Background
It is estimated that up to 1.9 million people worldwide rely on water from ←unimproved’ sources or ‘improved’ sources that are fecally contaminated [
1,
2]. Providing safe and adequate quantities of water to households has long been identified as indispensible to achieving the Millennium Development Goals (MDGs) and for progress beyond [
2‐
4]. Recent studies have found that water from “improved” sources does not necessarily equate to “safe” water. Hence, the current estimates of the number of people having access to safe water may have been overstated [
1,
2]. Populations relying on unsafe or unreliable water supplies are often exposed to water-associated pathogens, besides the additional burden of collecting and transporting water over long distances [
3,
5,
6].
Of the water-associated infections, diarrhea continues to be a leading child-killer globally, despite a decrease in under-five deaths by nearly 50 % between 2000 and 2013 [
7,
8]. Diarrheal infections disproportionately affect children living in Low and Middle Income Countries (LMICs). An estimated 276,000 children (nearly 38 % of all deaths attributable to diarrhea) are from India alone [
9,
10]. The impact of diarrheal disease is most severe on children under two years of age, with 72 % of all diarrheal deaths occurring in this crucial stage of life [
7]. Recent estimates suggest that up to 95 % of diarrheal deaths among under-five children in the high-burden countries can be averted by scaling up water, sanitation and hygiene (WASH) related interventions [
11,
12].
The most effective WASH interventions in order of their estimated reductions in diarrheal risk are hand-washing, proper excreta disposal and improved water quality [
5]. A 2014 systematic review and meta-analysis for LMICs revealed that even though high quality piped water offers the greatest protection against diarrheal disease, filter interventions promoting safe storage are most effective at the household [
13]. While interventions aimed at improving water quality have demonstrated success in improving health and reducing diarrheal diseases among children worldwide, lack of sustained use, post-intervention, has greatly limited their benefits [
14‐
19].
Reasons for poor adoption and lack of sustained use of water quality interventions are complex and understudied [
14,
17,
20]. In rural Kenya, point-of-use chlorination with a clay pot for storage, a low-cost intervention (cost ~ USD 3) had limited success [
17]. Poor adoption of the intervention appeared to be linked to cultural factors and community preferences [
17]. In Vellore, southern India, a region with highly contaminated water supply, testing of low-cost water quality interventions such as solar disinfection, closed valved containers (cost ~ USD 2) and domestic chlorination (cost ~ USD 0.04 per 5 L treated water) are most effective at disinfecting or preserving the quality of disinfected water [
18,
21‐
23]; but barriers to acceptance and sustained use prevent the targeted communities from benefiting from these interventions, placing young children at a higher risk of water-related diseases [
18,
23‐
25].
Qualitative methods can help overcome most limitations of quantitative-only methods, and their potential in exploring complex areas such as behavior change is increasingly being recognized [
26,
27]. Qualitative research has helped provide critical insight on public perceptions and preferences concerning drinking water quality in the past [
28‐
30]. We therefore designed this study to understand parental perceptions on the effects on health of unsafe drinking water, household drinking water treatment practices and factors that influence acceptance and sustainability of an ongoing water quality intervention, with the overall aim of addressing the barriers to adoption which are known to limit the health benefits of such interventions to children and their families in rural Vellore, southern India.
Methods
Study setting
This qualitative study was conducted as part of a year-long intervention testing the effectiveness of a commercially available, high throughput membrane filter for providing clean drinking water among residents of three villages in Kaniyambadi block (a rural administrative unit) of Vellore, Tamil Nadu, India. Earlier studies from a number of villages in this block have documented high levels of fecal contamination in the drinking water available for public consumption, and a limited acceptability of other point-of-use methods for water treatment [
18,
21,
31]. The membrane filter (Skyhydrant™ water filtration system, Skyjuice™ Foundation Inc, Australia) was a commercially available, gravity-fed, source based water filtration system that has previously been pilot-tested on the residential campus of Christian Medical College (CMC), Vellore and demonstrated potential as a low-cost solution for water disinfection in smaller communities [
32]. Maintenance was accomplished through designated workers performing a weekly-backwash (with water) and a once-monthly chemical cleansing using 100 ml liquid chlorine diluted in 400 ml of water to clean the membranes, costing about INR 1,500 or USD 30 every month. Each membrane filter was placed in a designated water kiosk,
i.e. a building (taken on lease) with a cast cement roof and there were a total of 5 kiosks in the intervention villages. Study households were encouraged to collect drinking water from the water kiosks and on average, each water kiosk supplied to 20 families every day, with up to 25 L of drinking water provided to each family for their daily use.
