ABSTRACT
In this study, we use water quality tests, epidemiological data, and ethnographic interviews gathered in Kathmandu in 2017 to provide an interdisciplinary analysis of water insecurity, illness, and targeted responses. Our findings fit the pattern of the past 20 years: the concentration of coliform bacteria across water sources remains unsafe. Our epidemiological data documents high rates of water-borne infectious diseases consistent with fecal contamination of water sources. Our ethnographic interviews suggest that social marginalization, poverty, and dislocation are major drivers in the incidence of illness. Evaluating some of the collective responses to water insecurity and contamination, we argue that effectively addressing water pollution, scarcity, and health outcomes in Kathmandu requires redirecting the focus of programs and the primary methods of implementation. Instead of large-scale top-down infrastructural interventions such as the Melamchi Water Supply Project, programs should target women's leadership in and subsidize funding for small-scale, community-managed interventions.
HIGHLIGHTS
The partnership which supplies the Kathmandu Valley is unable to provide half of the water needed by residents.
Concentrations of coliform bacteria across all water sources are unsafe.
Water users suffer widespread health impacts from exposure to contaminated water.
Water insecurity is severe and widespread.
We suggest redirecting resources to support women's leadership and community management of water.
INTRODUCTION
Access to affordable and safe drinking water is a crucial determinant of human health and well-being. Recent estimates show, however, that more than a quarter of the global population faces water scarcity (Sengupta & Cai 2019). And water scarcity severely limits possibilities for sustainable and equitable growth globally (https://sdgs.un.org/goals). Recently, public health advocates have drawn attention to the need for integrating interdisciplinary approaches in responding to health crises to promote equity in the distribution of resources and social justice in the organization of health services (Janes & Corbett 2009). In particular, within epidemiology, there is a need to address social and political inequalities as drivers of disease (Farmer 1999, 2003). High levels of social and economic inequality correlate with worse health outcomes both nationally and internationally (Nguyen & Peschard 2003; Nichter 2008). And vicious cycles have been well documented, for example, ill health due to contaminated water is a main driver of poverty and poverty is in turn linked to living in areas of contamination (Prakash Saravanan & Chourey 2012). Facing gendered expectations about household work and reproductive labor, women are particularly and differentially impacted by water insecurity (Davis 2006; Sultana 2009). Across Asia, water-borne diseases and contamination may be driven by interlinked socio-economic effects of rapid deforestation, industrialization, urbanization, and land reclamation due to agricultural intensification and extraction, as well as waste production (Orlove & Caton 2010).
Because of its geographic position Nepal is a key site for analyzing problems of water access, sustainable development and social justice. Nepal is one of the ‘water towers’ of Asia and, given its centrality, recent years have ushered in a proliferation of competing local level, state, NGO, and corporate projects aimed at generating power and profits through water. Exacerbated by climate change, these competing projects have many-layered effects on people across Asia. Nepal sits in the middle of the Himalayan water-energy-food nexus which affects the livelihoods of over one billion people (Rasul 2014). At the same time, the Nepali state is reforming after decades of civil conflict, a devastating earthquake in 2015, and failed privatization and liberalization initiatives. State reform goes hand in hand with hydropower to the extent that one analyst argues that ‘the production of the hydropower future ensures the economic and political coherence of the state, and vice-versa’ (Lord 2016, p. 150). However, studies of Nepali state policies and reforms, often overlook the most effective interventions in water access and affordability. Here, we examine social conditions and biological vectors of contamination in the urban center of Kathmandu in Nepal, as a case study of possibilities for promoting water equity.
In this study we draw together water quality tests, epidemiological data, and ethnographic interviews gathered in Kathmandu in 2017, to provide an interdisciplinary analysis of water insecurity, illness, and targeted responses. Based on our findings, we argue that effectively addressing water pollution, scarcity and health outcomes in Kathmandu requires redirecting the focus of programs and the primary methods of implementation. Instead of large-scale top-down infrastructural interventions, the government should target women's leadership in and subsidize funding for small-scale, community-managed interventions such as those already led by Water User Sanitation Committees.
With an estimated population of 3 million, Kathmandu is at the center of one of the most rapid urbanizations in South Asia (Elsey et al. 2016, p. 2). And rapid urbanization places pressure on available resources such as water and sanitation. Currently, the water supply network is operated by a public-private partnership, the Kathmandu Upatyaka Khanepani Limited (KUKL), which is unable to provide even half of the water needed by residents. Because KUKL cannot yet meet household water demand, all of the urban areas of Kathmandu are characterized as water-insecure, e.g. a 2016 survey found that 75% felt they were facing a water crisis (Molden et al. 2016). Given supply problems, residents of Kathmandu access water through multiple systems (often marking bottles for different uses within a household). KUKL sells water at subsidized rates through piped household connections, public stand posts, and water tanker suppliers. And private tanker trucks sell water at market prices – estimated as ten times higher than those of KUKL – determined by seasonally impacted supply and demand and source (Shrestha & Shukla 2014, pp. 261–262). Many households also have tube wells or ponds used for drinking water. Others access water through the traditional stone spout system (which is often free of charge). Installed 1,500 years ago, the stone spout system served the whole valley and, though many are in disrepair, some stone spouts remain accessible in public areas (Pradhan 2012; Molden et al. 2016). Many neighborhoods experience frequent shortages and disruptions in water supply (Joshi & Maharjan 2003; Pradhan 2012; Shrestha 2014) and, when water becomes available it often flows for a short time and may be contaminated.
