Household drinking water treatment in rural Indonesia: actual practice, determinants, and drinking water quality

Household water treatment (HWT) is various techniques used to treat drinking water, such as boiling or filtration, and a low-cost option to reduce the consumption of contaminated water at the household level (Sobsey et al. 2008). Improved water quality from appropriate treatment measures can decrease the number of water-related diseases (WHO 2019). HWT can be seen as an interim solution for an area where people still do not have access to safe water sources (Ren et al. 2013). However, evidence from previous studies described challenges in the adoption of this HWT practice, noting irregular use (Brown & Sobsey 2012; Rosa et al. 2014; Geremew et al. 2018, 2019). Poor adoption can undermine the potential benefits these WASH interventions can provide (Enger et al. 2013; Geremew & Damtew 2020).

HWT is commonly practiced in Indonesia. About 70% of Indonesia's population carries out HWT regularly and boiling is the most common method that people use (National Population and Family Planning Board et al. 2018). However, previous studies reveal that households in rural and undeveloped areas of Indonesia, still drink untreated water (Daniel et al. 2020a, 2021). Furthermore, previous studies related to HWT in Indonesia were conducted mainly in areas where many people still drink untreated water or do not perform HWT regularly (e.g., Daniel et al. 2020a, 2021). Those studies found some critical psychological factors that distinguish those who regularly and not regularly treat drinking water, e.g., coping planning (self-regulation) and attitude related to the taste of treated water.

In contrast to a popular view, we were curious to see the ‘level’ of behavioral determinants of people in areas where the majority of households treat their drinking water. We may expect that the level of psychological factors is optimal, i.e., at a higher level, in this kind of area. For example, communities with high HWT may perceive higher levels of health risks if they drink untreated water or think that treating water is a social norm everyone has to do as well.

This study then aims to assess the psychological factors related to HWT practice in the community, where most of the households already treat their drinking water regularly. Therefore, this paper can be seen as an explorative study to answer the question, ‘Why do most of the people there treat their drinking water?’.

A total of 425 households in three villages were visited in the first phase: Randoria, Tanali, and Fataatu Timur. These household numbers cover almost 90% of the total households in those three villages. We originally targeted all households in those villages. However, unavailable households, i.e., no one at home, were skipped. We mainly targeted the primary caregiver in the house, i.e., the mother, as the respondent. However, if the mother was not available, the household head or the oldest person in the house was interviewed. The participants were interviewed about their HWT practices, socio-economic conditions, HWT-related psychological factors based on the RANAS psychological framework (Mosler 2012), WASH promotion, sanitation, and observation about general hygiene conditions (see Appendix, Supplementary material). All participants gave written informed consent before the interview. The interview and observation were conducted by seven local experienced enumerators.

For the second phase, drinking water quality samples from 143 houses from the previous phase were randomly collected. We also asked whether the drinking water was treated or not.

Psychological theories or frameworks have been used to understand those behavioral determinants, one of them is the RANAS (risk, attitude, norms, ability, and self-regulation) framework. RANAS has been used widely in the WASH domain, including HWT (e.g., Lilje & Mosler 2017; Daniel et al. 2020a; Rahaman et al. 2022). The psychological-related questions were created following the RANAS psychological framework (Mosler 2012). RANAS consists of five main psychological factors: risk, attitude, norms, ability, and self-regulation. Furthermore, there are some sub-factors under those five main factors (Table 1). For example, there are three sub-factors under the norms factor: descriptive norm (others’ behavior), injunctive norm (others’ approval), and personal norm (personal importance). Most of the sub-factor questions are measured on a five-point Likert scale. Detailed information about RANAS can be found in Mosler & Contzen (2016).

There are three points from the questionnaire that can be used to assess the practice of HWT: (1) whether or not they use HWT (measured on four scales: ‘never’, ‘stop’, ‘intermittent’, and ‘always’); (2) frequency of drinking untreated water (five scales from ‘always’ to ‘never’); and (3) daily percentage of drinking water which is treated (five scales from ‘less than 20%’ to ‘more than 80%’). All these information are self-reported practice. The intention to use multiple information was to assess the consistency of the answer. For example, we may expect that regular HWT users will answer ‘always’ to the first question, ‘never’ to the second question, and ‘more than 80%’ to the last question.

