Health and economic consequences due to inadequate water quality: a case study from Eastern India

The densely populated Eastern India (Government of India 2011) faces severe water quality issues due to industrial discharge, agricultural run-off, and inadequate sanitation infrastructure. These pollutants lead to various health problems, including waterborne diseases such as diarrhea, cholera, and dysentery, disproportionately affecting vulnerable and poor populations in the region (Sarkar 2024). Recent national level household surveys in India found that 95.9% of surveyed households use improved piped water⁷ for their domestic use and 70.2% of surveyed households have improved sanitation facilities⁸ (Government of India 2021). In this current study, we are focusing on Odisha and West Bengal, two states in the eastern part of India. We find that Odisha has the highest level of neonatal mortality of 32 per 1,000 live births (International Institute for Population Sciences (IIPS) & ICF 2021a) and 15.5 per 1,000 live births in West Bengal (International Institute for Population Sciences (IIPS) & ICF 2021b), highlighting the need for improvement in healthcare services and sanitation practices. Additionally, 45.2% households in urban areas and 66.1% in rural areas do not perform any kind of treatment on water before drinking (International Institute for Population Sciences (IIPS) & ICF 2021a). In West Bengal, the proportions are higher in comparison to Odisha. In total, 73.5% households in urban areas and 93.3% households in rural areas do not perform any kind of treatment on water before drinking (International Institute for Population Sciences (IIPS) & ICF 2021b). Contaminated piped water can expose households to waterborne diseases, imposing significant economic burdens through treatment costs, productivity loss, and reduced labor capacity. In Odisha, diarrheal diseases constituted 7.6% of the state's Disability-Adjusted Life Years (DALYs) in 2016, highlighting a major public health challenge. Malnutrition accounted for 12.7% of DALYs, whereas poor WASH practices contributed 7.4% of DALYs. Children are particularly vulnerable to diarrheal morbidity and mortality, underscoring the critical need for improved water, sanitation, and nutrition (Indian Council of Medical Research et al. 2017). These factors strain household finances and emphasize the importance of addressing WASH-related issues to alleviate public health and economic impacts (Indian Council of Medical Research et al. 2017).

In West Bengal, diarrheal diseases accounted for 3.0% of the state's DALYs in 2016, with malnutrition contributing 10.4% and inadequate WASH practices 3.1% of the total disease burden. Although these figures suggest improved public health measures, diarrhea remains a concern, particularly among children (Indian Council of Medical Research et al. 2017).

The European Union (EU)–India project SARASWATI 2.0 addresses waterborne health impacts, employing the cost of illness methodology to quantify economic burdens on households and society. Using a health production function approach, the study developed a utility-maximizing model to evaluate health damages faced by urban households due to water-related illnesses.

Health damage from the use of contaminated water is a major cause of illness and mortality around the world and is even significant in the case of children (Schwarzenbach et al. 2010; Lin et al. 2022). Major factors that have been found affecting the water pollution level significantly are income of households, population density, literacy level, fertilizer consumption, livestock population, level of industrialization, poverty, and the annual mean temperature of water (Barua & Hubacek 2009). Rather than going into what factors affect water pollution, we look into how polluted water affects the households in terms of economic burden due to illness caused by waterborne diseases.