KUKL currently operates reservoirs and treatment plants to filtrate and disinfect the water that is piped to residential areas. However, in 2017 the director KUKL informed us that less than fifty percent of residents in Kathmandu receive treated water. KUKL also provides water delivered by trucks to ‘Hilltake’ tanks located in areas with no taps. But, according to their 2012 ‘KUKL at a Glance’ information pamphlet, KUKL operated only 27 tanker trucks. Private water tanker trucks also sell water and most people we interviewed reported using water from private trucks. Many of our interview participants also believed the water delivered by the tanker was untreated. A migrant from eastern Nepal explained that he buys bottled water as well as using the stone spouts but uses groundwater for all washing. In his family, they make sure not to use river water but they do use water from the stone spouts and also from pumps which were made ‘collectively by the community. Water always comes out, but it is smelly and problematic to bathe in [Interview 2, Site 1].’ He also explained that his family often experiences gastrointestinal upsets. A middle-aged mother who works as a cleaner for a power company and lives in the same neighborhood explained that she only allows her family to drink bottled water because she does not believe the KUKL supplied water is safe to drink. Despite these measures, in her family, they had experienced dysentery, diarrhea and fever as well as ‘many stomach aches’ within the year [Interview 3, Site 1]. In Kirtipur, several residents stated that, while they buy water from the tanker trucks which arrive each day, they believe it is untreated groundwater and several mentioned that the entire neighborhood was affected by an outbreak of ‘severe stomach upset’ (Interviews 22, 23, Site 4) during which ‘foreigners came and tested a few years ago and determined it was unsafe to drink (Interview 14, Site 4).’ Although the director of the KUKL assured us that the tank water is treated, other suppliers may collect water unhygienically. In 2017, we observed tanker trucks filling directly from groundwater and others reported similar observations before the 2015 earthquake (Shrestha & Shukla 2014; Pant 2016, p. 261). To indicate that water has been treated, KUKL uses a color-coded sticker system displayed on water tanker truck bumpers. Unfortunately, residents stated that they are unaware of this sticker system and the majority of tanker truck operators remain unregistered.
Drinking water may be contaminated with pathogenic microbes at the source, through the supply system, or at the users' level (World Health Organization 2003). Rapid population growth, unsanitary disposal of waste into water, and increased human activities in and around water sources, environmental settings that are typical of developing countries, can all lead to contamination of water with pathogenic microbes. In Nepal, unscientific disposal of human and agricultural wastes in water and lack of sanitation has led to increased levels of microbiological contamination in streams, springs and ground sources (Joshi & Maharjan 2003). In areas with high levels of microbial water contamination, this may lead to a significant increase in the incidence and spread of water-borne illnesses. The pathogenic bacteria found in contaminated drinking water include Salmonella, Shigella, E .coli 0157:H7, Vibrio cholerae, Yersinia enterocolitica and Campylobacter sps (Resource Network Nepal 2016). These organisms may cause diseases ranging in severity from mild gastroenteritis to severe and sometimes fatal dysentery, cholera or typhoid.
Nepal faces problems regarding both its drinking water quality and supply and throughout Nepal, people are exposed to severe health threats resulting from water contamination by sewage, agriculture, and industry. Owing to the impact of sewage, typhoid, dysentery, and cholera are endemic every summer (Khadka 1993). These diseases have accounted for 15% of all illnesses and 8% of total deaths, but those numbers increase to 41% of all illnesses and 32% of all deaths in children up to 4 years old (Sharma 1990). In the Kathmandu Valley, the chief concern is contamination from sewage lines, septic tanks, open pit toilets (Jha et al. 1997), and from surface water that has been polluted by direct disposal of sewage waste (Khadka 1992; Karn & Harada 2001; Shrestha et al. 2017).
In addition, school sanitation and water provision are another major vector for the transmission of disease among children. Numerous studies have been conducted which highlight the need for improvement of drinking water quality, sanitation, and hygiene conditions in schools. In Kathmandu, less than half of people were found to treat their drinking water themselves and water supplies tested in the months of May, June, and July were the most heavily contaminated. As discussed above, the quality of drinking water in Kathmandu has been deteriorating due to the deficiency of treatment plants, direct discharge of sewage waste into surface water, and inefficient management of the piped water distribution system. Currently, approximately 50% of the water supply that is used for drinking and other domestic requirements in the Kathmandu Valley is derived from groundwater (Tanaka et al. 2012, p. 170).