We also assessed the safety of the drinking water storage by assessing three main criteria: (1) conditions of the storage, i.e., whether there is a crack or not, the cleanliness of the inside container, or the storage cover; (2) the placement of the storage, i.e., whether the storage is placed above the ground or not, i.e., above a table or chair; and (3) the surrounding environment, i.e., presence of flies and animals around, the cleanliness of the floor, and the presence of animal feces and garbage. These factors may introduce recontamination into the treated water (Daniel et al. 2020b).

Participants were requested to fill a sterile sampling bag (Whirl-Pak bag, Nasco, Fort Atkinson, WI, USA) with approximately 200 mL from their drinking water container. The samples were stored in a cooler box and transported to the District Health Laboratory for analysis within 6 h after collection. The samples were processed using the most probable number (MPN) method to assess the level of fecal contamination.

The ordinal regression was performed to identify significant psychological factors associated with the frequency of drinking untreated water, i.e., measured on five scales from ‘always’ to ‘never’. We assume that if the respondent says ‘never’ to drink untreated water, it means that they regularly perform HWT and drink treated water wherever they are, i.e., inside or outside the home. All RANAS sub-factors were included in the analysis as independent variables. The Chi-squared test was conducted to assess the relationship between the risk of fecal contamination and HWT.

Most of the respondents were mothers (71.8%), followed by fathers (22.4%). Other respondents were the oldest or older child, grandparents, or relatives who live in the house. The mean respondent's age was 43 years old (SD = 13.7; range: 18–86). The mean number of years of respondent's education was 7.8 years (SD = 3.3) or equal to junior high school. Furthermore, the mean year of a household head's education was 7.6 years (SD = 3.4). There were 133 households (31.5%) that had, at least, one child below the age of five; a total of 167 children. Among these 133 households, diarrhea in children in the last 2 weeks occurred only in 16 houses (12%). Considering the economic status of the respondents, 72.2% of the respondents had a permanent floor, but only 18.6% of them had permanent walls. The mean monthly expenditure of the households was IDR 634,517 or $44.3. Most of the household heads worked as farmers (91%).

The main drinking water source was untreated tap (39.3%), i.e., private tap was 27.5% and public tap was 11.8%, followed by unprotected spring (21.4%), protected spring (16.7%), and river (11.3%). The remaining respondents used well water, rainwater, and commercial potable water. The existing pipe distribution system did not reach all households in those three villages, due to insufficient budget when the system was being built, so others relied on self-supply water sources. However, 29.6% of the respondents used more than one water source. Among 167 respondents who used tap water, 32.3% of them do not have 24 h available taps. About 63.3% of the respondents said that the water source is located on-premises or within less than 5 min of walking (go and back). However, there were still 11.5% of the respondents need to walk more than 45 min of walking to draw water.

There were only 0.5% of the respondents still practice open defecation, while 84.2% had their own toilet and 15.3% used a shared toilet. Among those who had a private toilet (n = 358), 53.1% had full water in the bathtub and 36.6% had about half-full of water. Furthermore, about 39.7% were considered to have clean toilet conditions.

About four-fifths of the respondents (81.8%) have ever received any WASH promotion in the past. Health workers and NGO officers were the main promoters, i.e., mentioned by 70.4 and 48.3% of the respondents, respectively, followed by the community leader, i.e., 27.4%. Furthermore, when we asked them who they prefer to be a WASH promoter, a health worker, and an NGO officer were mentioned, i.e., 65.1 and 16.5%, respectively.

About 37.9% of the respondents know more than one HWT methods, however, all respondents said that boiling is the most frequent HWT method that they perform. Almost all of the respondents, i.e., 99.5%, said that they always do HWT every day. On the other hand, 85.9% said that they never drink untreated water in a day and 83.1% said that almost all (81–100%) of their drinking water was treated. These three pieces of information show that HWT is regularly performed in this area, even though not all the water they drink has been treated.

Furthermore, we observed the storage conditions. As much as 92.2% of the storage containers were found with no cracks, 97.6% of the inside containers were considered clean and 98.4% had a cover (n = 425). In terms of storage placement, 16.9% were put directly above the ground or floor. Moreover, in terms of the surrounding environment, there were flies, animals, animal dirt, and garbage in 28.7, 40.2, 23.3, and 27% of the observed households, respectively. About 76.2% of the house floor was also considered dirty. Furthermore, 64.2% of the respondents claimed that they clean the storage more than once a week. The observation results imply that, in general, the respondents stored their drinking water in safe storage, but the storage was placed in an unsafe and unhygienic environment.

