholerae.

Grande Plaine was chosen as the study area because it is one of the two main areas in Haiti with a significant number of CBT users. This rural area is located in the municipality of Gros-Morne, Haiti, and has the following geographical coordinates: 18.52°N and 74.34°W (Google Earth 2022). The average annual temperature is 24.8 °C (Jean et al. 2017). Grande Plaine has nearly 2,000 inhabitants distributed across 192 households, 35 of which, i.e. 280 people, use the CBT (Association des Originaires de Grande Plaine 2022). The health centers and reference hospitals in the region mentioned by Jean et al. (2017) revealed that typhoid, gastroenteritis, and intestinal parasitosis are the most frequent diseases in the region (especially among children). In addition to these pathologies, cholera is a sporadic epidemic in the region.

Out of the 35 aforementioned households, 33 use a community composting platform to compost their feces and two use their own composter (Association des Originaires de Grande Plaine 2022). This composting platform consists of nine compost bins, including five community and four individual, which compost fecal sludge from CBT user households throughout the year (Association des Originaires de Grande Plaine 2022). The platform and the composting process are described by Jean et al. (2017) and Jean (2018).

This study focuses on nine pathogens. These pathogens were selected based on the following criteria in accordance with Westrell (2004), Schonning et al. (2007), and WHO (2022): (i) prevalence in Haiti, (ii) presence in feces, (iii) pathogenicity, (iv) ability to survive in the environment after excretion, and (v) availability of data (especially those related to the dose-response model) to allow their integration into a QMRA study. The target pathogens are Ascaris lumbricoides, Campylobacter spp., Cryptosporidium parvum, E. coli O157:H7, Giardia intestinalis, poliovirus, Salmonella spp., Shigella spp., and V. cholerae. The health problems generated by this organism are ascariasis, campylobacteriosis, cryptosporidiosis, hemorrhagic diarrhea, giardiasis, poliomyelitis, salmonellosis, shigellosis, and cholera, respectively (Feachem et al. 1983; Mara 2004). For the purpose of this study, Ascaris eggs were considered (not the worms).

The data on fecal pathogen content of feces are drawn from the scientific literature (Table 1). Most of these studies used the quantitative polymerase chain reaction (qPCR) method to quantify target pathogens in feces.

CBT users collect feces in approximately 20-L buckets, which are usually emptied once a week. In each household, sludge is manually emptied by an adult from the household (referred to as ‘emptier’ in this study) who carries the bucket of sludge to a community composting platform located approximately 200 m from the house.

The population exposed to microbial risks mainly includes emptiers and master composters. Farmers, who spread compost on their fields, as well as the potential consumers of the products grown, were excluded from the scope of this study. Thermophilic (co)composting is supposed to sanitize fecal sludge because of the increase in temperature during the second phase (Berendes et al. 2015; Jean 2018), which implies that the health risk can be considered negligible. The main known exposure routes are accidental ingestion, inhalation of bioaerosols, and skin contact. However, due to the absence of a dose-response model for the latter two exposure routes, only ingestion was considered.

The exposure scenarios considered the most plausible were those where the operators' hands were contaminated and accidentally brought to the mouth either directly or indirectly through eating, drinking, hand-to-mouth contact, or nail-biting. Two scenarios were developed: (i) contamination of emptiers during transport, unloading, and washing of the feces bucket and (ii) contamination of master composters while handling sludge during the composting process.

Two hypotheses were formulated based on previous studies to estimate accidental ingestion of feces by emptiers and master composters. The following values were considered, in cases where the operators did not sufficiently use personal protective equipment (PPE) in the course of their work:

• emptiers inadvertently ingest about 0.033 g of feces per operation (Schonning et al. 2007), which usually occurs once a week;

• master composters accidentally ingest between 0.05 and 0.48 g of feces per operation (Gerba et al. 2002; Brooks et al. 2012; Gholipour et al. 2020; Sadeghi et al. 2022). These operations mainly consist of turning the piles approximately once a week throughout the year (i.e. 52 weeks).