METHODS
Water sampling and collection
Water sampling, interview, and epidemiological data collection sites in Kathmandu Map obtained from Shah & Shah (2013), p. 1125, were modified to include our sampling sites.
Water sampling, interview, and epidemiological data collection sites in Kathmandu Map obtained from Shah & Shah (2013), p. 1125, were modified to include our sampling sites.
Epidemiological data collection
With permission, we gathered hospital intake registries for the prior 60 days (from the inception of our research on June 1, 2017) at major hospitals and one clinic located in the neighborhoods where we tested the water (Figure 1). These registries provide a record of the prevalence of gastrointestinal infections in Nepal at-large as well as within Kathmandu during the months from mid-March to mid-May of 2017. Because this data were recorded during the dry season in Nepal, these months provide a baseline for understanding and interpreting the incidence of water-borne disease in this rapidly expanding city. During this season, the volume of water decreases and the contamination level increases as does the bacterial count in comparison to the winter season. In the rainy season, the volume of water increases and bodies of water become more contaminated due to sewage wastewater and runoff flowing into tributaries (Joshi & Maharjan 2003). As water sources become more contaminated, it is expected that the incidence of water-borne diseases will also increase during rainy seasons. Our first site, Teku Hospital, was chosen because it is the major infectious disease hospital in Kathmandu. The second site, Bir Hospital, was chosen because it is the largest government hospital that provides low-cost medical care to patients from all over the country. The third site, Bhaktapur Hospital, primarily serves the population in the surrounding neighborhood along the Manohara River. Because this is an agricultural, peri-urban area, we argue that this data provides a window into point and non-point source contamination along this tributary of the Bagmati. Finally, we collected data from the Patan Clinic, which serves populations in the Lalitpur district and is bordered on three sides by the Bagmati River. We argue that, due to this geographic location as well as the cheaper cost of care, this clinic provides a cross section of the local Kathmandu population overall and therefore a window into contamination across the city. Hospital intake registries were then translated and coded for the geographic location of patients, gender, age cohort, and type of infection. We then interpreted the distribution and incidence of water-borne diseases over these months in relation to the population served by each hospital.
Water user interview data collection1
We conducted 60 interviews with individuals living in permanent and semi-permanent shelters in the neighborhoods of our water testing and epidemiological data collection (Figure 1). In these interviews, we paid attention to the social determinants of water scarcity, in particular to the effects of structural inequalities by asking questions regarding health and illness within families, work and workplace experiences, living conditions, water use and access, and migration. Interviews were conducted in Nepali and with consent by the team, pairing US students with Nepali students who translated and assisted in write-up and interpretation. Interviewees were selected through initial site surveys as well as through snowball sampling. Selection criteria included age, occupation, gender, caste, and proximity to our water sampling sites. We focused on tea shops, temples, malls, and other public gathering places (such as wasteland used for play and recreation) as prime locations for gaining representative views among affected communities. Our interview data is not, therefore, random but we argue that it provides a critical window into uneven development within the rapidly expanding and much displaced population of Kathmandu.
RESULTS AND DISCUSSION
Of the 27 river samples we collected from various sampling sites along three major rivers traversing the valley, 100% produced positive results in the Most Probable Number (MPN) test for the presence of coliforms in water. The average number of coliforms in the river water samples was found to be ≥1,100 ± 0 CFU/100 ml, the maximum detection level of the MPN test, thus indicating a high level of contamination of the rivers with fecal bacteria (Table 1). The detection of bacteria of fecal origin in water indicates that the water is contaminated with other potentially pathogenic bacteria of fecal origin, and hence poses a risk of infection to people who use such contaminated water for drinking, household or other uses that can directly or indirectly bring them in contact with the pathogens in the water. Of the 16 tap and tube well water samples collected from areas around the rivers, 31.25% (five samples) were found to be contaminated with coliforms. The number of coliforms in the positive samples ranged from 3.0 to >1,100 CFU/100 ml. This is in agreement with the average E. coli CFU determined by a previous study in tube well water samples collected from various locations in Kathmandu (Joshi & Maharjan 2003). While our tap and tube well samples seem to be relatively safer for use than river water, subsequent studies showed that the concentration of coliforms across all water sources (groundwater, municipal tap water, tanker truck) remains unsafe for drinking purposes (Bhandari et al. 2021). And recent studies show that the groundwater in the Kathmandu Valley has significant microbial contamination that exceeds the standard recommended by WHO (Ghimire et al. 2023). Currently, water samples from wells between 30 and 3 m depth as well as tap water samples that have been analyzed by researchers do not meet the standards for the Nepal national drinking water quality guidelines and thus are unacceptable sources for drinking purposes (Pant 2016; Bhandari et al. 2021). In addition, Shrestha et al. have shown that the water is also microbiologically unsafe for cleaning vegetables and irrigating plants (2017). It remains a common practice in Kathmandu Valley to pump water from polluted rivers to irrigate vegetables and to use well, river, or tanker water sources to wash raw vegetables.
