Critical review of household water treatment in Southeast Asian countries

Worldwide, an average gap of 32% was observed between urban and rural populations in terms of access to safe drinking water. Worryingly, 50% of the Southeast Asia population resides in rural areas, making the region highly vulnerable to threats from not having access to safe drinking water. The sparse population density and infrastructural complexities in rural areas have made centralized water treatment systems very challenging in terms of implementation and significantly increased cost. Hence, adopting a household water treatment (HWT) system would be a more suitable co‐existing water provision solution. However, data on the sustainability of HWT in Southeast Asia is still lacking. Therefore, this review aims to provide a critical overview of water poverty and current HWT implemented in Southeast Asian countries. The factors associated with feasibility and potentially sustained implementation of the HWT in Southeast Asian countries covering user preferences, user perception towards water safety, education and training, economic feasibility, collaborations, and supportive policy environment were also discussed. In a nutshell, there is a need for co‐designing the HWT with the targeted community before its implementation for better sustainability.

is explicitly reflected in Sustainable Development Goal (SDG) target 6.1, which states that "By 2030, achieve universal and equitable access to safe and affordable drinking water for all." To track progress towards the target, indicator 6.1.1 monitors the proportion of the population using safely managed drinking water services. There are currently 2 billion people living without access to safe water (UN, 2021). By 2030, current estimates predicted that nearly 700 million people will be displaced due to insufficient access to water (UN, 2020). Between 2000 and 2020, the global population using safely managed services has increased from 61.7% to 74.3%. Central and Southern Asia, together with Sub-Saharan Africa are lagging behind the global average with proportions of 62.4% and 30.0% (Ritchie & Roser, 2021). The gap in access to clean water between high-income countries and low-income countries is 68.8%, reduced by only 10% over 20 years (Ritchie & Roser, 2021). Even with all the mitigation efforts, 129 of the 193 UN member states worldwide are not on track to manage water resources sustainably by 2030 (UN, 2021). Significant inequalities in the access to safely managed water services are also seen between urban and rural areas, recording a gap of 32% (UNICEF & WHO, 2019).
A centralized water distribution system is feasible in densely populated urban areas due to existing infrastructure and benefits from economies of scale (Peter-Varbanets et al., 2009). However, these facilities are lacking in sparsely populated rural areas. The possible reasons are (i) rural localities situated further from centralized drinking water treatment plants, (ii) huge investment required for centralized systems in remote locations, and (iii) lack of financial resources and funding support (Boguniewicz-Zablocka & Capodaglio, 2017;Estévez et al., 2022;Huang et al., 2021). Some countries, including Mongolia, Bangladesh, and Vietnam face a funding gap of at least 61% to achieve SDG 6 targets (UN, 2019(UN, , 2020. Localities lacking centralized water treatment and distribution networks could benefit from adopting decentralized lower cost and safe water treatment systems such as household water treatment (HWT) (UNESCO World Water Assessment Programme, 2017). UNESCO World Water Assessment Programme (2017) reported that the investment, operation, and maintenance costs of decentralized systems are 20%-50% lower than large-scale systems. Unlike the centralized system that is dependent on the availability of large water resources, decentralized systems are feasible with limited water resources.
HWT, also commonly known as point-of-use water treatment, often provide a more practical and feasible co-existing water provision solution to obtain safe drinking water at home in localities facing intermittent water supply or remote areas, in which case they might need to depend directly on surface or unimproved water sources (Peter-Varbanets et al., 2009;WHO, 2017). Drinking water from unprotected or untreated sources could transmit waterborne diseases like diarrhea, cholera, dysentery, typhoid, and polio (WHO, 2019). According to Murray et al. (2020), unsafe water sources were responsible for 1.2 million premature deaths globally, which accounted for 14% of death in Chad,8.17% in South Sudan, 6.31% in India, 3.21% in Indonesia, 2.06% in Laos, and less than 0.01% across most European countries (Murray et al., 2020). The disaggregated data according to countries showed that lower-income countries, especially across Sub-Saharan Africa and Asia recorded higher unsafe water death rates (Ritchie & Roser, 2021). The disease burden linked to poor hygiene, sanitation, and clean water supply is largely preventable (Bartram & Cairncross, 2010;Cairncross et al., 2010). HWT is one of the proven cost-effective interventions that has been adopted in many low-and middle-income countries worldwide (Rosa et al., 2016). However, literature reflected conflicting reviews and evidence on the acceptability and scalability of HWT (Clasen et al., 2009;DuBois et al., 2010;Luoto et al., 2011Luoto et al., , 2012Pickering et al., 2019;Rosa et al., 2016;Schmidt & Cairncross, 2009). Out of three randomized efficacy trials in Kenya, Bangladesh, and Zimbabwe, WASH Benefits Bangladesh reported reduction in child diarrhea with the immense support of frequent contact of the promoter with the participants (Pickering et al., 2019). While this level of commitment from the promoters for WASH might not be scalable, however, most literature supported the need for more evidence on the use of HWT especially in developing countries in contributing to the realization of SDG 6 in a sustainable way (Deng, 2021;Luoto et al., 2012;Nielsen et al., 2022;Pickering et al., 2019;Rosa et al., 2016).
Based on the UN (2021), spotlighting the Asia-Pacific region, only two-thirds of the SDG 6 indicators had adequate data for progress tracking. More worryingly, 12 out of 46 least developed countries worldwide are from this region, and thus, under-reporting of safe water management might fail to spot vulnerable populations that need urgent attention. To get a better understanding of water management at the household scale, Southeast Asia (SEA) is selected as the study area for the following reasons: (i) one-third of least developed countries from Asia-Pacific belong to the SEA region, (ii) 50% of SEA population resides in rural areas, which makes the region highly vulnerable to threats from not having access to safe drinking water, and (iii) under-reporting observed, as only a small fraction of SEA countries has publicly available drinking water quality and service delivery quality reports (urban: 13%, rural: almost 0%; Maniam et al., 2021;UNICEF & WHO, 2019;WHO, 2019). Thus, this review aims to provide a critical overview of water poverty in SEA and the factors associated with feasible and sustained implementation of HWT in SEA countries.