The three villages (two intervention villages and one control village) for the intervention study were purposefully chosen from 85 villages in the Kaniyambadi block, and all households with a child less than two years of age were invited to participate. These households were largely dependent on intermittent public water supply for drinking and other domestic use and families stored their drinking water in plastic or metal pots. Chlorination of source water was done using bleaching powder fortnightly and water was mostly consumed without further treatment [
21,
23].
The intervention study commenced in October, 2013 and results of the testing of the membrane filter and the effect of providing filtered water on diarrhoea in children under-two years of age is the focus of another manuscript in preparation.
Selection of participants for the qualitative study
Selection of participants for the focus group discussions (FGDs) was purposive and focused on parents of children participating in the intervention study described earlier. Each parent was briefed about the nature and need for the FGDs through face-to-face meetings and invited to participate by field-workers performing the diarrheal surveillance in the study villages. The meetings were held at a time and place convenient to the participants, i.e., mostly in the mornings and in a village’s community hall. Meetings were held separately with both compliant (families collecting water regularly, i.e. ≥ 3 times a week) and non-compliant (families that withdrew from the study before the study end) households in the intervention villages, and with all participating households in the control village. Three key-informant interviews were held in the villages with the elected village leaders, as they are key stakeholders who directly oversee the functioning and maintenance of all basic amenities provided to the residents of their villages.
Methodology
This qualitative study used focus groups and key-informant interviews. Focus groups were used to capture parental attitudes, perceptions and beliefs which are not easily captured using other methods such as observations and surveys [
33]. Also, focus groups allow social interactions, helping highlight differences in perceptions between participants, sometimes missed with one-to-one interviews [
33]. One-to-one interviews of village leaders (key-informant’s) were also conducted to explore in-depth their attitudes and perceptions on water safety, the intervention and its sustainability in their villages.
A total of 8 FGDs were held - 3 FGDs in each of the 2 intervention villages (1 with fathers from compliant households, 1 with mothers from compliant households and 1 with either parent from non-compliant households) and 2 FGDs in the control village (1 with fathers and 1 with mothers). The FGDs and key-informant interviews were held six months after the commencement of the intervention study (between April and May, 2014). A semi-structured format was employed to collect data from the FGDs and the key-informant interviews [
33]. An FGD guide containing a set of open ended questions exploring various aspects such as the effects on health of unsafe water (e.g. In your opinion, how do you think water/water quality affects the health of the child?), water sourcing for drinking and other domestic uses, water treatment in the household, feedback regarding the intervention (e.g. Can you please tell us your opinion of water provided by the machine?) and willingness to support any future plans of implementation (e.g. Will you support any future plans to establish a service for providing this water?) was developed. The FGD guide was piloted in a village that did not participate in the intervention study and modified based on feedback from the piloting. The FGD guide was also used for the key-informant interviews. All the meetings were conducted in the vernacular language (Tamil) and audio-recorded.
The interviews were conducted by a field supervisor with extensive experience in community engagement, training on interviewer ethics, probing for detail and efficient management of the discussions. The lead investigator (MRF) was present as a facilitator during the meetings and interviews, and took notes to document participant observations relevant to the qualitative analysis. Field workers were also present for the meetings. The meetings lasted between 45–60 minutes and written transcripts of responses to the various questions were maintained (in addition to the audio-recording) to validate the data, post-transcription. Any information that could identify participants was not available during transcription or retained during analysis or subsequent reporting of findings.