The parameters α, β, and r are specific to each of the pathogens considered. The values chosen for each of these organisms are presented in Table 2.

The data demonstrate variability in the ingestion of feces by master composters, and the concentrations of G. intestinalis and V. cholerae in the feces were subjected to a Monte Carlo simulation to address the inherent uncertainties. The @RISK software, version 8.4.0 developed by Palisade Corporation, was used for this purpose. Log-normal distribution was chosen as the appropriate probability distribution, in accordance with Schonning et al. (2007). A total of 10,000 iterations were executed. The median (50th percentile) was utilized for result interpretation, representing the realistic scenario, while the 95th percentile was employed to represent the pessimistic scenario. The variables selected for uncertainty analysis are listed in Table 3.

In the context of this study, n = 52 for emptiers, since the bucket is emptied once a week on average. Similarly, for the master composters, n = 52, as they work throughout the year at a frequency of once a week.

A risk classification model inspired by the work of Westrell et al. (2004) was used to facilitate the interpretation of the study results. This model classifies the risks as insignificant, minor, moderate, major, and highly elevated, in accordance with Table 4. Insignificant risk corresponds to the level of acceptable risk mentioned in the Introduction, which is equivalent to 10⁻⁴ per person per year (10⁻⁴ pppy).

The calculations carried out using the selected dose-response functions and the ingestion doses per exposure made it possible to determine the probability of infection associated with each operation. These results are summarized in Table 5.

These results show that operators are highly exposed to a risk of infection if basic precautions are not taken. For emptiers, the highest risks are related to poliovirus (100%), V. cholerae (approximately 65%), A. lumbricoides (nearly 45%), Shigella spp. (nearly 37%), and G. intestinalis (nearly 28%). For master composters, in the realistic scenario, the highest risks are linked to poliovirus (100%), G. intestinalis (approximately 82%), V. cholerae (approximately 78%), and C. parvum (nearly 58%). In the pessimistic scenario (95th percentile), the highest risks were associated with poliovirus (100%), G. intestinalis (100%), C. parvum (nearly 94%), V. cholerae (approximately 86%), Shigella spp. (approximately 65%), and A. lumbricoides (nearly 60%).

The most likely pathogens to cause infection during an operation are ranked in descending order as follows: poliovirus > V. cholerae > G. intestinalis > A. lumbricoides > Shigella spp. This ranking is consistent with the information provided by Jean et al. (2017) on the most prevalent pathologies in the region, which are presented in the section ‘Presentation of the Study Area’.

The results show that the yearly probability of infection is, on average, two times higher than the probability of infection per operation, equal to 1.00 for all the pathogens considered, except for E. coli O157:H7 and Salmonella spp., where it is 9.25 × 10⁻¹ and 8.28 × 10⁻¹, respectively, among emptiers.

The yearly risk of infection was 8,281.36–10,000 times higher than the acceptable level of risk (10⁻⁴ pppy) depending on the pathogen considered. However, it should be noted that infection does not necessarily lead to illness. The probability of illness following a given infection depends on a range of factors, including age, the host's immune system status, and previous exposure to other pathogens (De Giudici et al. 2011; U.S. EPA 2012).

The calculated values of the probability of illness per exposure are summarized in Table 6.

Analysis of these data revealed that the probability of illness per operation was 2.58–1,071.19 times lower than the probability of infection per operation. During each operation, the emptiers and the master composters were exposed to two major risks, namely those related to poliovirus and V. cholerae; other risks were considered moderate and/or minor. The pessimistic scenarios indicate that master composters were exposed to two highly elevated risks of illness, approximately 39% for poliovirus and 25% for V. cholera, and two major risks, almost 8% for A. lumbricoides and nearly 9% for Shigella spp. Therefore, the most likely pathogens to cause disease are ranked as follows: poliovirus > V. cholerae > Shigella spp. > A. lumbricoides.

