Abstract

This study aimed to identify the main sources of drinking water, evaluate public perception and satisfaction with regards to tap water and compare it with the distance from provider in Florianopolis city, Southern Brazil. Physicochemical and microbiological parameters of water at daycare centers were also assessed. Questionnaires were applied to 1,298 residences in six districts regarding water quality and consumption preferences and analyzed by adjusted binary logistic regression. Tap water physicochemical and microbiological parameters at the selected districts were assessed during summer and winter seasons. Of the 581 questionnaire respondents, 93% did not drink tap water, mostly due to a lack of confidence in water safety and taste. Only 39% were satisfied with water quality and approximately 30% reported that water is worse in summer season. Most water samples collected were not in accordance with Brazilian regulations. Thirty percent of samples were positive for total coliform, and one sample was positive for infectious adenovirus (38 PFU/L). Chlorine concentration was higher in some districts closer to the provider and during the summer season. These results could be useful in informing political resolutions aiming to improve the quality of drinking water, and to protect human health.

Introduction

Management of the quality of drinking water and sanitation is a key factor for preventing and controlling waterborne diseases. In 2010, the United Nations declared access to clean drinking water for domestic and personal use to be a human right, stating that it should be safe, acceptable in taste, odor and color, accessible and affordable (UN, 2010). When such basic conditions are not met, people face serious risks to their safety caused by poor health and disruption to livelihood.

Additionally, among the global sustainable development goals created by the UN in 2015, there is one specifically for water and sanitation: ensure the availability and sustainable management of water and sanitation for all in the next 15 years. Moreover, there are specific targets in other goals that are related to clean water (UN, 2015).

According to the International Water Association, drinking water standards should be based on the protection of human health and consumer acceptability (IWA, 2004). In this respect, public surveys can provide critical information on the perception of and satisfaction with water quality (Doria, 2010). Several studies in many countries have analyzed the public perception of tap water quality and the reasons why bottled water is consumed (Turgeon et al., 2004; Doria, 2006; Jones et al., 2006; Doria et al., 2009). In Brazil, surveys reporting public satisfaction are scarce, and despite the fact that most people seem satisfied with tap water, they buy or filter it to drink. However, these studies were conducted with few participants (up to 60 respondents), resulting in limited information (Da Silva et al., 2010; Freitas et al., 2012; De Queiroz et al., 2013).

Public perception is influenced by several psychological and physiological factors. These factors include personal experiences and external stimulation, as well as organoleptic aspects and health risks associated with water consumption. Contextual elements, such as social demographics, source of water supply, and distribution piping network, may also be associated with water quality and risk perception (Dietrich, 2006; Doria, 2010). Turgeon et al. (2004) reported that general satisfaction and risk perception vary according to distance from provider, source of water supply, and chlorine concentrations. In tourist areas, the population increment during summer seasons is another factor that may affect water quality and perception, mostly in coastal cities, such as Florianopolis, Brazil.

Chlorine is the most common disinfectant used by water treatment plants. It provides pathogen inactivation, and is a preservative and a marker in water distribution systems (Fawell & Nieuwenhuijsen, 2003). Despite its efficiency as a disinfectant, chlorine can affect the taste and odor of water for consumption. Moreover, water contamination may occur during the distribution process, when the chlorine concentration drops to inefficient values. Of all outbreaks related to drinking water between 1971 and 2006 in the United States, 50% were related to flaws detected during the water treatment and distribution process (Craun et al., 2010).

Ligon & Bartram (2016) also identified that most waterborne disease outbreaks are related to deficiencies in water treatment and/or distribution systems. Furthermore, they reported that viruses and bacteria are responsible for 9 and 23% of outbreaks related to drinking water, respectively. However, outbreaks caused by viruses affect more people than other infections.

Diarrhea is the main disease related to contaminated drinking water. 1.7 billion cases per year are reported by the World Health Organization (WHO), of which, 88% are caused by inadequate hygiene, sanitation, and water supply. Children are more vulnerable to infections from drinking contaminated water, mainly due to their immature immune systems. Diarrhea is estimated to cause 1.5 million child deaths per year, mostly among children under five living in developing countries (WHO, 2013).

Thus, the aim of this study was to evaluate public perception and satisfaction with tap water quality in a Brazilian coastal city, and to identify the main source of drinking water (tap, filtered, or bottled). Physicochemical and microbiological parameters of tap water at municipal daycare centers were also measured during winter and summer.

Methods

Area of study

The study was conducted in Florianopolis, the capital city of Santa Catarina State, Brazil (Figure 1). The city has 421,000 inhabitants, in an area of 675 km2 and an average family income of USD$880. Nearly 96% live in urban areas, 84% have access to appropriate sanitation, and 98% of people aged over 15 are literate (IBGE, 2010).

Fig. 1.

Location of Florianopolis city (Island and mainland), Southern Brazil, and sources of water provider and served regions.

Fig. 1.

Location of Florianopolis city (Island and mainland), Southern Brazil, and sources of water provider and served regions.

A municipal water provider is responsible for the uptake, treatment, and distribution of drinking water for the population. There are three different sources of water supply used for this purpose: (a) the Cubatão river, which supplies the Central region of the island and mainland, (b) the Peri lagoon, which supplies the Eastern and Southern regions of the island, and (c) the Ingleses aquifer, which supplies the Northern region of the island (Figure 1). For this study, two districts in each region were selected according to their distance from the water provider (closest to and furthest from the respective water provider), in a total of six districts. The population size (number of inhabitants) of the districts evaluated were: Central-close, 5,883; Central-far, 15,665; Eastern/Southern-close, 5,279; Eastern/Southern-far, 4,925; Northern-close, 10,756; and Northern-far, 3,620 (IBGE, 2010).

Evaluation of perceived drinking water quality

To evaluate public perception of and level of satisfaction with municipal water quality, a questionnaire was developed and distributed in houses located in six districts. The questionnaires were posted in mail boxes or delivered to residents for later collection, or applied in an interview.

Sample size was defined using a formula described by Pocock (1983) for binomial outcomes, assuming 5% for type I error and 80% for power. The prevalence of expected outcomes was unknown, therefore was established as 50%, maximizing the sample size. A difference of at least 15 percentage points among the districts was defined as significant. Considering all these parameters, 84 houses were required in each district, giving a total of 504 residences.

Due to the low response rates using the posting strategy, the sample size was increased to 160%, achieving approximately 220 residences visited in each district. Only houses that appeared to be inhabited and that do not belong to a private condominium were selected. Streets were chosen to cover the largest area of each district and every house in the selected streets received one questionnaire to be answered by a person aged over 18. The data were collected between September and November 2015 (between winter and spring). Informed consent of the participants and the study protocol were approved by the Federal University of Santa Catarina Ethics Committee on Human Research (protocol no. 33243014.1.0000.0121, approved on 8 September 2014).