| METHODS
Data for water poverty, access to different water sources, and various aspects of HWT focusing on SEA were drawn from a vast literature search of both the gray and academic literature from 2000 to 2021. The keywords used were "household water treatment," "point-of-use," "drinking water," and "Southeast Asia." For gray literature, first, a custom Google search was carried out to identify relevant organizations and websites publishing data and documents on water poverty and HWT practiced in SEA. Data were also extracted from major international survey programs that closely collaborate with WHO/UNICEF JMP, including the Multiple Indicator Cluster Surveys (MICS) and Demographic and Health Survey (DHS). For peer-reviewed literature, the following databases were searched using Pubmed, Embase, Web of Science, and Proquest. Next, each interest organizations, websites, and databases were searched for potentially relevant documents by skimming and scanning executive summaries, abstracts, and/or tables of contents. The screening of reports, websites, and papers was followed by a full-text run for data extraction, analysis, results reporting, and discussion. Finally, 50 peer-reviewed papers and eight official reports from international organizations were selected to formulate this review paper. The breakdown of the number of documents according to specific SEA countries is shown in Figure 1.

| FEASIBILITY AND SUSTAINABILITY OF HWT IN SOUTHEAST ASIA
This review gathered 50 peer-reviewed journal papers and eight official reports from international organizations (e.g., DHS, MICS, and WHO/UNICEF JMP) that targeted each of the SEA countries in the context of HWT (Table 1). The most common subjects discussed in these documents are the study of existing drinking water source quality and the necessity for a proper HWT. Other peripheral topics include the drivers of sustainable HWT, frequent recontamination of water due to poor efficacy of HWT and safe storage, and user's willingness to pay (WTP) for HWT or a better drinking source. As summarized in Table 1, all the studies in Cambodia cover either rural or peri-urban areas; more than 50% of the studies in Vietnam cover the rural areas; most studies in Indonesia cover both rural and urban areas. More papers study HWT in rural areas (27 papers) compared to urban areas (20 papers).
There are 31 papers discussing HWT on improved water sources and 22 papers on unimproved water sources. According to the official reports, 54% of the urban population in Cambodia has piped water connected to their household, while 30% of the rural population depends on public tap water (The DHS Program, 2015). However, the preference for rainwater increases dramatically to 41.4% in the rural population during the rainy season as it is abundant, and many households favor the taste. In Laos and the Philippines, the preference for bottled water is very high in both urban and rural populations ( Program, 2018b). In Thailand, 44.9% of the urban population has piped water, followed by 39.5% using bottled water while 52.6% of the rural population uses bottled water (National Statistical Office of Thailand, 2020). In Indonesia, 47.3% of the urban population uses bottled water, while 20.9% of the rural population uses protected dug well (The DHS Program, 2018a). In Myanmar, 47.7% of the urban population uses bottled water, and 32.5% of the rural population uses borehole . In Vietnam, 55.9% of the urban population has piped water connection while 24.6% of the rural population uses boreholes (General Statistics Office and UNICEF, 2015). Lastly, 43.3% of urban and 30.3% of rural populations in Timor Leste use piped water connections (The DHS Program, 2018c). However, studies often show that the said improved water source was contaminated after testing (Sobsey et al., 2008). This implies the deterioration in groundwater quality due to various human activities and the lack of facilities and infrastructure management (Akhtar et al., 2021). Based on these studies, the factors leading to a more sustainable implementation of HWT is poor existing drinking water sources, education on the correct perception of drinking water safety and quality, providing microfinance schemes, social influence, and trust in local, nongovernmental organizations (NGOs) have also shown positive contributions to the implementation of HWT.