Looking at the economic or social cost of waterborne diseases, the cost increases significantly with the affected individual's age (Kumar et al. 2022) and overall wealth or income bracket (Malik et al. 2012). A study in Pakistan found that in the case of a population below the poverty line, the average per day cost of illness ranges from US$ 0.6 to US$ 1.2, but in the case of middle- or high-income households, it increases to US$ 2.3 per day (Malik et al. 2012). However, the relative burden of diseases is higher for poor households than for middle- or higher-income households (Bedi et al., 2015). In the case of previous studies in India, we find that outbreaks of waterborne diseases have been seasonal, with high occurrence in the summer (Bedi et al., 2015). In a study carried out in Ludhiana City of Punjab, it was found that the annual per capita cost of illness for waterborne diseases ranges from INR 5,727 (for high-income households) to INR 3,385.90 (for low-income households) (Bedi et al. 2015). In the case of low-income households, the cost is less because of the financial hindrance to adopting any coping mechanisms adopted by high-income households, like the use of reverse osmosis water purifiers (Bedi et al. 2015). Another study conducted in Kolkata, West Bengal, found that the total average monthly cost of illness due to waterborne diseases is INR 97.34 (Majumdar & Gupta 2009), which annually amounts to INR 1,168.08. Still, the monthly cost can vary depending on the season. It has also been found that households are willing to pay more for improved water infrastructure to increase the quality of water received (Majumdar & Gupta 2009). The cost of using polluted or contaminated is not limited to the costs incurred due to illness but also due to productivity losses like in agriculture, the occurrence of diseases in livestock, and the loss of reproductive capacity (Reddy & Behera 2006). Waterborne diseases are caused by recreational water use, especially in coastal areas (Dwight et al. 2005). Dwight et al. (2005) found that recreation in marine waters is causing four major ailments among users in California – gastrointestinal illness, acute respiratory disease, ear ailment, and eye ailment. The cost of ailments is estimated by aggregating the lost income per illness episode and medical costs. The economic burden from gastrointestinal illness is estimated at US$ 36.58, from acute respiratory at US$ 76.76, from ear ailment at US$ 37.86, and from eye ailment at US$ 27.31 (Dwight et al. 2005). Health damages due to wastewater or polluted water can also be attributed to vector-borne diseases like malaria. Vector-borne diseases can have extensive and long-term economic effects at the micro- and macro-levels of the economy. McCarthy et al. (2000) found that malaria causes an annual average growth reduction of 0.55% in sub-Saharan nations.

Waterborne illnesses in developed nations, such as the USA, incur substantial economic costs, with surface water recreation alone causing 90 million illnesses annually, costing US$ 2.2–3.7 billion (DeFlorio-Barker et al. 2021). Domestic water supply led to 17 waterborne diseases between 2000 and 2015, resulting in US$ 3.33 billion in hospitalization costs (Collier et al. 2021). Outbreaks like Ghana's cholera epidemic imposed costs of US$ 60.68–100.46 per person depending on incidence rates (Awalime et al. 2017).

One of the methods to mitigate the economic burden due to waterborne diseases can be the establishment of decentralized wastewater treatment plants. Decentralized wastewater treatment plants mitigate such risks by removing contaminants, improving water quality, and reducing illnesses (Lienhoop et al. 2014). The health benefits can be monetized using the cost of illness method, encompassing medical expenses, lost wages, and other economic impacts of pollution-induced diseases (Hunter et al. 2010). These interventions also address social inequalities worsened by water pollution (Ridzuan 2021.

Based on the current scenario of water pollution and contamination, the Government of India released the Drinking Water Quality Monitoring & Surveillance Framework in October 2021. This framework aims to train citizens at the village level to conduct water quality tests and surveillance of water sources to reduce pollution and contamination of potable water for households (Ministry of Jal Shakti 2021).

The monetization of households' health losses provides a measure of the minimum willingness to accept the household's cleaner technologies to avoid damage from contaminated water. The findings of this study will contribute to a better understanding of the health and economic implications of inadequate water quality, thereby highlighting the importance of implementing water safety measures and prioritizing resources for designing effective interventions for water treatment and sanitation improvement.

The primary objective of this study is to investigate the various factors affecting the health and water purification behaviors of households in Bhubaneswar in the Khurda district (Odisha) and Kharagpur in tthe Paschim Medinipur district (West Bengal), with a focus on waterborne diseases. Specifically, the aims of the study are as follows:

• (1) To examine the factors contributing to waterborne illnesses.

• (2) To identify the determinants of water purification practices.

• (3) To estimate the cost of illness for a representative household.

Through this multi-faceted analysis, the study aims to provide a comprehensive understanding of the factors affecting waterborne diseases and water purification practices, and how these factors interplay with the socio-economic burden on households.

We have developed an extensive questionnaire to collect information from 281 households of the Khurda district in Odisha and the Paschim Medinipur district in West Bengal. The pilot plants are installed at the Indian Institute of Technology, Kharagpur and the Indian Institute of Technology, Bhubaneswar. The survey was conducted in two adjacent cities of the Paschim Medinipur district in West Bengal – Midnapore Sadar and Kharagpur, and peri-urban areas of Bhubaneswar in the Khurda district of Odisha. The households for the interviews were selected randomly from the two study locations. Figure 1 and Figure 2 provide the map of the study sites.