The questionnaire enquires about the source of drinking water (‘What is the main source of drinking water at home?’; options: tap, filtered, or bottled) and reasons not to drink tap water (‘What are the reasons not to drink tap water at home?’; options: taste, odor, color, safety (none or more than one could be selected)). Additionally, satisfaction with the overall quality, taste, odor, color, safety, healthiness, and cost of water during the last 7 days before the survey and during the summer season were evaluated using a Likert scale ranging from 1 (very bad) to 5 (very good). Sociodemographic information (birth date, number of residents, number of rooms, and educational level) were also collected. Number of rooms was used as a proxy for family income.

Evaluation of drinking water quality

To assess the physicochemical and microbiological parameters of tap drinking water, 12 municipal daycare centers were chosen, two in each district previously selected. It was decided for daycare centers as representative sample points of water distributed into the respective districts and because they are managed by the municipality, therefore have public access. Sampling was carried out in one region per day (four samples) for 3 consecutive weeks, during the summer and winter seasons of 2016. Sampling was done prior to water reaching the tank storage, in order to represent the water distributed to respective districts and to not be influenced by the water tank cleaning conditions. Water temperature, pH, and free chlorine concentration were measured. Ten liters of water were collected and 10 mL of 5% sodium thiosulfate were added to chelate free residual chlorine. In the laboratory, the turbidity was measured and 100 mL of water was used to check the presence of total coliforms and Escherichia coli using an Aquatest kit (Laborclin™).

For viral analyses, 10 liters of water were concentrated with organic flocculation using the skimmed milk technique, as described by Calgua et al. (2008), until 10 mL final volume was reached. Briefly, water sample conductivity was increased by adding 5 g of artificial sea salts (Sigma), pH was adjusted to 3.5 and 10 mL of 1% skimmed milk (Sigma) solution at the same pH was added. After stirring and a rest period of 8 hours each, the bottom 500 mL were centrifuged at 3,800 × g, for 30 minutes at 4 °C (Avanti J30I; Beckman). The pellet was then dissolved once again in phosphate-buffered saline at pH 7.5.

For the detection of human adenovirus (HAdV) infectious particles, samples were evaluated using plaque assay, using non-cytotoxic dilutions in A549 cells. The method was adapted from Cromeans et al. (2008), filtering the diluted sample with a 0.45 μm membrane and using antibiotics (penicillin 100 U·mL−1; streptomycin 100 μg·mL−1) and an antifungal (Amphotericin B 250 μg·mL−1) at 2% each.

Briefly, A549 cell monolayers were grown in six-well tissue culture plates at a density of 5 × 105 cells/well for 24 hours. Diluted and filtered samples were inoculated into cells and after 1 hour of incubation, the samples were removed and cell monolayers were overlaid with 0.6% Bacto-agar (BD) in a 2× concentrated growth medium supplemented with 26 mM MgCl2, antibiotics and antifungal. After 7 days of incubation the agar overlay was removed, cells were stained with 20% Gram's crystal violet solution and plaques were counted macroscopically after stain removal.

After each water collection, one sample was always artificially contaminated with 2.0 × 106 PFU (plaque forming units) of murine norovirus (MNV) to be used as an internal control for the water concentration process. Plaque assay for MNV was performed according to Gonzalez-Hernandez et al. (2012). Briefly, RAW 264.7 cell monolayers were grown in six-well tissue culture plates at a density of 2 × 106 cells/well for 24 hours. Diluted samples were injected into cells and after 1 hour of incubation, the samples were removed and cell monolayers were overlaid with 3% Sea Plate (Lonza) in a 2× concentrated growth medium supplemented with antibiotics and antifungals. After 48 hours of incubation the cells were stained with 0.3% neutral red solution and the plaques were counted.

Data analysis

For the survey, descriptive analysis was carried out and the results were presented as absolute and relative frequencies, with a 95% confidence interval (95% CI) for categorical variables. For continuous variables, mean values and standard deviation (sd) were presented. As for the inferential statistics, simple and multiple binary logistic regressions were performed. Control variables were tested for each exposure–outcome pair, and included into the models using the forward method, those remaining significant (p < 0.05) or that had reduced by 10% or more the coefficient of already included parameters. To this end, the following was determined.

Outcomes: (a) taste, odor, color, and safety as reasons not to drink tap water (yes or no), (b) satisfaction with overall quality, taste, odor, color, safety, healthiness, and cost of water (satisfied (good and very good) or not satisfied (very bad, bad, and neither bad nor good)), and (c) worse taste, odor, color, safety, and healthiness in summer (yes or no).

Exposures: (a) region (Central, Eastern/Southern, and Northern), (b) distance from provider (close and far), and (c) interaction between region-distance (districts).

Control variables: (a) number of rooms in house, (b) educational level, (c) age, (d) water tank cleaning frequency, and (e) drinking water source.

The results were displayed as odds ratio (OR) and 95% CI. If OR > 1.0, the odds of outcome occurrence were higher in the exposed group compared with the reference group; when OR < 1.0, the odds were lower. Central region and distance far from provider were taken as references. This implies that the OR presented between regions is related to the distance far from provider (distance reference). The same applies comparing distances far from and close to the provider: the OR is related to the Central region (region reference). In order to compare the OR of regions setting the close distance as a reference instead of the far distance, the chosen region's OR must be multiplied by the corresponding interaction term (regions × distance).

Using taste attribute as an example to calculate the OR comparing Eastern/Southern against Central regions using close to provider as the reference distance, one has to multiply the corresponding interaction term (2.90) by the respective OR presented in Eastern/Southern region (0.74), resulting in an OR of 2.14. The same process needs to be done to compare the effect of distance from provider within a region other than Central. Using taste attribute again, to calculate the OR comparing close against far distances within the Eastern/Southern region, one has to multiply the corresponding interaction term (2.90) by the respective OR presented in close to provider (0.61), resulting in an OR of 1.77. The correct 95% CI cannot be calculated using this process, so regression models were run changing the reference categories to obtain the proper intervals.

Dependent and independent two-sample Student's t-tests were performed to compare chlorine concentration between the seasons (summer and winter) and the distance from provider (far and close). Levene's test was used to check the homoscedasticity. Sensitivity analysis was conducted using equivalent non-parametric tests (Mann–Whitney and Wilcoxon) and results were similar. All analyses were performed using IBM SPSS 19.0 software, and a p-value <0.05 was considered significant.