Data analysis
The audio-recordings of the FGDs and the key informant interviews were transcribed verbatim from Tamil to English by a designated multi-lingual transcriber with previous health research experience. Transcripts from the audio-recording were compared to the written transcripts by the field supervisor (who conducted the FGDs and interviews) as much as possible to ensure consistency of the data transcription. Due to paucity of research on the topic of study, an inductive approach was used to identify themes from the transcripts [
34]. A thematic analysis was considered appropriate to make interpretations of the collected information. Data for each of the questions were entered in spreadsheets (Microsoft Excel). Responses to the various questions were deconstructed into emerging categories and sub themes using open coding on separate spreadsheets [
27,
33]. The consistency and relevance of the coding was examined through triangulation, which was conducted independently by investigators not associated with the data-collection (GN and VB). All participant quotes within the scope of this study have been presented to support the results, with additional text for clarification placed within square brackets, as necessary. Data collection and reporting of results are as per the RATS guidelines for qualitative studies (
http://www.biomedcentral.com/authors/rats) and the RATS checklist for the study is available as Additional file
1.
Ethics
The intervention study was approved by the Institutional Review Board (IRB) of CMC, Vellore; this also included ethical clearance for the qualitative study. Written informed consents were obtained from the parents of children recruited for the intervention study. This report focuses on the findings from the FGDs and key-informant interviews that involved only adults (parents of participating children and village headmen) from whom verbal consent was obtained prior to participation. No compensation was offered for participating in the meetings.
Discussion
Qualitative methods employed in this study allowed the researchers to gain insight into participant perceptions regarding the health-effects of consuming unsafe water and water treatment in practices in rural Indian households. The inductive, “bottom-up” approach used for analysis of the responses [
34] helped identify key factors that influence acceptance and sustainability of water quality interventions in rural southern India. This information can be used in the future for better planning and effective implementation of such interventions in low income settings.
Parents of young children linked unsafe water not only with diarrheal disease, but also with other unrelated diseases such as cough, cold and fever. A possible reason for this could be the concept on “hot” and “cold” foods that is widely prevalent in rural Indian families [
35]. Consumption of “hot” and “cold” foods (and beverages) are believed to cause illness and are avoided, especially when a person is ill. In an earlier study in the same area, participants perceived “heating” of the body to be an important cause of diarrhea; very few felt that diarrhea was caused through consumption of unclean water [
23]. This lack of consistency of beliefs is an important impediment to acceptance of such interventions and highlights the need for repeated educational and behavior change interventions in these communities for maximum impact.
Boiling (a relatively expensive method [
36]) cloth (an ineffective method [
13]) were the commonest household water treatment, methods practiced. Boiling was resorted to only for young children and sick adults. In addition some families “warmed” water before giving it to young children, which is also related to “hot” and “cold” food concept mentioned earlier. A few parents felt that water bought while travelling (“outside” water) can cause illness, and placed greater trust on water sourced from their village. Parents from the control village compared their water to “mineral” water (generally meaning that it appeared clean and tasted good), which conferred a false sense of protection. The practice of warming (vis-à-vis boiling) and the false sense of protection from the locally available drinking water may place young children at the greatest risk of diarrheal illnesses as high levels of fecal coliform contamination have consistently been documented in public drinking water samples from urban and rural areas of Vellore region [
18,
21‐
23]. Hence, there is a pressing need for educating local communities, especially parents of young children on making drinking water potable, one of which is boiling and safe storage [
18,
21,
31].
The importance of taste and odor of drinking water has been highlighted in studies from this region and elsewhere [
17,
18,
37]. A few parents reported that the water lacked taste (saltiness) while others described the water as “bitter”. The membrane filter is not known to alter the physical composition of water drastically [
21], and this feedback could represent the participants’ preference for drinking water to have a taste similar to that of untreated water. Any water intervention, hence, needs to be sensitive and address this aspect.