The survey sites in the Medinipur district needed proper sanitary latrine facilities and sewage disposal. The municipal water supply is the primary water source for most households, and also many have their own wells or groundwater boring facilities. The families suffer from various types of water-related problems ranging from faulty pipelines, leakages, foul smell, reduction of groundwater level due to extensive use of groundwater by multiple industries located in Kharagpur, reddish-brown color of water due to the presence of iron, irregular water supply, etc. Deterioration in the quality of water results in water-related illnesses faced by the households. A similar situation has been reported in the case of Bhubaneswar and the whole of Khurda district. In the case of Bhubaneswar, rapid urbanization has put pressure on the disposal of waste and sewage, with cases of untreated waste disposed of directly to local rivers, Kuakhai and Daya (Sahu et al. 2006). Urbanization has also led to encroachment in many areas of Bhubaneswar, reducing groundwater recharge and groundwater contamination (Mishra et al. 2021). Recently, cholera cases have been reported in the district of Khurda (Khuntia et al. 2021). Cholera has been categorized as endemic in the coastal districts of Khurda, Puri, and Cuttack due to regular reporting of cases (Khuntia et al. 2021). Against this backdrop, in the case of West Bengal and Odisha, the current study presents a case for the cost of illness incurred by an average household in our sample due to waterborne diseases and points toward policy implementation for reducing the cost of illness in the future.

This paper is based on limited geographical area with households only from the Khurda and Paschim Medinipur districts; therefore, it can limit the broader applicability in other areas. Furthermore, the data collected in the questionnaire have been self-reported by the respondents, and hence can have recall bias.

The questionnaire focused on the households' profile and health related to water sources, water use, and wastewater disposal. In the health aspects section of the questionnaire, we also collected information about the types of diseases that the households (or members of the families) have faced in the past 3 months (i.e., 3 months before the interview).

The first section of the questionnaire is the household profile and dwelling characteristics. This section gathers basic demographic information about the household and the respondent. It includes questions about the respondent's name, address, contact information, gender, age, education level, and occupation. Additionally, it collects data on the household background, size, income, and access to basic infrastructure like sanitary facilities, sewage connections, and municipal waste disposal services. The dwelling characteristics include whether the respondent owns or rents their home, the type of building they live in (individual house, apartment, or semi-attached quarter), and whether they face issues such as water scarcity or infrastructural problems like faulty pipelines or hard water. Furthermore, it investigates the household's main water supply sources and frequency of water conservation practices. This section aims to establish a socio-economic and infrastructural profile of the household, which is critical for analyzing the variations in perceptions and practices regarding water use and wastewater management.

The second section of the questionnaire is on the health aspects of the households. This section addresses the impact of water quality on household health. Respondents are asked if they believe water quality affects their health and whether they have experienced diseases linked to water pollution such as diarrhea, vomiting, or abdominal pain. The questions asked under these sections include the following (Table 1):

• Do you feel that the water supplied at your house (for potable and non-potable purposes) is not entirely free from disease-causing contaminants (presence of foul smell/unusual taste, etc.)? ______ (Code:1=Yes; 0=No).

• What water purification method do you follow in your household to ensure safe drinking water? __________ (Code:1=Water purifier/ RO system; 2=boiling and cooling (batch method); 3=chlorination; 4=already use purified (canned/bottled) water; 5=any other (please specify); 6=does not follow any method to purify water).

• Did you or any member of your household experience these diseases linked to water pollution in the past 3 months (recall) for which you required treatment? Please provide information in the table below. (Fill in.)

• Note: 1. Diarrhea is defined as the passage of loose, liquid, or watery stools more than 3 times a day.

The questions inquire about their water purification practices, such as whether they use water purifiers, boil water, or use bottled water, as well as the costs associated with these methods. The section also investigates hygiene practices, including whether respondents wash their hands after defecation and before meals. Additionally, the questionnaire asks how frequently storage facilities like water coolers or buckets are cleaned. This section's objective is to evaluate the health risks associated with water quality and hygiene behaviors, as well as the financial burden households bear for ensuring safe water. It helps to determine whether health concerns drive the willingness to adopt wastewater reuse technologies.