Results

Evaluation of perceived drinking water quality

1,298 residences were contacted, of whom, 581 answered the survey, representing a response rate of 45%. Table 1 shows the number of residences participating in each region, their distance from provider and sociodemographic profiles. The mean age of the respondents was 50 (sd = 15) years. The Central region seemed to be the one with the highest economic profile due to greater educational level, number of rooms, and residents per house.

Table 1.

Sociodemographic attributes, reservoir cleaning frequency, and main drinking water source by region, Florianopolis, Brazil, 2015.

Variables Total (n = 581)
 
Central (n = 184)
 
Eastern/Southern (n = 201)
 
Northern (n = 196)
 
n n n n 
Distance to supplier 
 Close 291 50 97 53 94 47 100 51 
 Far 290 50 87 47 107 53 96 49 
Educational level 
 Incomplete ES 37 24 13 
 Complete ES 63 11 10 27 14 26 14 
 Complete HS 184 32 54 30 68 34 62 33 
 Complete Co 287 50 113 62 95 48 79 41 
No. of rooms 
 1 96 17 25 14 41 21 30 16 
 2 205 36 49 27 85 43 71 37 
 3 197 35 69 38 55 28 73 38 
 4 or more 71 13 39 21 15 17 
No. of residents 
 1 45 26 13 12 
 2 160 28 39 22 61 31 60 32 
 3 162 29 56 31 53 27 53 28 
 4 or more 197 35 79 44 55 28 63 34 
Reservoir cleaning frequency per year 
 0 164 30 70 39 62 34 32 18 
 1 288 53 90 51 82 46 116 64 
 2 73 14 15 33 18 25 14 
 3 or more 15 
Main drinking water source 
 Tap 38 16 15 
 Filtered 215 37 81 44 58 29 76 39 
 Bottled 326 56 96 52 125 63 105 54 
Variables Total (n = 581)
 
Central (n = 184)
 
Eastern/Southern (n = 201)
 
Northern (n = 196)
 
n n n n 
Distance to supplier 
 Close 291 50 97 53 94 47 100 51 
 Far 290 50 87 47 107 53 96 49 
Educational level 
 Incomplete ES 37 24 13 
 Complete ES 63 11 10 27 14 26 14 
 Complete HS 184 32 54 30 68 34 62 33 
 Complete Co 287 50 113 62 95 48 79 41 
No. of rooms 
 1 96 17 25 14 41 21 30 16 
 2 205 36 49 27 85 43 71 37 
 3 197 35 69 38 55 28 73 38 
 4 or more 71 13 39 21 15 17 
No. of residents 
 1 45 26 13 12 
 2 160 28 39 22 61 31 60 32 
 3 162 29 56 31 53 27 53 28 
 4 or more 197 35 79 44 55 28 63 34 
Reservoir cleaning frequency per year 
 0 164 30 70 39 62 34 32 18 
 1 288 53 90 51 82 46 116 64 
 2 73 14 15 33 18 25 14 
 3 or more 15 
Main drinking water source 
 Tap 38 16 15 
 Filtered 215 37 81 44 58 29 76 39 
 Bottled 326 56 96 52 125 63 105 54 

ES, Elementary School; HS, High School; Co, College.

Less than 16% clean their water tank twice or more per year, which is the recommended frequency. Most households drink bottled water, and, from those who drink tap water (n = 38), only four reported boiling before consumption. Regarding knowledge about the water treatment used by the provider, 18% reported knowing the basic procedures used by the municipal company.

Safety was the main reason not to drink tap water, reported by 75% of respondents, followed by taste (41%). Odor and color were considered the main reasons by 30% and 32%, respectively. The prevalence of each reason for each region and distance from provider, as well as results from the multiple logistic regressions, are presented in Table 2 (see example on how to interpret results described below in the ‘Data analysis’ section).

Table 2.

Association between reasons for not drinking tap water and region and distance to supplier, Florianopolis, Brazil, 2015.

Variables n (%) Crude logistic regressiona
 
Adjusted logistic regressionb
 
OR 95% CI OR 95% CI 
Region 
 Eastern/Southern 81 (45) 1.33 0.87–2.03 0.74 0.38–1.43 
 Northern 72 (40) 1.10 0.72–1.68 0.58 0.30–1.13 
 Central 67 (38) 1.00  1.00  
Distance to supplier 
 Close 116 (44) 1.24 0.88–1.75 0.61 0.32–1.17 
 Far 104 (39) 1.00  1.00  
 Region × distance      
Eastern/Southern, close    2.90 1.147.38 
 Northern, close    3.01 1.21–7.46 
Region 
Eastern/Southern 46 (26) 0.97 0.60–1.56 0.32 0.150.70 
 Northern 68 (38) 1.73 1.10–2.72 0.99 0.51–1.93 
 Central 46 (26) 1.00  1.00  
Distance to supplier 
 Close 89 (34) 1.42 0.98–2.06 0.53 0.26–1.08 
 Far 71 (26) 1.00  1.00  
Region × distance 
Eastern/Southern, close    6.44 2.2418.51 
Northern, close    2.94 1.147.58 
Region 
Eastern/Southern 49 (27) 0.59 0.38–0.93 0.38 0.200.73 
Northern 56 (31) 0.72 0.47–1.12 0.30 0.150.59 
 Central 68 (39) 1.00  1.00  
Distance to supplier 
Close 86 (33) 1.01 0.70–1.45 0.46 0.240.86 
 Far 87 (32) 1.00  1.00  
Region × distance 
 Eastern/Southern, close    1.95 0.77–4.96 
Northern, close    5.26 2.0913.23 
Region 
 Eastern/Southern 130 (72) 0.79 0.49–1.27 0.67 0.31–1.44 
 Northern 138 (77) 1.02 0.62–1.68 0.58 0.27–1.24 
 Central 135 (77) 1.00  1.00  
Distance to supplier 
 Close 193 (73) 0.77 0.52–1.14 0.49 0.23–1.03 
 Far 210 (78) 1.00  1.00  
Region × distance 
 Eastern/Southern, close    1.13 0.41–3.08 
 Northern, close    3.08 1.09–8.74 
Variables n (%) Crude logistic regressiona
 