Coliform counts in the water samples
Samples . | Coliform counta . | Number of samplesb . |
---|---|---|
Manahara River | ≥1,100 | 9 (100) |
Bishnumati River | ≥1,100 | 6 (100) |
Bagmati-Bishnumati confluence | ≥1,100 | 3 (100) |
Bagmati River – upstream from confluence | ≥1,100 | 6 (100) |
Bagmati River – downstream from confluence | ≥1,100 | 3 (100) |
Drinking water (tap, tube well, tank water)c | 3 to >1,100 | 5 (31.25) |
<3 | 11 (68.75) |
Samples . | Coliform counta . | Number of samplesb . |
---|---|---|
Manahara River | ≥1,100 | 9 (100) |
Bishnumati River | ≥1,100 | 6 (100) |
Bagmati-Bishnumati confluence | ≥1,100 | 3 (100) |
Bagmati River – upstream from confluence | ≥1,100 | 6 (100) |
Bagmati River – downstream from confluence | ≥1,100 | 3 (100) |
Drinking water (tap, tube well, tank water)c | 3 to >1,100 | 5 (31.25) |
<3 | 11 (68.75) |
aNumbers connote the most probable number (MPN) of coliform CFU/100 ml sample.
bNumbers in parentheses represent percentages of results out of the total samples collected from the corresponding sites.
cCollected from neighborhoods located along each of the river sites.
Distribution of cases of potential water-borne illnesses among male and female patients seen at (a) Teku Hospital, (b) Bir Hospital, (c) Bhaktapur Hospital, and (d) Patan clinic, from mid-April to mid-June 2017 (over a period of 60 days). Striped bars represent females and solid bars represent males. Gray areas represent cases of potential water-borne illnesses.
Distribution of cases of potential water-borne illnesses among male and female patients seen at (a) Teku Hospital, (b) Bir Hospital, (c) Bhaktapur Hospital, and (d) Patan clinic, from mid-April to mid-June 2017 (over a period of 60 days). Striped bars represent females and solid bars represent males. Gray areas represent cases of potential water-borne illnesses.
Cases of infectious diseases of potential water-borne origin among different age cohorts reported at (top): Teku Hospital (blue), Bhaktapur Hospital (Red), Patan Clinic (Green), and (bottom): Bir Hospital (Black) from mid-April to mid-June 2018. Striped bars represent cases in females and solid bars represent cases in males.
Cases of infectious diseases of potential water-borne origin among different age cohorts reported at (top): Teku Hospital (blue), Bhaktapur Hospital (Red), Patan Clinic (Green), and (bottom): Bir Hospital (Black) from mid-April to mid-June 2018. Striped bars represent cases in females and solid bars represent cases in males.
Based on the hospital record, the individuals admitted for treatment came from a wide range of locations within the country. Therefore, a particular geographical focal point of potential water-borne infection for the entire population could not be identified. However, for the patient population based in Kathmandu, a potential source of infection can be proposed. The hospital record showed that all infected individuals from Kathmandu were living in close proximity to the Bishnumati and Bagmati rivers at the time of the infection. Furthermore, most of the infected individuals lived at locations very close to the sampling sites tested in this study. As revealed by the MPN tests, the water in these rivers contains high levels of fecal bacteria and thus poses a potential health hazard to the people living in areas close to the rivers. It is likely that these patients contracted the disease by direct or indirect exposure to the contaminated water from the river or the groundwater sources. In a prior analysis of the microbial pathogens distributed in the groundwater pumped from deep tube wells in the Kathmandu Valley, Teku Hospital reported that 16.5% of all deaths were due to complications related to water-borne diseases (Tanaka et al. 2012, p. 171). The results of this study revealed that the groundwater contained fecal microbes and a variety of aerobic microbes including C1 compounds-utilizers. This suggests that groundwater pollution due to human activities is an ongoing challenge in the valley. As discussed above, school water and sanitation continue to be challenges in Kathmandu and we believe they form a major disease vector.
Interestingly, despite having a broad patient population, the hospital in Teku did not have any patients from the Bhaktapur district. The epidemiological data from the Bhaktapur district hospital showed considerable numbers of cases from the local area registered at this hospital. Considering that Bhaktapur Hospital would be a closer commute for the patients living in Bhaktapur or surrounding areas close to Bhaktapur, it is reasonable to speculate that the local people preferred getting medical service at the Bhaktapur Hospital. As Bhaktapur Hospital also offered discounted or free checkup services for the residents of the district, this could have influenced the decisions of these people to seek treatment at this hospital.