The third section is about awareness and the respondent's opinion about wastewater treatment and reuse. This section gauges respondents' awareness of wastewater disposal methods and sewage treatment plants (STPs). Questions explore familiarity with the concept of treated wastewater reuse and whether respondents support or oppose its use. Respondents are asked about acceptable uses of treated wastewater for activities such as irrigation, toilet flushing, or domestic cleaning. The section also explores reasons for supporting or opposing wastewater reuse, including water scarcity, environmental sustainability, or health concerns. It highlights public perception and knowledge gaps regarding wastewater reuse. This section is essential for understanding societal attitudes and identifying potential challenges in promoting the adoption of wastewater reuse technologies. It also helps assess how well-informed respondents are about the benefits of using treated wastewater for non-potable applications.

We have employed two methods for the analysis of the data to achieve our objective. The first method is the use of bivariate probit regression. The regression has been carried out on the R software. This regression methodology has been used to identify the major factors causing the disease incidence among the sampled households and the water purification techniques used by the households. This helps us address the first and second objectives of this study.

For the third objective of the study, we follow the cost of illness methodology from the study by Dasgupta (2004). We have collected the required information from the households in the questionnaire.

We start this section with the discussion of the descriptive statistics. Later, we discuss the results from the bivariate probit regression. From the regression, we obtain the statistically significant factors influencing the incidence of disease in the households and use of water purification methods by the households.

Of the total sample of 281 households, 247 households received municipal water, and 152 of these households expressed concerns about the water supply. Among them, 96 households reported specific problems with the supplied water, and 44 households mentioned experiencing diarrhea. The reported issues included foul smell, leaking pipes, and red-colored water caused by high iron content, and hardness of the water. Figure 5 represents how the information was filtered from the data collected from the questionnaire survey.

Table 2 shows the descriptive statistics of the independent variables that we have used for the bivariate probit regression. The first variable, water source, is binary, indicating whether households use a non-piped water source. With a mean value of 0.068, it suggests that only 6.8% of the households rely on non-piped sources, while the majority had access to piped water.

The presence of garbage indicates whether households experience irregular garbage disposal. A mean value of 0.58 shows that 58% of the households report garbage dumps not being regularly cleared. On the other hand, municipal waste disposal, another binary variable, reflects whether waste disposal is conducted regularly by the municipality. The high mean value of 0.932 indicates that 93.2% of households benefit from regular waste disposal services.

Table 2 also includes data on access to sanitary latrines, with a mean value of 0.979, indicating that 97.9% of households have access to proper sanitary facilities. Similarly, sewer connection, with a mean value of 0.961, suggests that nearly 96.1% of the surveyed households have a sewer connection for waste disposal.

The variable foul smell, indicating the presence of an unpleasant odor in the area, has a mean value of 0.241, showing that 24.1% of households experience such issues. Meanwhile, the income of households, reported in rupees, has a mean value of INR 45,017.79 (US$ 535.48) with considerable variability (standard deviation of INR 32,778.34 (US$ 389.90)), ranging from INR 25,000 (US$ 297.39) to INR 200,000 (US$ 2378.99).

Regarding storage facilities for water storage, the variable has a mean value of 0.964, 96.4% of the households have a separate storage facility for water. Cleaning frequency of storage facilities, measured on a scale of 1–5, has a mean value of 1.178, indicating relatively infrequent cleaning.

The variable water supply timing represents the number of hours water is supplied per day, with an average value of 1.37, although it ranges widely from 0 to 7 h. Finally, the education level of the respondent, measured on a scale from 1 to 6, has a mean value of 4.036, indicating a moderately educated population with respondents' education levels ranging from primary school to higher education.

Table 3 presents the results of a bivariate probit model, examining the relationship between illness as the dependent variable and various household socio-economic and infrastructural factors as independent variables. In a bivariate probit model, the coefficients represent the marginal effects of the independent variables on the probability of the dependent outcome, in this case, illness. Negative coefficients imply a reduction in the probability of illness, while positive coefficients suggest an increased likelihood.