Adjusted logistic regressionb
 
OR 95% CI OR 95% CI 
Region 
 Eastern/Southern 81 (45) 1.33 0.87–2.03 0.74 0.38–1.43 
 Northern 72 (40) 1.10 0.72–1.68 0.58 0.30–1.13 
 Central 67 (38) 1.00  1.00  
Distance to supplier 
 Close 116 (44) 1.24 0.88–1.75 0.61 0.32–1.17 
 Far 104 (39) 1.00  1.00  
 Region × distance      
Eastern/Southern, close    2.90 1.147.38 
 Northern, close    3.01 1.21–7.46 
Region 
Eastern/Southern 46 (26) 0.97 0.60–1.56 0.32 0.150.70 
 Northern 68 (38) 1.73 1.10–2.72 0.99 0.51–1.93 
 Central 46 (26) 1.00  1.00  
Distance to supplier 
 Close 89 (34) 1.42 0.98–2.06 0.53 0.26–1.08 
 Far 71 (26) 1.00  1.00  
Region × distance 
Eastern/Southern, close    6.44 2.2418.51 
Northern, close    2.94 1.147.58 
Region 
Eastern/Southern 49 (27) 0.59 0.38–0.93 0.38 0.200.73 
Northern 56 (31) 0.72 0.47–1.12 0.30 0.150.59 
 Central 68 (39) 1.00  1.00  
Distance to supplier 
Close 86 (33) 1.01 0.70–1.45 0.46 0.240.86 
 Far 87 (32) 1.00  1.00  
Region × distance 
 Eastern/Southern, close    1.95 0.77–4.96 
Northern, close    5.26 2.0913.23 
Region 
 Eastern/Southern 130 (72) 0.79 0.49–1.27 0.67 0.31–1.44 
 Northern 138 (77) 1.02 0.62–1.68 0.58 0.27–1.24 
 Central 135 (77) 1.00  1.00  
Distance to supplier 
 Close 193 (73) 0.77 0.52–1.14 0.49 0.23–1.03 
 Far 210 (78) 1.00  1.00  
Region × distance 
 Eastern/Southern, close    1.13 0.41–3.08 
 Northern, close    3.08 1.09–8.74 

Significant interactions are in bold font.

OR, odds ratio; 95% CI, 95% confidence interval.

aCrude model.

bModel including region, distance to supplier, and the interaction term, adjusted by number of rooms in the house, educational level, age, water reservoir cleaning frequency, and drinking water source.

Regarding the water odor, residents from Eastern/Southern and Northern regions close to the provider had, respectively, 2.07 (95% CI: 1.03–4.14) and 2.92 (95% CI: 1.47–5.80) times the odds to report it as a reason not to drink tap water, compared with those living in the Central region. Among houses far from the provider, those from the Eastern/Southern region had 0.32 times the odds to state odor as a reason, compared with the Central (95% CI: 0.15–0.70) and Northern (95% CI: 0.15–0.68) regions. Moreover, in the Eastern/Southern region, those who lived close to the provider had higher odds than those who live far away of refusing to drink tap water due to odor (Table 2).

Among houses far from the provider, those in the Eastern/Southern and Northern regions were less likely to report color as a reason not to drink tap water compared with the Central region. Those from the Central and Northern regions close to the provider had, respectively, 0.46 (95% CI: 0.24–0.86) and 2.39 (95% CI: 1.23–4.64) times the odd to report color as a reason, compared with those who lived far from the provider (Table 2).

Likewise, among houses close to the provider, those in the Eastern/Southern region had 2.14 (95% CI: 1.12–4.11) times the odds to report water taste as a reason not to drink from the tap, when compared with the Central region. As for water safety, no relationship was observed, but it had the highest prevalence in all regions (Table 2).

Only 39% of residences were satisfied with the overall quality of water. Color and odor were the best attributes, with 57 and 52% acceptance rates, respectively. Taste was considered satisfactory by 38%. Safety, healthiness, and cost of water were approved by 21, 25, and 25%, respectively. Figure 2 shows the mean satisfaction score for overall and each attribute quality (except cost), comparing district scores with the survey grand mean of overall quality on the 7 days before the survey. Prevalence of satisfaction by region and distance from provider, as well as logistic regressions, are presented in the Supplementary material (see Table S1, available with the online version of this paper).

Fig. 2.

Mean satisfaction score for overall water quality, taste, odor, color, safety, and healthiness in each district, during the summer season (Δ) and during the survey period (•) (winter and spring), Florianopolis, Brazil, 2015. Dotted lines represent the survey grand mean for overall quality on the 7 days before the survey. Score represents Likert scale: 1 – very bad, 5 – very good.

Fig. 2.

Mean satisfaction score for overall water quality, taste, odor, color, safety, and healthiness in each district, during the summer season (Δ) and during the survey period (•) (winter and spring), Florianopolis, Brazil, 2015. Dotted lines represent the survey grand mean for overall quality on the 7 days before the survey. Score represents Likert scale: 1 – very bad, 5 – very good.

Among houses close to the provider, those in the Northern region had 0.51 (95% CI: 0.27–0.98) times the odds to be satisfied with the water color, when compared with the Central region. Moreover, residents from the Northern region close to the provider were less likely to consider the color good or very good than those who lived far from the provider. Interactions were not found between region and distance from provider for water overall quality, taste, odor, safety, and healthiness.

A high frequency of households reported that water quality is worst in summer. For organoleptic attributes, 36, 33, and 30%, assigned lower scores in the summer season for color, taste, and odor, respectively. Regarding safety and healthiness, 22 and 23% of respondents, respectively, reported that these factors were worse during the summer. Figure 2 shows the mean score for each attribute in all districts during summer, which was always lower.

The prevalence of reports that drinking water is worse in summer by region and distance from provider, as well as logistic regressions, are presented in the Supplementary material (see Table S2, available with the online version of this paper). Among those who lived far from the provider, Northern region residents had 1.91 (95% CI: 1.00–3.65) and 2.23 (95% CI: 1.20–4.15) times the odds to report that the odor and color were worse in summer, respectively, compared with houses from the Eastern/Southern region.

There was an interesting correlation with regards to water odor. In the Northern region, the likelihood of reporting odor as a reason not to drink tap water and of this attribute being worse in summer is higher, and the likelihood of considering odor as satisfactory is lower, when compared with the Central region. Moreover, in general, those who lived close to the provider were less likely to report good or very good odor.

Water quality assessment

Free chlorine, pH, temperature, turbidity, total coliforms, E. coli, and infectious adenovirus were assessed at 12 daycare centers. Values for these parameters on each sampling day from all daycare centers are shown in Table 3 (winter) and Table 4 (summer). The average water temperature was 19.0 °C in winter and 23.3 °C in summer.

Table 3.

Drinking water physicochemical parameters and total coliforms in winter, assessed in municipal daycare centers, Florianopolis, Brazil, August 2016.