The levels of RANAS psychological factors are shown in Table 1. In general, the mean scores of all sub-factors were quite high, i.e., the upper 20% of the possible scores. These high scores indicate that the respondents have a ‘favorable’ mindset, which then facilitates the regular practice of HWT. If we normalized the scores of action planning and coping planning to five scales and then calculated the mean of sub-factors for each of the five main RANAS factors, the mean of sub-factors in norm and ability were the highest compared to others, i.e., mean = 4.30. On the other hand, the mean of sub-factors under risk was the lowest, i.e., 3.35. The mean scores of the sub-factors under attitude and self-regulation were 4.28 and 4.08, respectively. These findings may denote that the respondents were not fully aware of the health aspect of HWT practice, i.e., this is not the main driver of HWT practice, instead of other psychological factors. Additionally, the top three highest mean scores were: commitment (self-regulation), descriptive (norm), and confidence in performance (ability).

Table 2 shows the results of the ordinal regression analysis. There are five variables significantly associated with the frequency of drinking untreated water: belief about taste (attitude), descriptive (norm), injunctive (norm), action planning (self-regulation), and remembering (self-regulation). Positive coefficients indicate that the chance of drink treated water increases if households have those perceptions in mind. The action planning (self-regulation) was the most influential psychological sub-factor (the highest β value in Table 2), followed by descriptive (norm) and belief about taste (attitude).

Almost all respondents (99.5%) claimed that they always treat their drinking water. Slightly lower percentages are found in questions on the proportion of households who never drink untreated water (85.9%) and the amount of water being treated (83.1%) in the households. The second household visit even found that only 76.7% of the sampled drinking water was treated. Based on our further discussion with some locals, they always treat their drinking water, but they may not drink treated water when they spend some time on the farm, which is often far from home and there is no treated water there, or if the treated water runs out. The mean score of the psychological factor coping planning indicates that about 30% of the respondents have no clear idea how to always treat water if any barrier arises, e.g., when there is no resource to treat water outside the home, which confirms those discussions with locals. We then can conclude that even though HWT is commonly practiced in this area, people do not fully and regularly drink treated water.

The previous study indicated that a decline in adherence of drink treated water from 100 to 90% reduces the health benefits of HWT by more than 95% or can be considered that they actually drink untreated water (Brown & Clasen 2012). Brown & Clasen (2012)'s study also found that the health benefits decrease sharply if the water source or untreated water is heavily contaminated. Based on our discussions with the public health center (‘Puskesmas’ in Bahasa’; a government clinic at the sub-district level) and district health agency, the fecal contamination in the water source in that area is relatively high, even though there is no visible source of contamination nearby, e.g., the presence of animal or open defecation, as stated also by another documentation (Marifa et al. 2012). Therefore, we argue that people in the study areas may receive only small health benefits from drinking treated water in their homes, especially because they do not drink 100% treated water, i.e., approximately 15% of the respondents.

High mean scores in almost all RANAS psychological sub-factors indicate that commonly HWT practiced in this area because people have an adequate or positive mindset related to HWT, including a mindset related to the intention to practice HWT and also the repetition of the HWT practice. The psychological sub-factor commitment (self-regulation) had the highest mean score, followed by descriptive (norm) and confidence in performance (ability), while in general, sub-factors under risk had the lowest mean score. Commitment and confidence in performance contribute significantly to the behavior, i.e., repetition of the HWT practice and descriptive norm create a social pressure, in which all these three sub-factors facilitate regular action, i.e., the behavior. On the other hand, sub-factors under risk contributed mainly to the intention to treat water (Mosler 2012), i.e., respondents intend to treat water if they feel a health risk from untreated water. These findings imply that people practice HWT automatically and regularly, even without a complete understanding of the health aspects of treated drinking water. However, this practice could backfire because people may easily change their behavior due to a poor understanding of the health aspect of WASH behavior.

Action planning (self-regulation) is the most influential sub-factor in the regression analysis that distinguish the frequency of drinking treated water among respondents. People without a clear plan about when, where, and how to perform the behavior have a lower chance of regularly performing HWT because repetition of the behavior is supported mainly by planning (Tobias 2009). The typical intervention for this is by prompting the target group to make specific plans to perform the behavior (Mosler & Contzen 2016), e.g., in our case, when people are outside the home.