Bir Hospital data show that from Mid-March to Mid-May, 2017 (over a period of 60 days), a total of 68,718 cases of diseases were documented at Bir Hospital. Of these, 3,756 cases were registered in the in-patient (148 individuals) and out-patient (3,608 individuals) units of the gastroenterology department. The addresses of the patients were not provided in the report. Bir Hospital is the largest government hospital that provides medical services to patients from all over the country. Considering this and the number of total cases recorded in the in-patient and out-patient registries it can be inferred that the patient population was from a larger geographical area. Of the total cases documented in the gastroenterology department, 1,989 patients were diagnosed with infectious diseases of likely water- or food-borne origin. Of these, 1,214 (64% of male patients) were males and 775 (41.8% of female patients) were females (Figure 2(b)). The age-based distribution of cases in male and female patients is shown in (Figure 3). Considering the diagnoses of acute gastroenteritis, acute hepatitis, diarrhea, enteric fever and typhoid fever, the age-wise distribution of the total number of typhoid fever cases was noticeably higher in both males (10) and females (4) compared to other disease categories. This is consistent with the data from Teku Hospital and other studies which found that, as a result of the earthquake, the people of Kathmandu are at a higher risk than ever before ingesting contaminated water (Pant 2016).
With regard to the age-based distribution of the water-borne illnesses identified above, the highest number of cases for male patients (six cases) was identified in the age cohort 15–19 years, followed by 20–29 years (five cases). However, only one female case was identified in each of these age cohorts. Although the numbers seem to be considerably different in males and females, the number of cases is not large enough to establish a statistically significant difference. Our interviews, however, showed that only in the most serious cases of illness will patients seek treatment at hospitals. Problems of healthcare access became evident when we visited the Teku Hospital in the Thapathali neighborhood which was crowded with patients waiting for many hours. Many people interviewed reported only going to the hospital if their illness was very serious. Instead of seeking to access medical care, they relied on pharmacies, home remedies, and/or ignoring pain or infections. In addition, patients perceive a significant difference in the quality of care between government and private hospitals. For many families we interviewed, the severity of illness corresponds to where families seek treatment. A pharmacist for example stated that it is very common for people to come to the pharmacy seeking treatment for typhoid, noting that everyone in that community has gastric problems (Interview 9, site 1). In our interviews, many women in particular stressed the difference in quality between private and public hospital care (Interviews 6 and 16, Site 1). One young mother, who said her family has stomach aches very often, explained that she takes her child to a private hospital but all other family members only visit the government hospital (Interview 2, site 2). Our interview data shows that cost barriers are yet another contributing factor amplifying the cycle of contamination that puts individuals at risk for water-borne illnesses. As discussed below, risks are particularly exacerbated for recent migrants.
Bhaktapur Hospital data show that from Mid-April to Mid-June, 2017 (over a period of 60 days), a total of 210 people including 86 (40.95%) males and 115 (54.76%) females between the ages of 14 and 95 years were admitted into Bhaktapur Hospital for treatment for various ailments. Of these, 13 (15.11%) males and 17 (14.78%) females were diagnosed with infectious diseases of water or food-borne origin (Figure 2(c)). Notably high numbers of enteric fever cases were identified in both males and females (12 and 11 cases, respectively). The highest numbers of cases were identified in the age cohort 20–29 years for both male and female patients (four and five cases, respectively) (Figure 3). Notably, four female cases were reported but no male cases were reported in the age group 50–59 years.
The patient population of this hospital was fairly homogenous in distribution. Based on the hospital record, most cases were from within the Bhaktapur district. A few of the cases were from further east, including two cases from as far as Ramechhap district located 132.8 km east of Bhaktapur district. With regard to the distribution of the potential water-borne diseases, all identified cases were individuals living in proximity to locations within Bhaktapur that are close to river water sources; most of the addresses are at short distances along the Manohara river belt. Like the Bishnumati and Bagmati rivers, the Manohara River runs amidst polluted urban areas in the city with high levels of human activity in and along the banks of the river. Many farmers who bring vegetables and fruits from Bhaktapur to sell in Kathmandu use the water of the Manohara River both to irrigate fields and to wash the produce (Shrestha et al. 2017). Our interviews show that this is an ongoing practice and an important vector for contamination. For example, two vegetable farmers explained that they would rather use river water because the chlorine from tap water would change the color of vegetables and make them unappealing to consumers (Interview 1, Site 5).
Patan Clinic data show that from Mid-April to Mid-June, 2017 (over a period of 30 days), a total of 790 people including 281 (35.56%) males and 509 (64.43%) females between the ages of 5 and 90 years were registered in the Patan Outpatient Clinic in Lalitpur district of Nepal. Additionally, 78 patients aged 0–5 years were examined in the pediatric department of the clinic. The registry that was provided for data recording did not disclose the sex of the pediatric patients. Of the people examined at the clinic, 58 (20.6%) males and 74 (14.5%) females were confirmed to have diarrheal illnesses and parasitic intestinal infections frequently caused by consumption of contaminated food or water (Figure 2(d)).