Region Distance to supplier Daycare center Residual chlorine (mg·L−1pH Turbidity (Tu) Total coliforms 
Central Close 1.43 2.23 1.35 7.0 6.1 6.2 0.53 2.38 1.55 − − 
2.52 2.01 2.18 6.7 6.6 6.4 1.14 0.66 1.33 − − 
Far 1.67 2.06 2.53 6.6 6.6 7.0 4.64 2.35 0.96 − − 
2.47 2.19 2.77 6.4 6.4 7.1 1.74 1.42 2.24 − − 
Eastern/Southern Close 2.05 2.53 2.85 5.0 5.0 5.2 2.21 3.60 2.48 − 
2.50 2.83 3.15 4.3 4.5 5.1 2.18 4.34 3.09 − 
Far 1.16 0.78 1.05 4.8 5.0 5.3 1.86 2.64 2.67 − 
0.00 0.82 0.91 4.7 4.9 5.1 >10.0 5.66 2.59 − 
Northern Close 0.00 1.20 2.05 6.1 5.3 5.8 7.02 1.25 1.89 − 
10 1.49 1.16 2.82 6.1 5.4 5.9 0.85 1.22 >10.0 − 
Far 11 1.74 1.33 2.35 6.0 5.0 5.8 1.79 >10.0 0.75 − 
12 1.62 1.21 0.80 6.2 5.4 5.8 5.41 1.12 1.20 − 
Legislation limitsa 0.20–5.00 6.0–9.5 Below 5.00 Absence 
Region Distance to supplier Daycare center Residual chlorine (mg·L−1pH Turbidity (Tu) Total coliforms 
Central Close 1.43 2.23 1.35 7.0 6.1 6.2 0.53 2.38 1.55 − − 
2.52 2.01 2.18 6.7 6.6 6.4 1.14 0.66 1.33 − − 
Far 1.67 2.06 2.53 6.6 6.6 7.0 4.64 2.35 0.96 − − 
2.47 2.19 2.77 6.4 6.4 7.1 1.74 1.42 2.24 − − 
Eastern/Southern Close 2.05 2.53 2.85 5.0 5.0 5.2 2.21 3.60 2.48 − 
2.50 2.83 3.15 4.3 4.5 5.1 2.18 4.34 3.09 − 
Far 1.16 0.78 1.05 4.8 5.0 5.3 1.86 2.64 2.67 − 
0.00 0.82 0.91 4.7 4.9 5.1 >10.0 5.66 2.59 − 
Northern Close 0.00 1.20 2.05 6.1 5.3 5.8 7.02 1.25 1.89 − 
10 1.49 1.16 2.82 6.1 5.4 5.9 0.85 1.22 >10.0 − 
Far 11 1.74 1.33 2.35 6.0 5.0 5.8 1.79 >10.0 0.75 − 
12 1.62 1.21 0.80 6.2 5.4 5.8 5.41 1.12 1.20 − 
Legislation limitsa 0.20–5.00 6.0–9.5 Below 5.00 Absence 

Sampling was made for 3 consecutive weeks (column values), one region per day.

Tu: Turbidity unit; (−): below method detection limit.

aAnnex XX of the Consolidation Ordinance no. 5 of the Ministry of Health.

Table 4.

Drinking water physicochemical parameters and total coliforms in summer, assessed in municipal daycare centers, Florianopolis, Brazil, March 2016.

Region Distance to supplier Daycare center Residual Chlorine (mg·L−1pH Turbidity (Tu) Total coliforms 
Central Close 3.60 3.60 4.00 5.0 6.7 6.7 0.25 1.19 0.98 − − − 
2.52 4.00 3.60 5.0 5.9 6.0 1.01 0.55 0.36 − − − 
Far 2.99 1.31 4.00 5.0 6.4 6.7 2.16 1.61 1.02 − − − 
3.90 1.99 2.90 5.0 6.6 6.5 1.98 1.74 1.32 − − − 
Eastern/Southern Close 1.37 2.04 1.69 5.2 4.9 5.2 1.26 2.34 2.45 − − − 
2.84 2.48 2.56 5.6 5.1 5.5 1.13 4.22 2.56 − − − 
Far 0.06 0.00 0.08 5.1 5.4 4.9 1.64 1.73 2.40 − − 
0.00 0.00 0.49 5.2 5.0 5.0 2.13 1.66 2.16 − − 
North Close 1.58 1.65 1.10 5.5 6.4 5.4 1.79 2.38 1.37 − − − 
10 1.35 1.87 1.49 5.2 6.3 5.4 1.35 0.99 >10.0 − − 
Far 11 NS 1.57 1.07 NS 6.4 5.2 NS 1.57 1.39 NS − − 
12 1.67 1.36 1.55 5.5 6.0 5.8 1.13 1.68 1.06 − − − 
Legislation limitsa 0.20–5.00 6.0–9.5 Below 5.00 Absence 
Region Distance to supplier Daycare center Residual Chlorine (mg·L−1pH Turbidity (Tu) Total coliforms 
Central Close 3.60 3.60 4.00 5.0 6.7 6.7 0.25 1.19 0.98 − − − 
2.52 4.00 3.60 5.0 5.9 6.0 1.01 0.55 0.36 − − − 
Far 2.99 1.31 4.00 5.0 6.4 6.7 2.16 1.61 1.02 − − − 
3.90 1.99 2.90 5.0 6.6 6.5 1.98 1.74 1.32 − − − 
Eastern/Southern Close 1.37 2.04 1.69 5.2 4.9 5.2 1.26 2.34 2.45 − − − 
2.84 2.48 2.56 5.6 5.1 5.5 1.13 4.22 2.56 − − − 
Far 0.06 0.00 0.08 5.1 5.4 4.9 1.64 1.73 2.40 − − 
0.00 0.00 0.49 5.2 5.0 5.0 2.13 1.66 2.16 − − 
North Close 1.58 1.65 1.10 5.5 6.4 5.4 1.79 2.38 1.37 − − − 
10 1.35 1.87 1.49 5.2 6.3 5.4 1.35 0.99 >10.0 − − 
Far 11 NS 1.57 1.07 NS 6.4 5.2 NS 1.57 1.39 NS − − 
12 1.67 1.36 1.55 5.5 6.0 5.8 1.13 1.68 1.06 − − − 
Legislation limitsa 0.20–5.00 6.0–9.5 Below 5.00 Absence 

Sampling was made for 3 consecutive weeks (column values), one region per day.

Tu, Turbidity unit; NS, Not Sampled; (−), below method detection limit.

aAnnex XX of the Consolidation Ordinance no. 5 of the Ministry of Health.

From all 71 samples, 63% had physicochemical parameters that are not accepted by Brazilian regulations. Water pH was below 6.0 in 62% of samples and in the Eastern/Southern region all of them were inappropriate, with the lowest pH registered. Six samples in winter and one in summer were above the limit for turbidity.