Parents from all study villages generally thought that filtered water was useful and provided health benefits especially to their children. However, a few parents (mostly from the non-compliant families) did not recognize any health benefit and found no need for the filtered water. The control village headman’s statement below captures the need for educating such parents and the public on the risks of consuming unsafe water:
“People may only realize the need for clean water when they experience diseases like diarrhea and vomiting after drinking unclean water.” (Headman, control village)
The ability to recognize health benefits from interventions is crucial to their long term acceptance and sustained use.
Distance of households from the water distribution centers was identified as an important factor affecting compliance to the intervention: non-compliant families tended to reside further away from the water distribution centers than compliant families. Lack of support from the male members to fetch water was also found to negatively influence household compliance, even though the female respondents recognized the benefit of filtered water for their children. Likewise, willingness to pay for future maintenance of the membrane filter was an important decision that mothers would not make without consulting their husbands. The dynamics of decision-making in rural households of India and South Asia is under researched and studies capturing the “process” of decision-making are vital to the success of future community based interventions.
Many parents were willing to pay approximately INR 30 (USD 0.50) every month for the maintenance of the membrane filter. Few parents recognized monetary savings with regards to expenditure towards treatment of illness in children and fuel costs for boiling. Parents who recognized these savings were willing to pay to support future maintenance of the membrane filter. Even though a service fee is currently levied for households with a private connection, (which is same as that of maintaining the membrane filter) the village headmen felt that any model with a service fee will not succeed. Most services in Tamil Nadu state (health care, education, food rationing and water and sewerage) are heavily subsidized or free in the villages [
38]. Therefore, residents of the study villages generally expected clean drinking water from the government at no additional cost to them.
An important limitation of the FGD is the potential for confirmation bias and ‘socially desirable replies’. Although, individuals were encouraged to express alternate viewpoints through probing questions [
39], it is possible that some may only have revealed perceptions that the community believes in. The use of purposive sampling was yet another limitation: only mothers and fathers of children less than two years of age (aged 21 – 40 years) were invited to participate, thus potentially resulting in a selection bias. Inclusion of older participants would have provided a more comprehensive understanding of changing perceptions and needs across different stakeholders. Also, the findings of this study may not be representative of parental perceptions in other rural parts of India. Moreover, parents from non-compliant families were less responsive during the FGDs, and key-informant interviews may have captured their opinions better. Additionally, we could not collect information on other aspects such as storage of water in the household and participant opinions on what water was considered “safe” to drink which, if collected, may have provided a more comprehensive snapshot of existing perceptions.
Conclusions
This study describes gaps in the knowledge of the transmission pathways of diarrheal illnesses as well as changing perceptions over time. This is the first study in the region where the parents of young children were probed on factors likely to influence acceptance and sustainability of a low-cost water quality intervention in a low income setting. Recognizing health benefits, ease of access to water distribution centers and the willingness to pay for intervention maintenance were important factors facilitating acceptance and sustainability. Faulty perceptions of water treatment and a false sense of protection from water locally available not only place young children at the greatest risk of diarrheal diseases but mitigate the demand for water quality interventions. Taste and odor have been highlighted as key factors influencing the acceptance of drinking water in this study and in earlier ones from the region; the value of this in the acceptance and sustainability of future water quality interventions cannot be overstated. A lack of support from male members of the household further impeded long term sustainability of the intervention.
Notwithstanding the identified barriers, the need to effectively involve communities at important stages of implementation is crucial to the long term success of water quality interventions. Timely research on the factors influencing uptake of water quality interventions prior to implementation will ensure greater acceptance and sustainability of such interventions in low income settings.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The author(s) declare that they have no competing interests.
Author contributions
VB, VRM, MRF and GK conceived the intervention study; VB, VRM, RS and MRF designed the qualitative study, MRF conducted/moderated the focus group and key informant interviews and collected the data, GN, VB, MRF, VRM and RS analyzed the data, MRF drafted the manuscript, VB, VRM, GN, GK and RS provided critical feedback on the analyses and interpretation of the results. All authors have approved the final version of this manuscript.