The highest numbers of cases were identified in the age cohort 20–29 years for male patients and in the age cohort 30–39 years for female patients (14 cases and 21 cases, respectively) (Figure 3). The distribution was similar in both sexes in the age cohorts 5–14 years and 20–29 years. However, a notably larger difference was observed in the age cohort 30–39 years; the number of female cases reported (21) was almost double the number of male cases reported (11). This pattern was also observed in age cohorts 40–49 years and 50–59 years though the numbers of cases were much smaller than the younger age cohorts. Based on the clinic record, all documented cases were from the Lalitpur district. The record did not specify the particular address (town, street etc.) of the patients. In the Kirtipur neighborhood several residents mentioned that they prefer to visit Patan because it is private and therefore, in their view, safer. However, as this is an out-patient clinic, it can be expected that the majority of people visiting the clinic were those living in its vicinity and did not seek or need immediate hospitalization at the time of their visit. Lalitpur district spans an area of 385 km² and is bordered on three sides by the Bagmati River. The river further branches into smaller rivers that traverse the outskirts of the district. Therefore, it can be assumed that people in this district are in close proximity to river water sources regardless of their specific location within the district.
The out-patient pediatric record did not specify the sex of the children aged 0–5 years, hence no inference can be drawn on the distribution of the disease in children under 5 years of age. Nevertheless, the incidence of water-borne diseases in this patient population was noticeably high (10.81%). A high but similar distribution in males and females in the age cohort 5–14 years was observed. This could indicate an equally susceptible population of children under the age of 15 years regardless of their sexes. An explanation for this could be that this category includes children of school going age. Regardless of their sexes, the children would have equal chances of exposure to the pathogens that they are likely to encounter via the use of unsanitary or inadequately managed water supply in schools (Tandukar et al. 2013; Ram Dhital & Koirala 2016; Yadav & Prakash 2016). Based on the recommended treatment for the diagnosis, the diarrheal diseases were likely protozoal (Giardiasis or Amoebiasis) and the worm infections were most probably intestinal roundworm infestations. Similar results were reported among schoolchildren in Burkina Faso (Erismann et al. 2016). The survey revealed a lower prevalence of helminthic worms than intestinal protozoa infections, which the authors attributed to the national and regional-level deworming campaigns of 2004 and 2014. However, despite these efforts, no interruption in transmission of these infections was observed in the area. Almost 35% of people, primarily children, in Nepal take medicine against worm infections. Another notable observation of the study by Erismann et al. (2016) is that the modality of drinking water storage and children's water exposure through play and domestic chores were not significantly different among the schoolchildren. Rather, strong associations were found between the children's parasitic infection status and handwashing, sanitary, and hygiene behaviors. Based on our observations, the prevalence and transmission of helminth and protozoal infections in schoolchildren in Nepal is likely determined in part by the socio-ecological contexts that determine the prevalence and transmission of food and water-borne illnesses such as proximity to the river, sanitation, and hygiene behaviors.
Though there were some differences in the number of cases of water-borne illnesses between the age cohorts of the patients as well as among males and females, overall, men were more likely to be diagnosed in a hospital or clinic for water-borne diseases or typhoid. In all locations, the majority of people infected were above 15 years of age, but most of the cases occurred between the ages of 20 and 49 years. It is significant to note that Patan Clinic had a large number of children between the ages of 0 and 5 years who were registered for pediatric treatment, indicating a high number of infections in school-aged children.
In our data, adults are more likely to be diagnosed with a water-borne disease than younger children. Most people explained that they boil their water for about five minutes or filter it as an extra precautionary step. This is in direct contrast to a 2013 study of public understandings of health risks in slums across Kathmandu where the majority of participants viewed frequent bouts of diarrhea as deriving from the quality and freshness of food rather than from contaminated water (Elsey et al. 2016). We noted that parents preferred not to give water obtained from the municipality or tank trucks to their small children. This explains why very few incidences of water-borne illness occurred in children under the age of 15 years. The adults in our study, however, reported that they would drink tap and tank water after boiling and would use ground water for cleaning and washing. Adults thus have a higher probability of coming into contact with contaminated water at home. On the other hand, children who attend under-resourced schools are exposed to an increased likelihood of infection due to water insecurity and inadequate sanitation. Because parents are not there to watch over their children, they are more likely to come into contact with contaminated water through the water supply from schools, improper hand washing, etc. This increases their chance of becoming infected with a water-borne or parasitic infection and may explain the high incidences of young children receiving care from Patan clinic.