Piped water supply has a negative coefficient (−0.677), suggesting that households with piped water have a lower probability of reporting illness, though this effect is not statistically significant. Similarly, the availability of municipal waste services and more regular timing of water supply also show negative coefficients (−0.446 and −0.336, respectively), indicating a lower probability of illness, though these effects are not significant.

Income, measured as the log of per capita income, shows a slight negative association (−0.283) with illness probability, but even this is not significant. Notably, the presence of a garbage dump near the household significantly increases the probability of illness (coefficient: −0.383, p < 0.05), while higher education levels of the household head reduce illness likelihood. Specifically, households led by head with a graduate or advanced academic qualifications head exhibit a strong negative association (−0.878, p < 0.05), highlighting the protective effect of education.

Foul smells from water pipelines show a substantial positive association with illness (0.871, p < 0.01), significantly increasing the probability of illness. Other factors, such as piped sewage collection (0.495) and location dummy (0.278), were not found to have significant effects.

Overall, the results indicate that household education, infrastructure quality, and environmental conditions play crucial roles in influencing illness probabilities in the population studied.

Equation (2) has the dependent variable, water purification, employed by the households at their place. The dependent variable refers to the household and it takes the value 1, otherwise 0 if the households have used any water purification technique. The water purification techniques that the households use include usage of water purifiers, boiling and cooling techniques, chlorination, or using already purified water in the form of bottled or canned water. The per capita monthly income significantly and positively affects households' use of water purification techniques. Similarly, the education of the household head is significant and positive regarding household water purification. Table 4 presents the results of a bivariate probit model, where the dependent variable is the use of water purification techniques, and the independent variables include socio-economic factors, household infrastructure, and environmental conditions.

Income, as measured by the log of per capita income, has a positive and significant coefficient (0.572, p < 0.01), indicating that higher-income households are significantly more likely to use water purification techniques. Households headed by individuals with higher education, especially those with a graduate or above education, also show a strong positive association with water purification adoption (coefficient: 1.12, p < 0.01). This suggests that education, particularly at higher levels, plays a crucial role in influencing households to engage in water purification practices.

The availability of piped sewage collection exhibits a large positive and significant effect (coefficient: 1.579, p < 0.01), indicating that households with access to piped sewage are more likely to use water purification techniques. This may suggest a correlation between improved sanitation and the likelihood of households taking additional steps to ensure water quality.

Other variables, such as the presence of a piped water connection (0.033), school-level education of the household head (0.458), availability of sanitary latrines (0.03), and location dummy (0.243), do not show significant associations with the likelihood of adopting water purification techniques. The negative coefficient for foul smells from water pipelines (−0.011) also lacks statistical significance, implying that this environmental factor does not significantly affect the adoption of water purification methods.

The constant (−7.536, p < 0.01) is significant and large, reflecting the baseline likelihood of water purification in the absence of other factors. Overall, higher income, advanced education, and piped sewage collection are key drivers of household water purification behavior.

Table 5 shows the treatment cost of illness on an average for the households. Here, we follow the technique used by Dasgupta (2004). We use the model estimated in the above section to calculate the cost of illness on an average for the households. The univariate (marginal) predicted the probability of success in the outcome, defined as the probability of observing illness in a household in our sample. The average value obtained for this predicted probability is 0.541. The probability is used along with other statistical measures to determine the monetized treatment cost of the illness for the sample.

Table 5

Cost of illness estimates

Variable name .	.	Value .
The probability of a household being affected	λ	0.541
Probability of being a child from an affected household
Probability of being ill if the individual concerned is a child from an affected household
The probability of the child being ill		0.004352
Probability of being an adult from an affected household
Probability of being ill if the individual concerned is an adult from an affected household
The probability of the adult being ill		0.003876
Probability of being an elder from an affected household
Probability of being ill if the individual concerned is a child from an affected household
The probability of the child being ill		0.000765
Average cost of treatment for a child	(in INR)	368.75
The average cost of treatment for adult	(in INR)	546.36
The average cost of treatment for the elderly	(in INR)	275
Average family size		4.017
Cost for a representative household (in INR)
		47.69