Descriptive norm (norm) and belief about taste (attitude) are the next influential psychological factors in the regression analysis. A previous meta-analysis of HWT studies in developing countries concludes that there is a reverse causality between norm and attitude and HWT behavior (Daniel et al. 2022). The increase in levels of norm and attitude will increase the chance of regularly drinking treated water and this will increase back the levels of norm and attitude. This study reemphasizes the importance of social pressure and personal feeling or experience regarding HWT practice.

Discussions with stakeholders reveal that intensive WASH promotion in the whole district in the 2010s was the reason for a high level of HWT practice in this area. Previously, many people still practiced open defecation, drank untreated water, and had a high occurrence of diarrhea in children. It was found that the diarrhea epidemic regularly occurred in the coastal areas where people often defecated on the beach (Marifa et al. 2012). All these events triggered changes in WASH-related behavior, including household water treatment and safe storage (HWTS). However, based on our findings, promotion is still needed to make sure people are drinking 100% treated water, which is also accompanied by safe storage.

The observations of the drinking water storage and its surroundings reveal that the drinking water is stored in safe storage. However, the surrounding area is considered unhygienic, which can introduce recontamination into the treated water or nullify the effect of HWT (Trevett et al. 2004; Psutka et al. 2011; Sodha et al. 2011). The drinking water quality analysis reveals that HWT seems to improve the water quality (Figure 2), even though the association is not statistically significant, as shown in other studies (Daniel et al. 2020c; Imtiyaz et al. 2021; Bänziger et al. 2022). Based on all of these results, we argue that the currently practiced HWT and storage in this area are not fully appropriate to protect them from unsafe drinking water. Thus, people still need to be informed about the correct practice of HWT and HWTS. However, there is also a possibility of bias in the self-reported HWT in the drinking water sampling that results in an insignificant association, i.e., respondents report that they do HWT, but they do not.

There are some limitations of this study and some recommendations for future study. First, we did not check the whole process of household water management that may influence the water quality, e.g., collection from the water source and transport to the home. All these processes should be checked to assess the risk of contamination in the drinking water according to the WHO Sanitary Inspection Forms (WHO 2020), even though Daniel et al. (2020d) argued that steps after HWT are more critical and should be checked than steps before HWT. Future studies should also analyze the quality of water sources or points of collection (POC) so we have a complete understanding of the water quality in this area and also possibly link the water quality at POCs and homes. Lastly, the study was conducted in the rainy season, when there is plenty of water. It would be interesting to study whether people still treat drinking water regularly in the dry season, even though our in-depth interviews reveal that the water availability does not decrease much in the dry season.

This study found that the household surveys found a high level of psychological factors RANAS, which then assist in the regular use of HWT in the district of Ende, NTT Province, where most people already treat their drinking water. The sub-psychological factors that have the highest level of importance are commitment (self-regulation), descriptive (norm), and confidence in performance (ability). All these sub-factors confirm and indicate that HWT is commonly practiced in this area. However, our results imply that untreated water may be consumed when the adults are outside the home. This action may nullify the health benefits of HWT. The action planning (self-regulation) was the most influential psychological sub-factor, followed by descriptive (norm) and belief about taste (attitude), related to the frequency of drinking treated water. Furthermore, intensive WASH promotion activities in the past in the whole district seem to have significantly changed the poor WASH behavior, including HWT. Treated drinking water has better water quality than untreated one, but not statistically significant. Observations reveal that people store their drinking water safely, but the storage is placed in an unhygienic environment, which may introduce recontamination. Therefore, behavioral interventions are still needed, even in communities that commonly practice HWT, so people can perform regular and correct HWT and safe storage and get the maximum health benefit of HWT.

We thank the Ministry of Education, Culture, Research, and Technology of Indonesia (Kementerian Pendidikan, Kebudayaan, Riset, dan Teknologi in Indonesian) for all support in this study. We are also grateful to all enumerators and Wahana Visi Indonesia Area Program Ende for all support in the field data collection, especially Abner Radani Sembong, Oktaviana Sulu, and Krysogonus Tau Tonda.

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

The authors declare there is no conflict.