Age cohort and gender differentiated patterns may also be due to some of the aspects of social marginalization described in interviews. For example, Bir Hospital had a much larger number of males than females who were diagnosed with infectious diseases of likely water or food-borne origin. As previously mentioned, Bir Hospital provides medical services for the whole country, thus it can be hypothesized that patients seeking medical attention come from a larger geographical range. A possible reason for the higher number of cases of infection in men may be related to socializing patterns. We noticed many under or unemployed men spending their free time in tea shops. Prior research shows that hotels and restaurants are the largest consumers of private tanker water (Shrestha & Shukla 2014) and this industry remains unregulated. Women cooking in their own households may be less likely to become sick because they are hyperaware of the source of water. Further, while many of the women we encountered worked in manufacturing or as domestics, many of the men interviewed worked in water-intensive industries such as construction and mining.
Our data show internal migration and dislocation as primary factors in the uneven spread of pathogens and illness. In a 2011 census, there were 63 slum and squatter settlements in Kathmandu especially along the Bishnumati, Bagmati, and Manohara rivers (Oli et al. 2013). And, in our interviews within slums along these rivers, we found that displaced families were well aware of the need to sanitize water but they were often unable to access sanitary water due to infrequent delivery and radically inflated prices. Our findings are parallel to other ethnographic studies around the world which show that the poor pay radically more for water (Davis 2006; Shrestha 2014; Wutich 2020). At all of the sites where we conducted interviews, we found a large proportion of migrants, renters, and squatters. In sites 1 and 2, we encountered the highest number of internal migrants, and most families had been in residence in Kathmandu for less than ten years. But many families were long-term residents, e.g. one refugee from Illam who moved 15 years ago (Interview 2, Site 1) and another migrant from Solukhumbu working at a local hospital for 20 years (Interview 7, site 1). Prior studies conducted in Western Nepal show that the migrants' main motivation is economic (Jaquet et al. 2016). However, migration is only an option for those who can afford it, and the most destitute cannot use this as a strategy for economic improvement. Across all neighborhoods, migrants spoke of precarity and fears of forced relocation. In Dhalko, one recent migrant explained that the only thing he wants is to be allowed to continue living in Kathmandu (Interview 4, Site 2). This is in keeping with other studies showing that rural-urban migrants may encounter difficulties not only in finding employment and housing, but also in accessing health care, education, and credit (Acharya & Leon-Gonzalez 2015).
In Site 4, Kirtipur, one older man originally from Dharan was particularly vociferous in blaming recent migrants for theft and violence in the neighborhood (Interview 3, Site 4). And this theme was brought up by many middle-class residents in each of the neighborhoods studied. Both owners and renters of permanent housing often framed migrants' behaviors as ‘dangerous’ and their bodies as ‘dirty.’ In response, some of our interviewees who were living in temporary shelters alongside rivers reported prioritizing washing water over drinking water. Our results above show that these choices can have very negative impacts on health and well-being. In our view, this points to social boundary demarcation between neighborhood residents, who view themselves as community insiders and stigmatize migrants as outsiders who do not belong in the community. In Kirtipur, for example, one older resident living in the area for over 25 years explained ‘they do not want us here, the land is already owned by residents and overseas workers (Interview 6, Site 4).’ In the same neighborhood, recent migrants who had been displaced by the earthquake were running small shops and farms, however, they were spoken of as outsiders. These boundaries of community belonging are clearly demarcated in order to frame migrants as transients rather than as community members who have legitimate claims on community resources. The residents we encountered in a squatter settlement along the Bagmati River in Thapatali know full well the difficulties of establishing community belonging, as they have lived along the river for over 20 years but residents in permanent housing continue to speak of them as recent, temporary, or earthquake refugees. Similarly, in Lalitpur in 2013, researchers found that residents demarcated water access between residents and migrants. In this case, local homeowners, usually ethnically Newar, claimed first rights of access to collecting water from stone spouts and this was a source of frustration for migrant residents, who preferred this water because it was usually good quality, but had to wait for many hours (Molden et al. 2016, p. 988). Overall, we found that social marginalization and dislocation contribute to the spread of water-borne infectious diseases by curtailing access to affordable uncontaminated water. For those on the margins, public water resources are difficult to access and private water resources are prohibitively expensive, and thus both poverty and ill health are concentrated in neighborhoods facing water scarcity. And, with such uneven development, water management institutions and water infrastructure often slip into pre-existing social and spatial divisions which in turn generate further vectors of contamination and disease.