| Type of HWT used in Southeast Asia
Health burden due to poor drinking water quality is an example of the Swiss cheese model. Many layers are involved in ensuring access to clean water by HWT, including source water protection, safe water transportation, different treatment methods (refer to Tables S1 and S2), and safe water storage. The common HTW practiced by the SEA countries will be discussed here.
The concerning matter in urban and rural areas is that in some SEA countries, the use of untreated water has reached up to the 80th percentile and is more prevalent in lower-income nations (Dadonaite, 2019). The implementation of HWT technologies varies across the SEA countries, as summarized in Figure 2. Based on the data collected, most rural and urban households in Cambodia, Indonesia, Timor Leste, and Vietnam prefer to use boiling as their primary water treatment method. In urban population, the values are 56.6%, 59.9%, 66.5%, and 77.9%, while in rural population, 54.4%, 78.5%, 80.9%, and 83%, respectively (General Statistics Office and UNICEF, 2015; The DHS Program, 2015, 2018a, 2018c. On the other hand, most urban and rural populations in Laos, the Philippines, and Thailand choose not to treat their water. In urban population, the values are 80.4%, 83%, and 61.4%, while in rural population, 54.1%, 71.4%, and 80.8%, respectively (Lao Statistics Bureau, 2018; National Statistical Office of Thailand, 2020; The DHS Program, 2018b). There is one exception in Myanmar, where the most common water treatment is straining through cloth, with 39.3% and 61.3% in urban and rural populations, respectively . This is not unexpected, as boiling is one of the most common and effective methods used for HWT when practiced correctly. It is the only method that has successfully achieved large-scale use and sustainability (Ojomo et al., 2015;Sobsey et al., 2008). This could be because governments often promote boiling compared to other treatment methods (Sodha et al., 2011). Traditionally, rural households use firewood or biomass fuel for boiling as these resources are freely accessible (Du et al., 2018;World Bank, 2020). This is because even though the level of electrification in rural areas is quite high except in Myanmar (60%) and Cambodia (83%) (Ritchie et al., 2020), according to Stoner et al. (2021), in 2020, it was estimated that 600 million Eastern Asia and SEA population still relied on polluting fuels (i.e., wood, crop residues, dung, charcoal, or coal) for cooking. Therefore, on certain occasions when the access to fuel become unreliable, boiling can be a major barrier, especially for low-income household, as the fuel cost often takes up 20% of the household energy requirements (Juran & Macdonald, 2014;World Bank, 2020). Moreover, the traditional method of burning biomass is inefficient and associated with increased household air pollution, releasing large quantities of pollutants like particulate and carbon monoxide, jeopardizing human health (Juran & Macdonald, 2014;World Bank, 2020).
On the other hand, Laos, the Philippines, and Thailand have a high percentage of improved sources of drinking water in both urban and rural regions; 78.3% and 96.7%, 92.6%, and >97.3%, and both >99% in rural and urban Laos, the Philippines, and Thailand respectively (Lao Statistics Bureau, 2018; National Statistical Office of Thailand, 2020; The DHS Program, 2018b). Improved water sources, by the nature of their design, can supply clean water (UNICEF & WHO, 2022). Thus, households might assume that the water is safe to drink and there is no need to treat it at point-ofuse. These findings also suggest why households with reduced access to improved drinking water sources perceive that their water is unsafe and tend to treat their water first. Nevertheless, improved water sources may not be safe to consume in practice due to variability in terms of quality (Ritchie & Roser, 2021).
Secondary and tertiary HWT methods practiced in SEA are ceramic filtration, sand filtration, straining water through cloth, or allowing it to stand and settle. The most utilized ceramic pots in rural areas are those designed by Potters for Peace, the colloidal silver-impregnated ceramic filter (CWP; Potters For Peace, 2019). In a field study in Cambodia, 62% of control households had drinking water at home with more than 101 CFU/100 mL of Escherichia coli compared to 9% of CWP users' households (Brown & Sobsey, 2007). In another study, Curry et al. (2015) reported that biosand filter significantly reduced turbidity by 86.7%, with E. coli log reduction of greater than 2.2 for the lake water from floating villages of Cambodia. The simplest sedimentation method is plain settling. An assisted, faster sedimentation can be achieved with the use of coagulants or flocculants like P&G Purifier of Water sachet (containing ferric sulfate as coagulant additive) and calcium hypochlorite as a disinfectant (Charoenpo, 2017;CSDW, 2021). The latter is being used daily in a few villages around SEA, including Gwak Tauk Chaung village and Hlaing Thar Yar Township in Myanmar (Charoenpo, 2017;CSDW, 2021) and Malaka in Indonesia (CSDW, 2017). Based on the Vietnam dataset, the percentage of households using filters increases with increasing wealth index quintiles (General Statistics Office and UNICEF, 2015), suggesting that cost might be one of the major barriers to implementing filters as HWT. In addition, it is also worth noting that in all these countries, the percentage of the population using chlorine is only between 0% and 4.1%, with the highest percentage coming from both urban and rural Vietnam with 1.7% and 4.1%, respectively (General Statistics Office and UNICEF, 2015; The DHS Program, 2015, 2018a. This suggests a low preference for chlorine treatment due to its overwhelming taste and odor despite its convenience. Overall, data shows that the preference for HWT in both urban and rural populations in each country is often similar. Safe water storage is crucial among communities that store water at home in sufficient quantity for potable use to protect it from recontamination once the water has been treated (CDC, 2012). Recontamination of clean water during storage and handling can predispose consumers to the risk of diarrhea (WHO, 2017). Containers used for proper storage should have tightly closed lids, be durable with a simple structure for ease of cleaning and if possible, tap to collect water without touching the body of water. The common receptacles range from plastic containers, metal cans, and Jerry can to ceramic pots, depending on the local availability and cost (Holman & Brown, 2014;Lantagne & Clasen, 2012;Oezdemir et al., 2011;Vannavong et al., 2018). A study in Cambodia reported that 32% of the households practice unsafe water storage (uncovered; Brown et al., 2009). While in another study, 52% of the households stored boiled water in a covered container, and 49% had a dedicated dipper (Brown & Sobsey, 2012). However, using plastic cups or dippers to collect water often leads to recontamination as the hand meets the water (Thomas et al., 2013). A study in Indonesia reported 97% of households use covered containers and 50% use narrow-mouthed containers (Sodha et al., 2011). In Laos, the primary storage containers are smaller jars and plastic buckets cleaned biweekly, and in Thailand, the primary storage containers are large cement jars cleaned less frequently than the households in Laos (Vannavong et al., 2018). In Vietnam, many households collect and store rainwater in larger containers made of cement, plastic or ceramic materials with 65% of the households own one or more ceramic filters with a cover, and 14% own such containers without a cover (Li et al., 2016). The households reported that the materials of the containers affect the taste of rainwater and ceramic container gives the best tasting water (Oezdemir et al., 2011).