Variable name .	.	Value .
The probability of a household being affected	λ	0.541
Probability of being a child from an affected household
Probability of being ill if the individual concerned is a child from an affected household
The probability of the child being ill		0.004352
Probability of being an adult from an affected household
Probability of being ill if the individual concerned is an adult from an affected household
The probability of the adult being ill		0.003876
Probability of being an elder from an affected household
Probability of being ill if the individual concerned is a child from an affected household
The probability of the child being ill		0.000765
Average cost of treatment for a child	(in INR)	368.75
The average cost of treatment for adult	(in INR)	546.36
The average cost of treatment for the elderly	(in INR)	275
Average family size		4.017
Cost for a representative household (in INR)
		47.69

View Large

Table 5 presents the probabilistic and cost analysis of the impact of illness on the members within the households. Here, we have explored the likelihood of a household being affected by illness, the probabilities of individuals within those households – children, adults, and elders – contracting the illness, and the associated treatment costs for each group. In Table 5, we have shown the breakdown of the variables and their interrelationships, leading to an estimation of the total cost burden on a representative household.

We start by defining the probability of a household being affected by the illness, denoted by λ, with a value of 0.541. This means that there is a 54.1% chance that any given household will be affected by the illness. Then, we calculate the probability of being a child, adult, or an elder within an affected household, denoted as μ_c, μ_a, and μ_e, respectively. The probabilities are calculated as the ratios of the number of children (165), adults (864), and elders (100) to the total number of individuals in the study (1,129). The probabilities are μ_c = 0.14 for children, μ_a = 0.76 for adults, and μ_e = 0.09 for elders.

Next, we examine the probability of falling ill for each age group within affected households. For children (ϕ_c), this probability is 0.16, meaning there is a 16% chance that a child from an affected household will become ill. For adults (ϕ_a), the probability is lower at 0.03, while for elders (ϕ_e), the probability is 0.05. These probabilities are calculated as the ratios of the number of ill individuals in each age group to the total number of individuals in that group within the affected households. These probabilities are calculated as the ratios of the number of ill individuals in each age group to the total number of individuals in that group within the sample of households.

The overall probability of illness for each group within the population is calculated by multiplying the probability of a household being affected (λ) by the probability of being in a specific age group (μ) and the probability of being ill within that group (ϕ). The results show that the probability of a child being ill is 0.004352, the probability of an adult being ill is 0.003876, and the probability of an elder being ill is 0.000765.

The study examines the prevalence of waterborne diseases in households in Odisha and West Bengal and suggests mitigating measures through the implementation of decentralized wastewater treatment plants.

The results indicate that improvement in water supply and treatment can go a long way in ensuring the good health of the households and reduce the economic burden of waterborne disease by reducing the cost of illness as well as workdays lost due to illness. Further benefits also include better-quality water being available for personal hygiene and food preparation that associate with other benefits to increase the overall well-being of the households.

The results from the bivariate probit models highlight the critical role of socio-economic factors, education, and infrastructure in determining household health outcomes and water purification practices. Higher education, particularly at the graduate level, emerges as a significant protective factor, reducing the likelihood of illness and increasing the probability of adopting water purification methods. Income also plays a crucial role, with wealthier households being more likely to purify their water.

Although the presence of piped water supply and municipal services show some influence, their effects are less significant compared to other factors like piped sewage collection, which strongly drives water purification practices. Additionally, environmental factors, such as foul smells from water pipelines, significantly increase the risk of illness but do not appear to influence water purification behaviors. These findings suggest that improving education and income levels, along with enhancing sewage infrastructure, could be key strategies for reducing illness and promoting safer water practices in households. The results are similar to findings from other parts of West Bengal where supply water users have reported higher incidence of waterborne skin infection (Sarkar 2024).

Further benefits in terms of household well-being can be derived from increasing the availability of water to households by increasing piped water supply. This can reduce the burden on households for securing a regular water supply. Clean water to households via pipe connection can ensure minimal pollutants and contaminants in the water. Additionally, the intermittent water supply from the municipality can cause microbial build in the supply lines causing diseases among households (Kumpel & Nelson 2016). Also, if household members can spend less time in securing water, they will find more time for caring and nursing ill household members. This can be especially helpful in reducing the severity of disease among children (Pickering & Davis 2012).