CONCLUSIONS
Nearly 40 years ago the government of Nepal initiated a large-scale, long-term project known as the Melamchi Water Supply Project (MWSP). Initially financed and proposed by the World Bank (WB) as its flagship project in Nepal, the WB pulled out of the project amid the political upheavals of the 1990s. Subsequently, the Asian Development Bank (ADB) and the Japan International Cooperation Agency have provided financing. The project was designed to provide more water to the Kathmandu Valley by diverting water from three rivers in the Indrawati River Basin to the Kathmandu Valley. The first phase of the project diverts 170 million litres/ day from the Melamchi River through a recently completed 27 km mountain tunnel. However, the historic floods of 2021 buried the intake headworks for the Melamchi river diversion. Climate change, quake impacts and design flaws have all been blamed for this failure. And, although the Melamchi River is fed by glacier melt, the initial design of the MWSP did not account for increased flooding risks due to climate change. The initial project was a USD 400 million dollar scheme and, although it is unclear how much it will cost to clear the headworks, estimates are that another USD 400 million will be needed for project completion (Bhattarai et al. 2023). Meanwhile, the Kathmandu Post reported that 1/3 of the financing for MWSP was used to pay external consultants (Rest 2019, p. 1204). Hence many residents of Kathmandu see MWSP as a pipe dream (Molden et al. 2016) and many of our interviewees voiced similar feelings.
Many of our interviewees were displaced by the earthquake and a few indicated that they had received small compensation from the government for resettlement. All, however, were critical of the governments' infrastructural plans and policies, e.g. one respondent stated that Nepal was a failed state ‘there is no government in Nepal [Interview 2, Site 3].’ Across the city, residents are thus trying to address water scarcity and quality issues for themselves. Comparing two neighborhood-level experiments, Butcher found that the most significant criteria of success are: the economic conditions attached to the projects and the prioritizing of women's leadership within community resource initiatives. In the first case, she examined a community-managed water project, the Low Income Consumer Services Unit Program (LICSU), which was organized by a local women's federation group in partnership with KUKL in 2009. In this program on the east side of the Bansighat neighborhood, KUKL constructed and rehabilitated 100 community taplines and also installed new water tanks in public areas. Neighborhood residents organized the management of the tanks, with a particular emphasis on women's leadership and community participation. Two tanks were installed to provide free water for one year (provided by KUKL) and then subsidized to 50% of market prices and the water was then resold by community leaders to collective members who paid a set fee for a fixed amount of drinking water. This project was so successful that over time LICSU leaders were able to push for formal representation on the utility board. The second project, installed on the west side of the neighborhood, was less successful. In this case, KUKL did not subsidize the infrastructure or provide water at reduced rates. West side residents were required to contribute almost half of the infrastructure costs. In consequence, only a few families contributed and they controlled management and access and operated the water tanks as a for-profit business, rather than a collective venture (Butcher 2021). Excluded west side residents commented ‘we don't think we can be part of developments in this community (Butcher 2021, p. 85).’ In this case, infrastructural interventions mirrored the pre-existing social divisions of resident/owner/community insider versus migrant/renter/community outsider that we found in our own study.
In rural areas, Water User Sanitation Committees (WUSC) have long been seen as effective in improving health outcomes (Chandra Regmi 2005; Anderson et al. 2021). However, sanitation and other water system improvements show better durability and improved long-term health outcomes when women have 50% representation on these committees (Shrestha et al. 2017). Because women's primary role in the agriculture and small business sectors was largely ignored by NGOs, responses to the 2015 earthquake magnified inequalities in access to healthcare, clean water, and gendered violence (Sthapit 2015). Using data from the Nepal Demographic and Health Survey of 2016, analysts found that any worsening of household water access consistently elevates Nepali women's risk of intimate partner violence which in turn impacts women's ability to command resources and power (Choudhary et al. 2020, p. 581). Thus, in both rural and urban settings across Nepal, scholars and activists have noted the need for women's participation in designing and implementing water systems. However, community leaders remain mostly unpaid and this is a significant barrier to participation for those with limited resources and social capital (such as recent migrants). A middle-aged mother on the leadership committee of a local Federation of Drinking Water and Sanitation User (FEDWASUN) committee in the rural area of Chitwan told us that, while she herself receives a small stipend funded from membership fees, 70% of the women who participate in the water user associations are unpaid volunteers. She argues that since ‘women are in charge of household work’ they should lead water and sanitation efforts and for this reason, she is constantly recruiting more women to lead efforts to monitor the water system ‘from source to mouth.’ Her arguments are echoed by a recent study of water, sanitation, and hygiene (WaSH) programs across Nepal, which found that women are perceived as much more effective than men in implementing these programs (Anderson et al. 2021). These results suggest that improving water quality, access, and health outcomes in Kathmandu requires: (1) redirecting government resources to subsidize community participation in building, managing and maintaining water infrastructure and (2) supporting women's leadership in these community organizations.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge research funding from the Freeman Foundation Student-Faculty Fellows program of the ASIANetwork and the laboratory facilities and cooperation of Dr Basant Giri and students at the Kathmandu Institute of Applied Sciences.
Free and informed consent of the participants or their legal representatives was obtained and the study protocol was approved by the appropriate Committee for the Protection of Human Participants, the Institutional Review Board of Saint Mary's College, May 7, 2017.
DATA AVAILABILITY STATEMENT
Data cannot be made publicly available; readers should contact the corresponding author for details.
CONFLICT OF INTEREST
The authors declare there is no conflict.