| Water poverty and safety in Southeast Asia
The lack of access to safely managed water services (located on-premises, available when needed, and free from chemical and microbial contamination) is multifactorial. From the analysis of documents reported under Section 3, economic water scarcity plays a major role in SEA (Digital Economy Act, 2017). In SEA, agricultural-producing countries have a higher percentage of the rural population, with the majority coming from Cambodia (75.8%), followed by Myanmar (68.9%), Timor Leste (68.7%), Laos (63.7%), and Vietnam (62.7%) as summarized in Figure S1. Rural areas mainly steer the agrarian economy, and it contributes to a smaller share of gross-domestic-product (GDP) in countries such as Myanmar (26.2%), Cambodia (23.4%), Laos (16.2%), and Vietnam (15.3%) compared to industrial sectors in urban areas (Roser, 2013;UNESCO, 2018). One of the reasons is low agricultural productivity, contributed by cheap labor, heavy reliance on small nonmechanized farms, and difficulties in gaining market access while competing with capitalintensive farmers globally (George, 2015). Thus, the economic gap further aggravated the social divide between urban and rural populations.
The socioeconomic divide between rural and urban people also reflects the household income disparities, with rural regions significantly linked to poverty . This has led to the megatrend of internal migration of rural communities to metropolitan cities. In Vietnam, almost half of the internal migrants (49.8%) moved from rural to urban, and over-urbanization was observed in big cities like Hanoi and Ho Chi Minh City (Vo, 2021). Urbanization leads to congestion, pollution, and improper management of water, waste, and sanitation services. Thus, the informal settlements facing economic and water poverty are forced to rely on unclean water sources. Urban migration is often assumed to be reducing the urban-rural gap at first. However, it further intricates the inequalities between the richest and poorest wealth quintiles (UNICEF & WHO, 2019). UNICEF and WHO (2019) reported that even though the urban-rural gap in Laos in terms of access to basic water coverage has reduced to less than 20 points, the gaps between the richest and poorest still exceed 30 points. Population from the poorest wealth quintiles in the rural localities face persistent inequality in water provision services (UNICEF & WHO, 2019).
Rural households tend to collect their water from a public standpipe, water service shared in the community, or open water sources. Figure 3 shows that both urban and rural areas of Brunei, Singapore, and Thailand have a high percentage (>99%) of water sources on the premises. In Vietnam, 99.2% of the urban population also has access to water on their premises. However, in rural areas, the percentage is slightly lower as 4.5% of the rural population utilizes surface water (UNICEF & WHO, 2021). For the remaining seven countries, the proportion of people that access water on their premises decreases significantly in rural areas as they need to travel within 30 min or more to get their water. However, in Cambodia, Indonesia, and Myanmar, accessible water on-premises is relatively low, even in urban areas. Cambodia has the lowest proportion of accessible water on-premise in urban and rural areas with 68.4% and 42%, respectively. Besides, it also has the highest proportion of people who travel more than 30 minutes to fetch water in urban and rural areas with 8.9% and 15.5%, respectively. In a Vietnam case study, 85% of households reported no difficulty, while 14% reported difficulty in obtaining drinking water (Li et al., 2016). In addition, 43% indicated that they must purchase water, and 66% will either store more water, consume less, or borrow water from their neighbor. Although a high percentage of households reported no difficulty, it is possible that they may be accustomed to the inconvenience (Li et al., 2016).
With focused attention, the intra-countries socioeconomic divide metamorphoses into water provision inequalities SEA countries. Singapore has complete access to safely managed water services, while Brunei and Thailand populations achieved 100% access to at least basic water services in urban and rural areas (Figure 4a). In 2020, the population percentages of using safely managed drinking water in Laos (17.7%), Cambodia (27.8%), the Philippines (47.5%), and Myanmar (58.8%) were significantly below the global average (74.3%; UNICEF & WHO, 2021). Cambodia has the highest urban-rural gap in access to at least basic water services (90.5% urban and 65.1% rural; Figure 4a). Although urban regions have better access to at least basic water services, most rural areas of SEA have a greater positive annual change in the access to these services than urban areas (Figure 4b). Laos has the fastest growth rate of 2.1% in improving these services in rural areas. However, Malaysia is an exception where rural and urban regions have a negative growth rate in improving access to at least basic water services with À0.14% and À0.01%, respectively. Physical water loss is a worrying issue in Malaysia, and it could be one of the contributing factors to the decline of at least basic water services growth rate (Jones et al., 2021;See & Ma, 2018). An unusually high non-revenue water compared to other SEA countries (Malaysia's NRW in 2017: 5929 million liters per day) due to aged water distribution infrastructure, added with increased water consumption due to population growth (increased population by 40%, from 23.19 million in 2000 to 32.37 million in 2020) were observed in Malaysia (Jones et al., 2021;See & Ma, 2018). The increased water consumption and water loss are further exacerbated by frequent pollution of raw water sources of drinking water treatment plants (Maniam et al., 2021). Ongoing studies are addressing the high NRW in Malaysia (Jones et al., 2021). The repercussion of lack of access to clean water comes down to human health risks.  WHO has listed safe water supply, improved hygiene, sanitation, and water management as key factors in maintaining health. Drinking water free from microorganisms and chemicals is a quintessential element because contamination could transmit waterborne diseases (WHO, 2019). A meta-analysis reported that water sources in low-income countries and rural areas are more likely to be contaminated with E. coli or thermotolerant coliforms with odd ratios of 2.4 for both (Bain et al., 2014). In 2016, rotavirus caused 27% of diarrheal deaths globally, followed by shigella (13%), adenovirus (11%), and cholera (11%) . Widmer et al. (2013) reported that 3.9% of the 564 isolated E. coli river water from Thailand, Indonesia, Cambodia, and Vietnam had at least one virulence gene such as Shiga toxin-producing E. coli. When considering the economic status and diarrheal death rate in children under 5 years old in each country, the lower-middle-income economic nations in SEA reported higher diarrheal death rates, except Vietnam (Table 2). However, among the low-income countries, the diarrheal death rate does not necessarily correlate with GDP per capita, but rather it reflects access to an improved water source. For instance, in the lower-middle-income countries, the Philippines reported lower and Vietnam exceptionally lower diarrheal death rates despite not having the highest GDP per capita. Yet, access to improved water sources is high in these countries. In the case of Vietnam, the key contributors to reducing diarrheal death are the introduction of community-based programs for water, sanitation, and infrastructure and prioritizing universal primary education (Kuruvilla et al., 2014). This framework could be replicated in other SEA countries. In a Cochrane Collaboration systematic review, household interventions were found to be more effective in preventing diarrhea than interventions at water sources (Clasen et al., 2006). This emphasizes the need for effective HWT and safe storage options for vulnerable populations, especially those under the poor wealth quintiles. In the next section, factors associated with the feasibility and sustained implementation of HWT in SEA will be explored.

| Factors affecting implementation of HWT in Southeast Asia
Historically, water supply interventions in developing nations had issues with a lack of sustained and positive project outcomes. The transfer and maintenance of clean water technology usually require constant external support for its sustainability due to its sophisticated technology. Five major factors namely, user preferences, user perceptions, education and training, economic feasibility and collaboration, and supportive policy environment, were identified to affect the feasibility and sustainability of the HWT adopted in SEA as shown in Figure 5.