The results of the annual probability of illness presented in Table 7 show that the yearly probability of illness was 1.16–24.87 times higher than the probability of illness per operation and 2.58–1,071.19 times lower than the annual probability of infection. Emptiers and master composters were only exposed to two major risks, which were related to poliovirus and V. cholerae; the risks associated with the other target pathogens were, overall, classified as moderate and/or minor, ranging from 0.08 to 2.58%. It is observed that master composters were 2.18–8.48 times more exposed to microbial risk than the emptiers. This confirmed that master composters are the most exposed to microbial risk. As for the risk per operation, the most likely pathogens to cause disease were poliovirus, V. cholerae, Shigella spp., and A. lumbricoides. It is noteworthy that the results of the pessimistic scenario are identical to those of the realistic scenario in terms of the annual probability of illness.

The QMRA is now recognized as an important tool for decision-makers in preventing infections and/or infectious illnesses related to water, excreta, and food. This tool has some limitations, such as not taking into account the potential immunity of a portion of the exposed population. However, it is important to note that its purpose is not to determine the quantity of infection and illness in a given area, but rather to assess the probability that infection and illness could occur in the area based primarily on available microbiological, epidemiological, and demographic data. This is precisely the perspective from which the QMRA was used in this study to quantitatively evaluate the potential microbial health risks associated with the use of CBTs.

Like any study of this type, this study is subject to uncertainties. Data on the concentration of pathogens in feces were not collected in Grande Plaine but mainly came from previous studies that were not carried out in Haiti. Data specific to the study area would be more relevant. Furthermore, the equation used in the QMRA framework assumes that the ingested dose is the same at each exposure and does not take into account the fact that some people who have previously been infected with certain pathogens may become immune to them (Health Canada 2019). In reality, the most vulnerable people are generally those who are immunocompromised (people with AIDS and others), seniors, pregnant women, infants, and people suffering from malnutrition (Haas et al. 1999; De Giudici et al. 2011; U.S. EPA 2012). Epidemiological and demographic data on the area (population health status, age groups, number of pregnant women, etc.) would allow for the identification of the most vulnerable groups and a more exhaustive analysis of the situation, but such data are not available in the existing literature.

Microbial risk management primarily falls under the jurisdiction of political authorities, namely the Ministry of Public Health and Population (MSPP). The MSPP is responsible for developing and enforcing barrier measures in accordance with the sanitation approaches adopted in the country to protect the health of the population. Implementing adequate hygienic and sanitary measures can significantly reduce the microbial risks associated with the use of CBTs. The measures adopted must prevent any contact with feces, such as the use of PPE (gloves, boots, and protective masks) and hand washing. Another way to prevent contact with feces is to reduce the concentration of pathogens in the feces by removing the full bucket from the CBT and allowing the sludge to dry, while another bucket is put into service right next to it. The full bucket would be covered with ash and a lid to reduce the nuisance caused by odors and harmful insects.

The study results show that the risk of infection by most of the targeted pathogenic organisms is high among operators (emptiers and master composters), which is not the case for the risk of illness. The results highlight the fact that the risks of illness associated with poliovirus, V. cholerae, and Ascaris are generally the highest, while those associated with E. coli O157:H7 and C. parvum are the lowest. The annual infection risks were found to be 8,281.36–10,000 times higher than the established acceptable risk level, while the annual disease risks ranged from 8.64 to 3,870.28 times higher than the acceptable risk level, depending on the pathogen considered. However, these results do not necessarily mean that the operators in question will be infected and/or fall ill, but rather illustrate what could happen if they do not take necessary precautions during their usual operations.

The obtained results suggest that political authorities should develop guidelines in this regard to ensure a safer use of the technology. This requires training and raising awareness in the population concerned, either by public authorities or local organizations mandated by the authorities. These actions would be a suitable lever for implementing barrier measures and self-protection mechanisms for regular monitoring of feces-composting operations to ensure that established guidelines have been respected and that the compost produced is truly sanitized.

The lead author would like to thank the Association des Originaires de Grande Plaine (AOG) for providing information about the study area with regard to the use of CBTs.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.