As for microbiological parameters, 55% of samples in winter and 8% in summer were positive for total coliforms, although E. coli was not detected in any of them. For infectious human adenovirus (HAdV), only one sample was positive (38 PFU/L) in summer at daycare center 7, in which chlorine was absent (Table 4). During the winter, a very strong rainfall occurred 2 days before starting the second sampling in Eastern/Southern daycare centers. Since then, all samples were positive for total coliforms. No significant rainfall was observed in the summer sampling.

Residual chlorine concentration was higher during summer in the Central region, but only significant when close to the provider (Figure 3). In the Eastern/Southern region, the concentration was lower when far from the provider, which also had the most samples with chlorine below that required by legislation (Tables 3 and 4 and Figure 3). In the Northern region, residual chlorine was consistent during summer and winter.

Fig. 3.

Residual chlorine concentration measured in municipal daycare centers during winter and summer representing each district, Florianopolis, Brazil, 2016. *p < 0.01.

Fig. 3.

Residual chlorine concentration measured in municipal daycare centers during winter and summer representing each district, Florianopolis, Brazil, 2016. *p < 0.01.

Discussion

In general, people in Florianopolis city were dissatisfied with the tap water quality, mostly due to taste and lack of confidence in water safety. Thus, the large majority did not drink tap water, choosing to filter it or buy bottled water. Most water samples at daycare centers were not in accordance with local legislation and presented some differences in chlorine concentration between the winter and summer seasons.

Population surveys can provide valuable information on water quality and population satisfaction. Furthermore, they can be useful for understanding and decision-making by the provider company, seeking to improve the service (Doria, 2010). Due to the population's ability to detect problems in the water, it is proposed that the distribution provider considers using consumers as monitors of water quality (Whelton et al., 2004; Dietrich, 2006).

Other studies also pointed out that taste is the main sensitive factor for not drinking tap water or for choosing bottled water (Levallois et al., 1999; Turgeon et al., 2004; Dietrich, 2006; Merkel et al., 2012). However, these studies reported that safety is not the main reason for not drinking tap water, which was reported by 75% of respondents in the present work. Doria et al. (2009) showed that in the United Kingdom and Portugal the participants assigned scores below average for safety, despite the organoleptic factors being considered good.

Florianopolis city is highly visited by tourists, reaching twice its resident population during the summer. This can lead to contamination of recreational and drinking water as the sewage treatment system may not adequately manage the increment, and illegal wastewater disposal often occurs (Moresco et al., 2012). The demand generated by this significant surplus during summer may be related to the worsening of water quality. In order to supply drinking water for all, the treatment process may be jeopardized, resulting in declining satisfaction from residents in the summer. Moreover, drinking water distribution is frequently interrupted during this season, increasing the necessity for water tanks at residences.

In Brazil, it is common to have one or more water tanks per house. Usually, the volume capacity of each tank varies from 500 to 2,000 L. For this reason, the tank cleaning frequency could interfere with perception of water quality. Nearly 30% of residences had not cleaned their tank during the year prior to the survey, allowing the accumulation of solids, which can modify the taste and color of water. According to Dietrich (2006), changes in public perception are generally caused by chemical and microbiological agents in water catchment, chemicals added or removed during treatment, and contamination during distribution.

The inconsistency in the distributed water quality may be a reason for people not considering it satisfactory, since consumers can detect variations in pH, minerals, and organic matter (Dietrich, 2006). During the present study, seasonal and spatial variations in pH, chlorine, turbidity, and pathogens were detected in the water. However, it is complex to identify the relationship between the physicochemical and microbiological parameters and the public perception of water, due to other factors not measured in this study, such as the mineral, chemical, and microbiological composition of the three water sources, and the quality of and repairs to the distribution piping for each region.

Only 25% of the participants were satisfied with the water service cost. This could be related to the high prevalence of people unsatisfied with the drinking water quality. As seen in Table 1, most people have a proper filtration system or buy bottled water for consumption. This certainly induces an unbalancing at the ‘cost/benefit’ of the service since they have to invest more to consume safe and palatable water instead of use the treated tap water provided by the company.

Brazilian regulations on drinking water are established by the Annex XX of the Consolidation Ordinance no. 5 of the Ministry of Health. It allows free chlorine concentration between 0.2 and 5 mg·L−1, turbidity of up to 5 turbidity units (Tu) and pH between 6.0 and 9.5 in the distribution system. Regarding microbiological parameters, absence of E. coli is mandatory. However, there is only one recommendation for surveying enteric viruses in the water supply source, even though it is known that viruses are more resistant than bacteria to chlorine treatment (Lechevallier & Au, 2004; Brazil Ministry of Health, 2017).

Chlorine concentrations in the Eastern/Southern region far from the provider were the lowest recorded and absent in summer, which may be related to residents being less likely to report odor as a reason for not drinking tap water, compared with those who live close to the provider. Moreover, odor was mostly considered satisfactory for those who live far from the provider, which is related to chlorine concentrations that are lower than when close to the provider.

In the Northern region, differences in free chlorine concentration were not observed, regardless of distance from provider. This may be due to a shorter distribution network, with it only being 10 km to the farthest district, while in the Eastern/Southern and Central regions it is 19 and 25 km, respectively. Furthermore, the higher temperature during summer may enhance chlorine evaporation. This could explain the difference between concentrations when close to and far from the provider, mostly in regions that have a more extensive distribution network.

Regarding the microbiological parameters, it is known that high chlorine concentration and low pH promote hypochlorous acid formation, which has a higher power and efficiency in water disinfection (Cheremisinoff, 2002; Page et al., 2010). This could explain why only one sample was positive for infectious human adenovirus. Since bacteria are less resistant than adenovirus to chlorine inactivation, the larger number of positive samples may be due to higher contamination in the network system.

It was not possible to correlate gastroenteritis cases with water consumption. However, of the 97 people that reported gastroenteritis symptoms only 48 sought medical aid, representing a 50% underreporting rate (see Supplementary material, Table S3, available with the online version of this paper). Ligon & Bartram (2016) showed that, in general, developed countries are those with most outbreak reports. This may be related to greater control and identification of diseases than in developing countries, not necessarily to a higher number of cases.

In Brazil, the underreporting of infectious diseases is common and when reported, they are often stated with delays, making it difficult to identify the source and to make decisions. Furthermore, the etiologic agent of gastroenteritis is often not specified, and can simply be categorized as intestinal infections (Grisotti, 2010). Cesa et al. (2016) studied the relation between water and wastewater treatment and waterborne diseases from 2002 to 2009 in Florianopolis city. In this case, people reported that tap water had an unreliable quality.