| User preferences
Technology preferences which are one of the user preferences, refer to the technical characteristics, including ease of operation and incorporation into daily life, and time required in practice which might influence people's preferences. A study in Cambodia shows a low uptake rate of ceramic filters as 26% of the households are not aware of the distribution points (Brown et al., 2009). The decline was reported to be 2% per month since implementation because the filters tend to break easily and cannot fulfill house demand as a result of slow flow rates (1-3 L/h). This is due to the tendency of the filter being rapidly clogged by insoluble iron contained in groundwater from a deep well (Brown et al., 2009). Cambodian households also think that boiling is time-consuming and costly as they need to purchase fuel (McGuigan et al., 2011). The average time for boiling takes up to 20 min, excluding gathering or purchasing fuel and waiting for the water to cool down (Brown & Sobsey, 2012). Meanwhile, in Indonesia, a study reported 83% saying boiling was practical, 73% said it was easy, and 90% believed it was cheap because their main fuel is subsidized propane (Lantagne & Clasen, 2012;Sodha et al., 2011). Although boiling is the most common practice of treatment, bottled water for drinking is also popular due to low pipe water coverage, drinking water quality, and added convenience (Prasetiawan et al., 2017). In rural Vietnam, sand filters are often used to remove arsenic in groundwater because it improves the taste of water and can treat large quantities of water (Lantagne & Clasen, 2012;Tobias & Berg, 2011).
Individuals often consider the taste and smell acceptability of treated water a crucial factor when choosing the HWT method that it even outweighs the health benefits provided (Jeuland et al., 2016;Orgill et al., 2013). In Cambodia, ceramic and sand filters are often seen in the household premises, however, not used by the households to produce clean drinking water (McGuigan et al., 2011). This is because the taste of filtered water is not the same as the more common boiled water (McGuigan et al., 2011). Respondents from a study in Cambodia also expressed their unwillingness to consume chlorinated water once the dosage surpasses 1.3 mg/L despite being lower than the recommended value of 2 mg/L (Jeuland et al., 2016). Similarly, chlorination and ceramic filter are not widespread among Jakarta's lower-income population despite the ease of preparation and use, and low cost due to lack of appreciation of taste and odor of treated water (Roma et al., 2014; Walter et al., 2017). Rainwater and treated water using SODIS are often F I G U R E 5 Relationship between factors that affect feasibility and sustainability of HWT in SEA. favored due to better appreciation of the taste and smell (McGuigan et al., 2011;Oezdemir et al., 2011;Roma et al., 2014). Many residents in Vietnam also prefer the taste of rainwater and are unlikely to purchase bottled water even when sold at low price because rainwater is available (Li et al., 2016).

| User perceptions toward water safety
Using a certain HWT highly depends on the user's perception of clean and safe drinking water and associated social status. This is affected by multiple factors such as level of education, knowledge about clean and safe drinking water, and who manages the household. A study in Cambodia reported that 62.3% of their respondents thought their water was good quality yet 97% still treat their water through boiling or filtering (Guppy & Shantz, 2011). Conversely, 37.7% of respondents reported that groundwater is not of good quality as it looks dirty, tastes bad, and makes people sick. It is also found that in-house water quality is often overestimated, and households' perception of water quality is not highly correlated with the actual E. coli count in the water (Orgill et al., 2013). In Indonesia, people were more likely to boil their water if they sought to protect stored drinking water, have knowledge indicators, and if female respondents attended school (Irianti et al., 2016;Lantagne & Clasen, 2012). Aini et al. (2007) and Daniel et al. (2020) supported that a mother's education or a woman's role is an important socio-economic factor influencing the use of HWT. A study in Flores Island, Indonesia, shows that despite villagers knowing the advantages of treating water, none reported health risk as a factor influencing technology use (Roma et al., 2014). The high consumption of bottled or refillable water in Indonesia is also associated with user's perception that it is a safer drinking water source, provides health benefits, and is more affordable despite having the second highest per unit cost after branded water bottle (Komarulzaman, Smits, & de Jong, 2017;Prasetiawan et al., 2017;Walter et al., 2017). In Indonesia, drinking bottled water has become a taste and lifestyle choice . Similarly, in Malaysian and Philippines studies, households often opt for bottled water or use a home purification system as it is convenient, has improved taste, is perceived to be safer and healthier, and is associated with rising social status (Aini et al., 2007;Francisco, 2014;Palanca-Tan, 2017).
In a study in Vietnam, a district was reported to have invested in piped water for the community as it is reliable and clean (Anh et al., 2020). Yet, households with piped water connections choose to draw their water from lower-quality groundwater due to high connection fees or poor perceived water quality (Anh et al., 2020;Wilbers et al., 2014). A study in the Philippines reported similar results (Sarajane et al., 2018). Another study reported that social and affective influences are more critical than knowledge of drinking water contamination (Tobias & Berg, 2011). This was observed when the study showed that critical knowledge regarding arsenic contamination in drinking water is unnecessary for the sustainable use of sand filters. In a developed country like Singapore, tap water is considered safe to drink without additional treatment, although it is still common for them to boil or filter their water despite high energy requirements, installation, and maintenance costs (Li et al., 2019). However, like other countries, they also perceive that bottled water is safer and has more stringent quality standards than tap water.

| Education and training
Education and training are pertinent ways of bridging the gap between household perceptions toward water management and the sustained implementation of HWT and safe storage (Graham et al., 2016;Jalal, 2014;The DHS Program, 2018c). Several studies in Cambodia had shown that education and informational campaigns are significant in improving the demand for HWT (Guppy & Shantz, 2011;Orgill et al., 2013). The study by McGuigan et al. (2011) also showed high compliance with SODIS HWT as the users first attended an information briefing before the implementation. However, educational intervention needs to change the misconception of the safety of existing water sources and not just on water quality alone (Orgill et al., 2013). Studies in Indonesia have reported the importance of the education of a female household head who manages the water provision in the house (Daniel et al., 2020;Irianti et al., 2016). Meanwhile, during an emergency response in Indonesia, the reason for the failure in the uptake of chlorine is the unfamiliarity of the product and the inability to understand the English instructions (Lantagne & Clasen, 2012). Komarulzaman, Smits, and de Jong (2017) suggested that education and training should be carried out to ensure HWT is executed effectively. This shows the importance of training in proper use and implementation of HWT. Many of the studies in Vietnam focused on education through a comprehensive health promotion or awareness campaign (Khuc, 2013;Seino et al., 2008;Vo & Huynh, 2017) while the training often focuses on proper treatment and safe storage practices (Grady et al., 2015;Heller, 2019;Seino et al., 2008).