The study offers two main conclusions to address waterborne diseases at both household and administration levels. At the household level, raising awareness and providing education, particularly to household heads, significantly reduces the incidence of waterborne diseases and promotes the adoption of water purification strategies (Jena et al. 2024; Awiyyah & Sufiyan 2017). Also, the need for generating awareness about water decontamination and hygiene has been highlighted from other parts of India such as Chandigarh (Ravindra et al. 2019) and Cooch Behar in West Bengal (Sarkar 2024). Household income also plays a crucial role in determining the use of water purification methods. Sarkar (2024) found that as the income level of households increases the better water decontamination techniques are used.

At the administration level, interventions aimed at promoting awareness of proper water purification techniques can improve the water quality at the household level. Initiatives such as the Jal Jeevan Mission, launched in 2019, focus on raising awareness in communities to encourage actions for water conservation and reuse (Ministry of Jal Shakti 2019). This can be particularly beneficial for households with low income and low education levels, helping them access the best water purification and cleaning techniques. Authorities can implement these initiatives in vulnerable areas to raise awareness about water purification practices. Furthermore, community involvement is necessary to promote water cleaning practices.

Policy interventions are also necessary to support the operation and maintenance of decentralized wastewater treatment plants at the two sites, as well as to enhance the quality of piped water supplied to households to minimize foul smells and water pollution. Improving sanitation has been linked to a 2.2% reduction in the risk of diarrhea among children in India (Kumar & Vollmer 2013).

Furthermore, there can be comparison between cost of illness due to waterborne diseases and the cost of infrastructure for piped water supply. This can provide crucial input in policy-making at the local level. Supplying the adequate quantity of water through the pipe system can improve the living standard of the households and confounding it with benefits like better-quality water and time saving from water collection can reduce the burden on members of the households. The households surveyed in this study are in the urban and peri-urban areas of Kharagpur in the Paschim Medinipur district (in West Bengal) and Bhubaneswar in the Khurda district (in Odisha). Any improvement in the water supply can open up newer economic opportunities for the households as well as improve the leisure time of the households. This will also reduce the economic burden on the households and help in reduction of inequality in the communities.

The study also emphasizes the importance of infrastructure related to waste disposal, water sources, and water purification. While these variables may or may not be under the households' control, government initiatives such as the National Mission for Clean Ganga and river rejuvenation programs in various states signify progress in this direction. However, collaboration between the community, local government, and planners is essential to enhance infrastructure, manage waste disposal and water supply, and ensure the long-term sustainability of decentralized wastewater treatment plants.

In conclusion, this study reveals the significant economic burden caused by waterborne diseases on the households of the Khurda district in Odisha and Paschim Medinipur district in West Bengal, with an average cost of illness per household estimated at INR 47.69. Households relying on piped water sources reported higher illness incidence than those using non-piped sources. Lower-income households (Q1 and Q2) experienced 24 and 16% of illness cases, respectively, compared to wealthier households.

The results from the bivariate probit models emphasize the critical role of socio-economic and environmental factors. Education significantly reduces illness probability, with households headed by graduates showing a 0.878 (p < 0.05) reduction in illness likelihood. Foul smells from water pipelines are strongly associated with increased illness probability (0.871, p < 0.01). Higher income and education also significantly drive water purification practices, with a 0.572 (p < 0.01) increase in likelihood for wealthier households and 1.12 (p < 0.01) for graduate-headed households. These results stress the importance of improving education, sanitation infrastructure, and water supply to reduce both the economic burden and health impacts caused by waterborne diseases.

This research was conducted within the project SARASWATI 2.0, ‘Identifying Best Available Technologies for Decentralized Wastewater Treatment and Resource Recovery for India’. This project was jointly funded within the framework of the EU–India Water Cooperation by the European Union (Horizon 2020 Research and Innovation Program, grant agreement no. 821427) and by the Government of India (Department of Science and Technology/Department of Biotechnology, sanction order DST/IMRCD/India-EU/Water Call2/SARASWATI 2.0/2018(G)).

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.