It should be taken into account that it is not possible to apply the results from this study to the whole city, which would require a different and larger sampling method. However, the response rate was 45% and probably most subjects were highly motived to participate in this study. Sample size was calculated expecting a response rate of 40%, so it is powered for the analyses conducted. Also, residence responses were represented by one person on behalf of the family. Some differences could be found if every person in the house were interviewed. Memory and availability biases might be present as participants were asked to indicate satisfaction about drinking water quality in summer. Nevertheless, this strategy allowed us to compare satisfaction between seasons. Confirmation bias could also be presented since the respondents might be inclined to validate their beliefs about quality of drinking water.

Analysis of additional drinking water samples may show more accurate results, making it possible to establish better correlation between water quality and perception. Furthermore, more tests to assess the quality of water attributes could be conducted. However, it is important to highlight that, to our knowledge, this is the largest study in Brazil on the public perception of water, relating it to the quality of water.

Regulatory impact assessment (RIA) could be a valuable tool for decision-making policies, since it can provide a more rational decision on regulations, based on evidence (Carvalho et al., 2017). A survey done in low-income countries showed that RIA has been applied but it is still in the early stage of adoption (Kirkpatrick et al., 2004). In Brazil, RIA is incipient and faces some problems to proper implementation, such as the lack of a centralized regulatory agency (Castro, 2014). Also, the lack of studies and information on this topic hampers the RIA application. Our work combined with future studies on drinking water quality and satisfaction can offer reliable information for RIA and further regulation reviewing.

Conclusions

In Florianopolis, most people are dissatisfied with the quality of water, mostly because of safety and taste. Dissatisfaction is justifiable considering the seasonal and spatial variation in pH, free chlorine, and microbial presence. In this study, a correlation was identified between the distance from provider and the perception of odor, probably due to chlorine concentration. More studies on public perception must be conducted, preferably with a greater sample size.

Nearly 60% of all tap water samples analyzed were not in accordance with Brazilian legislation. Thus, stricter enforcement of standards for drinking water is recommended. In addition, it is suggested that enteric viruses are included as a mandatory standard for microbial contamination in Brazilian regulations, as they are present in drinking water samples and are not related to total coliforms.

Usually, there is a deficiency in cooperation among water provider companies, municipalities, and consumers in Brazil. The data raised in the present work can be useful to notify the Florianopolis population about the quality of water supplied. Additionally, it can support water provider companies and policy-makers to look for solutions to improve the quality of drinking water. Thus, the surveillance on water quality and public satisfaction can be enhanced and routinely performed. RIA also should be considered for more studies and decision-making policies.

Acknowledgments

This work was supported by The Brazilian National Council for Scientific and Technological Development (CNPq) [grant numbers 420398/2016-3 and 400183/2014-5] and there was no involvement in the study. We thank Florianopolis City Hall for allowing us to collect water at municipal daycare centers, Camila Daminelli Schissi and Joceli Zaguini for the support during the study, and Heather Louise Godwig for reviewing the English language. The authors declare they have no actual or potential competing financial interests.