| Economic feasibility
Affordability and income constraint are common issues influencing the willingness, successful adoption, and sustainability of HWT (Blanton et al., 2014;Brown et al., 2009Brown et al., , 2017Clasen, 2015;Daniel et al., 2018;Ojomo et al., 2015;Orgill et al., 2013;Stubbé et al., 2016). In a study involving 506 households in three areas of Cambodia, 156 households regularly used ceramic filters, yet only 120 reported the willingness to purchase a new or replace the filter (Brown et al., 2009). The WTP for a new filter range from USD 1.5-3. However, respondents who used filters daily have higher WTP ranging from USD 2.5-5, which is in the range of replacing ceramic filter elements in Cambodia. This study also reported a more sustained filter use when cash investment is involved. In another study, PATH collaborated with HWT manufacturers and microfinance institutions (MFIs) to study the impact of consumer finance models on the uptake of HWT (Blanton et al., 2014). Small loans provided by MFIs to the low-income users had positively affected the uptake of water filters (Freeman et al., 2012). In the pilot study in Cambodia, Vision Fund collaborated with Hydrologic, a ceramic water filter manufacturer on the middle wealth quintile households (Blanton et al., 2014). The number of MFI clients increased from 6% to 43% during the study. It was reported that the demand for water filters also reached people who are beyond the MFI clients, where there is a substantial uptake rate for the lowest (poorest) and second quintile in Cambodia compared to India. The pilot study worked well in locations where water treatment is already commonly practiced. They feel the necessity to treat water and have substantial knowledge about the promoted water filter. Thus, this commercial model is considered successful in Cambodia as represented by the 90% of the households who felt the necessity to treat water and 95% of them who already practice treating their water in Cambodia. The pilot study also achieved 102% cost recovery.
A study in Vietnam across 235 households in urban and rural areas reported 51% of the households agreed to pay for improved water quality (Khuc, 2013). However, this is mostly contributed by the household in urban areas suggesting that location influences WTP. The WTP in Hanoi, Ho Chi Minh, and Hai Duong range from USD 0.3-7.5, USD 0-7.5, and USD 0.4-25, respectively. Besides location, income, and perception towards safe water are also important determinants affecting WTP. People active in learning water-related knowledge are more WTP for safe drinking water. Thus, people in rural areas aged 60 years or more and unemployed are less WTP. A study conducted in the Mekong Delta rural areas reported rainwater to be the main source of their drinking water, especially in the wet season (Li et al., 2016). They prefer rainwater taste, making it unlikely to purchase bottled water (safe water) even with limited containers for collection and low price. About 83% of 184 households reported a high possibility of purchasing 20 L bottled water at no cost in the wet season. In the dry season, the likelihood of purchasing bottled water increases from 52% to 66% at USD 0.2 and from 32% to 48% at USD 0.5 compared to the wet season. Like the previous study by Khuc (2013), age was again reported to be negatively correlated with WTP due to the perception toward water.
In Malaysia, a study across three districts in Kelantan shows that household income is directly proportional to the WTP of having water filters at home. Those with incomes lower than RM 950 are not WTP for water filters (Awang et al., 2020). In Butuan City, the Philippines, the WTP for an improved water supply is generally low as they prefer to buy ready purified water (Sarajane et al., 2018). The study reported 33.2% of the respondents were not WTP, 43.2% WTP less than USD 1, and 19.6% WTP between USD 1-2. Their WTP improved when the households had higher food consumption due to food security and general wellbeing. However, their ability and WTP reduced with higher electricity bills due to tightened budget. A study in Timor Leste reported that each household needs to pay USD 0.3 per month for rural water services' maintenance and repair fees (Neely & Walters, 2016). This amount is considered fair and relatively manageable for most people, and they may be willing to pay more, especially when they do not have to transport the water from a distance. The minister of internal affairs regulation no.23/2006 in Indonesia stipulates that piped water cost should not exceed 4% of household income ). Yet, low-income households are still willing to spend 7%-10% of their total expenditure on individual water connections. One study in Indonesia supports this showing that rural households are willing to pay up to 7.6% of their monthly expenses (Suparman et al., 2016). This may be attributed to better awareness of the importance of availability, quality, and the risks of consuming polluted water, increasing their preference for piped water. A study by Nastiti et al. (2017) shows that households with higher income pay more for water, contrary to the popular belief that the poor often pay more for water, driven by the reliance on water vendors who price water at 4-30 times more than public piped water. According to Walter et al. (2017), in the absence of cross-subsidy through a differentiated water tariff structure, poorer households are paying more per unit volume consumed. However, refill water vendors usually have an informal credit system to help poorer people refill water and pay when they can (Walter et al., 2017).

| Collaborations and supportive policy environment
The adoption status of HWT, especially in lower-income countries of Southeast Asia, is still low and variable. The success of HWT is interlinked between the degree to which they are made available to the public and the scale to which they are adopted, used consistently, and adequately (Naman & Montgomery, 2012). Therefore, collaborations and supportive policies are paramount to achieving this goal. A study in Timor Leste shows a causal loop between donors, international and local NGOs, and the community as stakeholders (Neely & Walters, 2016). It was stated that most NGOs tend to continuously engage in new projects to improve their reach and grow. This is driven by short-term funding donors prioritizing extended coverage instead of sustained outcomes. One village in an Indonesian study reported low implementation of HWT due to low trust in local NGOs (Roma et al., 2014). Community-based management in rural areas is also fragile and susceptible to failure. In Indonesia, they often do not have a comprehensive plan and rely more on existing culture (Irianti et al., 2016). The only statistically significant social variable was the participation in PKPS-BBM (Irianti et al., 2016). In Timor Leste, it is often due to a lack of incentives for the person managing, lack of trust from the committee, and funding shortfall (Neely & Walters, 2016). Meanwhile, WHO has identified five key policy and regulation criteria of HWT, such as incorporating HWT in national policies with a committee for coordination, compulsory regulation of chlorine and nonchlorine-based HWT technologies, and accessibility to at least HWT standards (WHO, 2016). Vietnam is one of the low-income SEA countries with at least four criteria, whereas Cambodia and Indonesia have achieved three out of the five elements (WHO, 2016). For Myanmar, no information on the inclusion in national policy, regulation, or standards of HTW products can be found (WHO, 2016).

| WHAT CAN WE LEARN FROM CURRENT PRACTICE?
Based on the discussion in the previous sections, it can be observed that the implementation of HWT is not a one-sizefits-all solution. In Indonesia, the highest rates of effective use of HWT combined three factors including targeting households that consume contaminated water, providing a HWT method that effectively treats the water, and providing to the population familiar with the product or practice, willing to be trained and eventually use it (Lantagne & Clasen, 2012). The implementation drops significantly when either one of these factors is left out. Solving and realizing this water problem requires understanding the complex interaction between multiple HWT implementation factors (DuBois et al., 2010;Luoto et al., 2011Luoto et al., , 2012 and can be categorized into three overlapping segments, namely social, economic, and environmental factors. User is the key component of social factors. Lantagne and Clasen (2012) reported that effective use of HTW technologies will not be greater than 15% when minimal training is provided and if the technology requires more than two steps to operate. Maintenance of biosand filter, for example, is challenging as the biofilm layer is sensitive to chlorinated water and heavy turbulence when pouring the water (Dangol & Spuhler, 2020). HWT technology preference is also a reflection of gender gap in water-related leadership (Jalal, 2014). A demographic survey showed noteworthy gender inequality in terms of responsibility to collect water for domestic use, where in 27.3% of rural households, females are responsible for collecting water compared to only 12% of males (The DHS Program, 2018c). While females are the primary household water decision-makers, their participation in decision making and implementation of water initiatives even at community levels is insignificant (Svahn, 2011). HWT generally does not fit into the niche of untrained female. This could explain the wide adoption of boiling in SEA, which is within the logistic and technical know-how capacity of most females without additional training.
HWT is more important for vulnerable communities relying on unsafe water and unfortunately, Indigenous people often fall under this bracket. In some communities, reliance on surface water such as streams and rivers is on one's own will. Indigenous people tend to show a special bond with nature and practice unique knowledge and beliefs, especially in the sustainable management of environmental elements (Maniam et al., 2021). Local water knowledge often fails to intersect with the scientific understanding of water treatment. For example, water services provided to some of the Indigenous communities of Peninsular Malaysia were reported to be lacking in sustainability due to incompatibility with their cultural value, in addition to the high cost of maintenance (Heller, 2019). The locals do not prefer drinking chlorinated water as it is not their traditional practice to consume water containing chemicals. Thus, promotion of the use HWT or even water delivery system that used chlorine or other additives for water treatment might not be successful among populations with similar beliefs and who find the chemical taste is unacceptable. Roma et al. (2014) emphasized the need for pre-implementation studies to consider factors promoting sustained use, which are most likely to be noneconomic components.
Water literacy, which encompasses water knowledge, attitudes perceptions, and behaviors are interlinked with sociodemographic factors such as gender, age, level of education, ethnicity, and economic status. Health perception plays a critical role in the HWT implementation (Roma et al., 2014). In Malawi, a year after an intervention in promoting chlorine-based disinfectant by the health care personnel among the pregnant women, an increased awareness and proper sustained use of the disinfectants observed in the communities (Ojomo et al., 2015;Sheth et al., 2010). Trusted sources such as health experts played a role in positively impacting water literacy and adoption of HWT (Ojomo et al., 2015;Sheth et al., 2010). In this review, it can be observed that some households have the perception that their piped water or other drinking water sources like harvested rainwater and groundwater is safe to drink and there is no need to treat their water. In Timor Leste, people are willing to pay a higher amount for piped water to their houses. Main pipe water is often assumed to be safe for consumption, as most would have been receiving water with quality meeting WHO standards. However, this is not always the case. Many times, the disparity in water quality was observed between treated water leaving the treatment center and water reaching the household premises. The deterioration in water quality is multifactorial, including due to old piping systems as aforementioned for Malaysia or underground water leakages, which might introduce fecal and priority chemical contamination during distribution. Flint water crisis in 2014 that involved lead contamination in the municipal water supply system is an example of public health crisis that challenged the water safety perception towards piped water supplies. Awareness of the Flint residents on the health issues rooting from the contaminated water was the ignitor to resolve the issue. In a nutshell, education is being the core of the social factors influencing the sustainability of HWT which encapsulate the HWT awareness creation, technology training, vulnerable communities' empowerment in water services.
Economic factor is one of the major elements affecting the clean water provision service or product purchasing power of households and organizations. Higher preference of public piped water in certain countries like Singapore is contributed by the low water tariff and availability of centralized water distribution throughout the country. (UNESCO World Water Assessment Programme, 2019) reported that Indonesia, Laos, Myanmar, Philippines, Thailand, and Timor Leste have less than 50% financial resources needed to implement drinking water provision plans, especially in rural areas. This explains why usage of unimproved water sources is still observed in SEA. Traditionally, the provision of water and sanitation services was the responsibility of governments. However, as the population grew with rapid urbanization and industrialization, the challenges of managing water supply and catchment areas took its toll on the governments' financial stress. In order to improve the efficiency and sustainability of these services, governments; started introducing public-private partnerships to foster cooperation between governments and the private sector in water projects. This includes the construction and maintenance of storage dams and treatment plants, installation of pipes for water distribution to users, and repair works of water sector facilities. Treatment plants that need enormous capital but are remunerative, dominate the privatized projects list and are located mainly in urban areas. Thus, particularly the remote areas are left behind in the expansion of the centralized water supply due to low return of investment. While use of HWT has been proven to reduce water-borne diseases, affordability of the technology is a hurdle to overcome. In a study in rural Kenya from 2003 to 2005, it has been observed that initial use of a HWT with the motivation of reducing water turbidity eventually weaned and replaced with cheaper and more accessible product even though it did not address the turbidity issue (DuBois et al., 2010).
Many times, these services and products are offered for free. However, according to Brown et al. (2009) andOjomo et al. (2015), the use of HWTs is more sustainable when there is a form of capital investment compared to receiving in for free. Free goods results in the lack of buy-in or investment by the users, and possibly a lack of ownership in the usage (Blanton et al., 2014). This was proven in a case study in Tanzania where biosand filters were not used when given for free. Nevertheless, its use increased after they began buying the filters (Ojomo et al., 2015). In developing countries with limited funding, subsidies from nongovernmental organizations are crucial in providing a more affordable price for water treatment technologies. The governments should actively collaborate with MFIs such as Vision Fund and Water Credit to develop microcredit schemes to help households under water poverty to have access to HWT that have proven health benefits.
Unfortunately, many people, especially in the rural areas, still need to travel a distance to fetch water which mainly falls under the responsibility of women and children. Rural localities, especially remote areas, are not infrastructurally well-equipped with tarred roads and streetlights. Thus, traveling in these conditions, especially during hours of darkness, is unsafe. Some villages get water from open surfaces such as lakes and river streams. Thus, walking to water sources for a refill will be time consuming and exhausting. Systems that require a manual refill of water cannot provide a consistent supply of freshwater. Spending a huge portion of their time fetching water leads to the females missing out education or job opportunities and in turn affects the proper and continuous use of HWT as it becomes an added burden. HWTs are not a competitor for improved water sources supply, but rather complements the service to provide clean water. For example, in Singapore and Malaysia, even though a high percentage of the households receive piped water, boiling or filtering water is still commonly practiced. In areas where centralized improved water sources are not available, small networks of water service managed by the municipality and local community can be developed to further support the use of HWTs. Small town water service provider in Nepal is an example of sustainable model of delivery service which integrated thorough planning and designing, including (i) project funding (external aid, internal resources, community WTP), (ii) eligible small towns or municipalities registration, (iii) project designing with input from consultants, (iv) feasibility studies, (v) detailed design reports, (vi) recommended tariffs for cost recovery, (vii) operation and maintenance services agreement, and (viii) procurement of services (Pokhrel & Adhikary, 2017;Rautanen & White, 2018). These small networks could be supplemented by HWT integrated into the delivery service model.
The small water service projects in Nepal are an example of successful story that illustrated the importance of co-designing concept in its implementation. Many HWT studies reported that despite the efforts to overcome barriers that exist in the scaling up of HWTs including cost and education, low sustained usage of HWTs products were observed (DuBois et al., 2010;Luoto et al., 2011Luoto et al., , 2012Nielsen et al., 2022;Pickering et al., 2019;Rosa et al., 2016;Schmidt & Cairncross, 2009). However, they agreed on the need for better understanding of end-user's choices, preferences and aspirations when it comes to HWTs. Recent papers discussing the design of sustainable HWTs, emphasized the importance of focusing on users (Deng, 2021;Nielsen et al., 2022;Ojomo et al., 2015). While most previous studies were designed to address a user's need, co-designing concept exemplified in the Nepal's initiatives, bring the users to the table of discussion during the planning stage itself. This will further strengthen the sustainability of the introduced HWT in a certain community or region.
Environmental factors are the least discussed component of HWT sustainability (Deng, 2021). Every water provision initiative be it in the form of a product or service has environmental impact associated with it. Boiling has been widely promoted to safely manage water for potable use in SEA (Ojomo et al., 2015;Sobsey et al., 2008;Sodha et al., 2011). Murray et al. (2020) reported that indoor air pollution is globally one of the largest environmental problems, especially for the poorer countries who still heavily rely on unclean fuels for cooking and heating. In 2016, 281 million SEA people had no clean fuels such as natural gas, ethanol, or electric technologies and 896,382 deaths in 2019 is contributed by indoor air pollution (Murray et al., 2020;Ritchie et al., 2020). Besides, postconsumer plastic waste generation with other types of HWT such as coagulants, flocculants, and disinfectant bottles and sachets not widely reported. While these HWT products are commonly distributed to populations from the poor wealth quintiles who need clean water, these low-to-middle income countries are documented to have poorer waste management facilities (Ritchie & Roser, 2022). Philippines, for example, reported to have hundred times mismanaged plastic waste (either litter or inadequately disposed) per capita compared to United Kingdom and third contributor (6.52%) of global mismanaged plastic waste (Ritchie & Roser, 2022). While environmental impact assessment of HWTs might not have an immediate effect on users, the significance of the impact cannot be rendered unseen.