References

References
Brazil Ministry of Health
2017
.
Consolidation Ordinance no. 5 of the Ministry of Health, of September 28 2017. Provides on the Procedures of Control and Monitoring of the Quality of Water for Human Consumption and its Standard of Potability (in Brazilian Portuguese)
.
Federal Official Gazette
,
Brasilia
,
October 3, 2017
, p.
365
.
Calgua
B.
,
Mengewein
A.
,
Grunert
A.
,
Bofill-Mas
S.
,
Clemente-Casares
P.
,
Hundesa
A.
,
Wyn-Jones
A. P.
,
López-Pila
J. M.
&
Girones
R.
,
2008
.
Development and application of a one-step low cost procedure to concentrate viruses from seawater samples
.
Journal of Virological Methods
153
(
2
),
79
83
.
http://dx.doi.org/10.1016/j.jviromet.2008.08.003
.
Carvalho
B. E.
,
Marques
R. C.
&
Netto
O. C.
,
2017
.
Delphi technique as a consultation method in regulatory impact assessment (RIA) – the Portuguese water sector
.
Water Policy
19
(
3
),
423
439
.
http://dx.doi.org/10.2166/wp.2017.131
.
Castro
C. M.
,
2014
.
Some aspects of implementing Regulatory Impact Analysis in Brazil
.
Revista de Administração Pública
48
(
2
),
323
342
.
http://dx.doi.org/10.1590/0034-76121359
.
Cesa
M.
,
Fongaro
G.
&
Barardi
C. R. M.
,
2016
.
Waterborne diseases classification and relationship with social-environmental factors in Florianópolis city – Southern Brazil
.
Journal of Water and Health
14
(
2
),
340
348
.
http://dx.doi.org/10.2166/wh.2015.266
.
Cheremisinoff
N. P.
,
2002
.
Handbook of Water and Wastewater Treatment Technologies
.
Butterworth Heinemann
,
Woburn
, p.
636
.
Craun
G. F.
,
Brunkard
J. M.
,
Yoder
J. S.
,
Roberts
V. A.
,
Carpenter
J.
,
Wade
T.
,
Calderon
R. L.
,
Roberts
J. M.
,
Beach
M. J.
&
Roy
S. L.
,
2010
.
Causes of outbreaks associated with drinking water in the United States from 1971 to 2006
.
Clinical Microbiology Reviews
23
(
3
),
507
528
.
http://dx.doi.org/10.1128/CMR.00077-09
.
Cromeans
T. L.
,
Lu
X.
,
Erdman
D. D.
,
Humphrey
C. D.
&
Hill
V. R.
,
2008
.
Development of plaque assays for adenoviruses 40 and 41
.
Journal of Virological Methods
151
(
1
),
140
145
.
http://dx.doi.org/10.1016/j.jviromet.2008.03.007
.
Da Silva
S. R.
,
Heller
L.
,
De Campos
V. J.
&
Cairncross
S.
,
2010
.
Relationship (or its lack) between population and a water and sanitation service: a study of users’ perception in Vitória (ES) Brazil
.
Journal of Water and Health
8
(
4
),
764
778
.
http://dx.doi.org/10.2166/wh.2010.067
.
De Queiroz
J. T.
,
Doria
M. F.
,
Rosenberg
M. W.
,
Heller
L.
&
Zhouri
A.
,
2013
.
Perceptions of bottled water consumers in three Brazilian Municipalities
.
Journal of Water and Health
11
(
3
),
520
531
.
http://dx.doi.org/10.2166/wh.2013.222
.
Dietrich
A. M.
,
2006
.
Aesthetic issues for drinking water
.
Journal of Water and Health
4
(
Suppl. 1
),
11
16
.
http://dx.doi.org/10.2166/wh.2005.034
.
Doria
M. F.
,
2006
.
Bottled water versus tap water: understanding consumer's preferences
.
Journal of Water and Health
4
(
2
),
271
276
.
http://dx.doi.org/10.2166/wh.2006.008
.
Doria
M. F.
,
2010
.
Factors influencing public perception of drinking water quality
.
Water Policy
12
(
1
),
1
19
.
http://dx.doi.org/10.2166/wp.2009.051
.
Doria
M. F.
,
Pidgeon
N.
&
Hunter
P. R.
,
2009
.
Perceptions of drinking water quality and risk and its effect on behaviour: a cross-national study
.
Science of the Total Environmental
407
(
21
),
5455
5464
.
http://dx.doi.org/10.1016/j.scitotenv.2009.06.03
.
Fawell
J.
&
Nieuwenhuijsen
M. J.
,
2003
.
Contaminants in drinking water: environmental pollution and health
.
British Medical Bulletin
68
(
1
),
199
208
.
http://dx.doi.org/10.1093/bmb/ldg027
.
Freitas
I. S.
,
Figueredo
B. L.
,
Almeida
J. V.
,
Vieira
J. F.
,
Barbosa
M. M. C.
&
Pinto
F. R.
,
2012
.
Social Perception on the Quality of Public Water Supply in the Community of Planalto Renascer, Quixadá/Ce
.
Annals of VII North/Northeast Congress of Research and Innovation
.
5
pp.
(in Brazilian Portuguese). Available at: http://propi.ifto.edu.br/ocs/index.php/connepi/vii/schedConf/presentations. Accessed April 17 2017
.
Gonzalez-Hernandez
M. B.
,
Cunha
J. B.
&
Wobus
C. E.
,
2012
.
Plaque assay for murine norovirus
.
Journal of Visualized Experiments
66
,
e4297
.
http://dx.doi.org/10.3791/4297
.
Grisotti
M.
,
2010
.
Emerging infectious diseases and disease emergencies: a conceptual review and new issues
.
Science & Collective Health
(in Brazilian Portuguese)
15
(
Suppl. 1
),
1095
1104
.
http://dx.doi.org/10.1590/S1413-81232010000700017
.
IBGE – Brazilian Institute of Geography and Statistics
2010
.
Population Census of 2010 (in Brazilian Portuguese)
.
Available at: http://cod.ibge.gov.br/7MP. Accessed April 18 2017
.
IWA (International Water Association)
2004
.
The Bonn Charter for Safe Drinking Water
.
IWA
,
London
, p.
20
.
Available at: http://www.who.int/wsportal/bonn/en/. Accessed April 16 2017
.
Jones
A. Q.
,
Dewey
C. E.
,
Doré
K.
,
Majowicz
S. E.
,
McEwen
S. A.
,
David
W.
,
Eric
M.
,
Carr
D.
&
Henson
S. J.
,
2006
.
Public perceptions of drinking water: a postal survey of residents with private water supplies
.
BMC Public Health
6
(
94
),
11
.
http://dx.doi.org/10.1186/1471-2458-6-94
.
Kirkpatrick
C.
,
Parker
D.
&
Zhang
Y.
,
2004
.
Regulatory impact assessment in developing and transition economies: a survey of current practice
.
Public Money & Management
24
(
5
),
291
296
.
http://dx.doi.org/10.1111/j.1467-9302.2004.00436
.
Lechevallier
M. W.
&
Au
K
, .
2004
.
Water Treatment and Pathogen Control: Process Efficiency in Achieving Safe Drinking-Water
.
WHO/IWA Publishing
,
London
, p.
136
.
Levallois
P.
,
Grondin
J.
&
Gingras
S.
,
1999
.
Evaluation of consumer attitudes on taste and tap water alternatives in Quebec
.
Water Science
40
(
6
),
135
139
.
http://dx.doi.org/10.1016/S0273-1223(99)00549-1
.
Ligon
G.
&
Bartram
J.
,
2016
.
Literature review of associations among attributes of reported drinking water disease outbreaks
.
International Journal of Environmental Research and Public Health
13
(
6
),
527
544
.
http://dx.doi.org/10.3390/ijerph13060527
.
Merkel
L.
,
Bicking
C.
&
Sekhar
D.
,
2012
.
Parents’ perceptions of water safety and quality
.
Journal of Community Health
37
(
1
),
195
201
.
http://dx.doi.org/10.1007/s10900-011-9436-9
.
Moresco
V.
,
Viancelli
A.
,
Nascimento
M. A.
,
Souza
D. S. M.
,
Ramos
A. P. D.
,
Garcia
L. A. T.
,
Simões
C. M. O.
&
Barardi
C. R. M.
,
2012
.
Microbiological and physicochemical analysis of the coastal waters of southern Brazil
.
Marine Pollution Bulletin
64
(
1
),
40
48
.
http://dx.doi.org/10.1016/j.marpolbul.2011.10.026
.
Page
M. A.
,
Shisler
J. L.
&
Marinãs
B. J.
,
2010
.
Mechanistic aspects of adenovirus serotype 2 inactivation with free chlorine
.
Applied and Environmental Microbiology
76
(
9
),
2946
2954
.
http://dx.doi.org/10.1128/AEM.02267-09
.
Pocock
S. T.
,
1983
.
Clinical Trials: A Practical Approach
,
1st edn
.
John Wiley & Sons
,
Chichester
, p.
265
.
Turgeon
S.
,
Rodriguez
M. J.
,
Thériault
M.
&
Levallois
P.
,
2004
.
Perception of drinking water in the Quebec City region (Canada): the influence of water quality and consumer location in the distribution system
.
Journal of Environmental Management
70
(
4
),
363
373
.
http://dx.doi.org/10.1016/j.jenvman.2003.12.014
.
UN (United Nations)
2010
.
Resolution 64/292 from 28 July 2010. The human right to water and sanitation. General Assembly, p. 3
.
UN (United Nations)
2015
.
Sustainable Development Goals
. .
Whelton
A. J.
,
Dietrich
A. M.
,
Burlingame
G. A.
&
Cooney
M. F.
,
2004
.
Detecting contaminated drinking water: harnessing consumer complaints
. In:
AWWA Water Quality Technology Conference Proceedings
,
San Antonio
, p.
11
.
WHO (World Health Organization)
2013
.
Diarrhoeal Disease
.
Fact sheet no. 330 from April 2013. Available at: http://www.who.int/mediacentre/factsheets/fs330/en/. Accessed April 17 2017
.

Supplementary data