| CONCLUSION
There is an existing urban-rural gap in terms of having at least basic water services in most of the SEA countries. The proportion of the population in Laos, Cambodia, Philippines, and Myanmar using safely managed drinking water is far below the global average of 74.3%, and water provision inequalities predominate in the rural localities. Areas without access to at least basic water services could benefit from HWT with proven health benefits. Current HWT available in SEA include boiling, chlorination, SODIS, assisted sedimentation, and filtrations. The systems could be used as a multibarrier strategy for HWT. The pros and cons of these systems could be used to customize the choice of HWT to a particular locality. However, the sustainability of HWT implementation in SEA countries is considerably low because most households prefer to treat their water by boiling, opt for bottled water, or not to treat their water at all as they believe the water is safe to consume. Improper storage practices also often result in recontamination of drinking water. There are also other factors that affect the feasibility and sustained used of HWT including user preferences, user perception toward safety, education, economic feasibility and collaborations, and supportive policy environment. To provide a holistic HWT solution, these interdependent factors need to be considered. This can be done in various ways such as considering the geographical and demographic context of the households, providing trainings to correct their perception of water treatment and practice HWT properly, communicating the cost of supplies transparently, making consumables and supplies more accessible, and developing microfinance schemes through intergovernmental initiatives to encourage and target HWT use more effectively. In a nutshell, co-designing the HWT implementation with the users will help in its sustainability.

CONFLICT OF INTEREST STATEMENT
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.

DATA AVAILABILITY STATEMENT
Data sharing is not applicable to this article as no new data were created or analyzed in this study.