3.1. Parameters identified in the monitored hospital effluents
Table 2 presents the descriptive statistics for the Surfactants parameter, being analyzed the substances that react with methylene blue, LAS (ABNT NBR 10738:1989, SMWW 5540:2017, ASTM D4251-89:2016, ISO 7875-1:1996/COR 1:2003) (ABNT 1989; ISO 2003; D12 Committee 2016; Baird and Bridgewater 2018), monitored in the analyzed hospitals. Since the data do not follow a normal distribution, the median is the best estimator of the central position of the data, and the absolute deviation from the median (MAD) as the data dispersion estimator. The number of establishments that presented effluent measurements was increasing over the years. In 2007 only one establishment was monitoring it in the effluents. In 2018, a total of 72 establishments were performing the same monitoring. For the year 2019, there are data only for January, which explains why there are only 31 establishments for the year.
Table 2
Descriptive statistics of surfactant measurements (mg⋅L− 1) over the years
Years
|
Health care facilities
|
Samples
|
Mean
|
SD
|
Median
|
MAD
|
Min
|
Max
|
IQR
|
2007
|
1
|
2
|
1.50
|
0.83
|
1.50
|
0.87
|
0.918
|
2.09
|
0.58
|
2008
|
6
|
18
|
1.63
|
0.93
|
1.54
|
0.93
|
0.24
|
3.26
|
1.16
|
2009
|
20
|
118
|
3.05
|
9.4
|
1.38
|
0.97
|
0.08
|
80.21
|
1.28
|
2010
|
29
|
140
|
2.02
|
2.6
|
1.28
|
1.2
|
0.0
|
17.80
|
1.79
|
2011
|
50
|
258
|
7.42
|
65
|
1.14
|
0.93
|
0.0
|
829.38
|
1.50
|
2012
|
54
|
292
|
2.04
|
2.1
|
1.50
|
1.5
|
0.0
|
18.10
|
2.31
|
2013
|
54
|
301
|
2.07
|
3.6
|
1.18
|
1.4
|
0.0
|
45.40
|
2.11
|
2014
|
56
|
307
|
1.99
|
3.7
|
1.10
|
1.1
|
0.0
|
44.00
|
2.05
|
2015
|
63
|
335
|
1.68
|
2.8
|
0.997
|
1.1
|
0.0
|
31.00
|
1.51
|
2016
|
66
|
342
|
1.81
|
2.7
|
0.938
|
1.2
|
0.0002
|
22.40
|
1.97
|
2017
|
71
|
380
|
2.71
|
6.0
|
0.738
|
0.94
|
0.0
|
49.20
|
1.81
|
2018
|
72
|
378
|
1.52
|
2.7
|
0.500
|
0.59
|
0.0
|
22.75
|
1.36
|
2019
|
31
|
31
|
1.05
|
1.4
|
0.680
|
0.76
|
0.0
|
6.96
|
0.96
|
All
|
78
|
2852
|
2.49
|
19.8
|
1.00
|
1.1
|
0
|
829.38
|
1.84
|
Legend: SD: standard deviation; MAD: median absolute deviation; Min: minimum value; Max: maximum value; IQR: interquartile range. |
As observed in Table 2, the median value for the monitored hospitals between 2007 and 2019 was 1 mg.L− 1 while the mean value was 2.49 mg.L− 1. In a research hospital in Istanbul, Turkey, with 780 beds, anionic surfactants were also identified in the raw hospital effluent (Top et al. 2020), ranging from 0.26 to 2.08 mg⋅L− 1 and with a mean value of 1.96 mg⋅L− 1. This shows that the values found during the monitored period in hospitals in Belo Horizonte have similarities in some years, being below those presented by this author. Studies of hospital effluents in France showed lower values, and cationic and non-ionic surfactants were also analyzed. In the study conducted by the SIPIBEL project, France, near the Swiss border, anionic surfactants were analyzed in the dissolved fraction of samples collected from hospital and municipal effluents for two years. The data obtained showed the highest presence of anionic surfactants in urban effluents, with values similar to those found in Brazilian hospital effluents: 2 mg⋅L− 1; and in hospital effluents, the median obtained was 0.40 m⋅L− 1. In hospital effluents, cationic and non-ionic surfactants predominated (Wiest et al. 2018), which are not yet required by Brazilian legislation.
Selecting variables to be used in the multiple linear regression model, twenty regression models were initially created using random sampling of 50% of the data. For each model, separated sampling was done. Of the 27 independent variables used in the models, only the variables Chromium VI, Chromium, COD, Phenols, and Sulfides were significant in all models created in this step and were selected as variables for the final regression model. The final regression model was created using 67% of the randomly sampled data. The remaining 33% were used for model validation. Table 3 summarizes the coefficients of the final regression model.
Table 3
Coefficients of the final regression model
|
Estimate
|
Std. Dev.
|
Error
|
p value*
|
(Intercept)
|
0.790
|
0.12
|
6.8
|
1.08x10− 11
|
Chrome VI
|
10,1
|
1.4
|
7.4
|
1.47x10− 13
|
Chromium
|
-6.16
|
1.4
|
-4.4
|
1.33x10− 5
|
COD
|
0.00251
|
0.00022
|
12
|
< 2.00x10− 16
|
Phenols
|
-0.635
|
0.22
|
-2.9
|
0.00399
|
Sulfide
|
0.991
|
0.13
|
7.5
|
7.18x10− 14
|
Legend: Std. Dve.: standard deviation; *: significative values when p < 0.05 |
The regression model had an adjusted R² of 0.1446. The standard error of the model residuals was 2.879, which corresponds to a coefficient of variation of 1.48. For the validation data, the R² was 0.1418, and the standard error of the residuals was 2.786, which are similar values to the values obtained for the regression model, which showed the validity of the model.
In the final regression model, positive correlations were found for the variables Chromium VI, COD, and Sulfides, as presented in Table 2, suggesting a high value of these parameters associated with high value of surfactants. Considering negative correlations, Chromium and Phenols, the lower these parameters, the higher the value of the surfactants. The presence of surfactants in effluents could lead to an increase in COD values, and the positive correlation presented in Table 2 confirmed this hypothesis. In presence of surfactants the phenol could be solubilized in the micelles, and cations (M2+ and M3+) promoted the capture of these micelles, corroborating the negative correlation. These cations are present in the hospital wastewaters (Sultana, et al., 2021; Cavalcante et al, 2018). Sulphites can be formed in the degradation of bacteria.
3.2. Surfactants concentrations identified in the literature
The effluents generated in hospital laundries present different characteristics from the others, due to the high volume of water and great complexity, such as the presence of different classes of surfactants and detergents, besides the high microbial and viral load, pharmaceuticals and common parameters in all effluents (Kern et al. 2015; Zotesso et al. 2017; Lutterbeck et al. 2020). Table 4 presented the surfactant concentrations of studies conducted in countries such as France, Switzerland, and Turkey. In the survey conducted by Lutterbeck et al. (2020), treatment processes and toxicity of influents and effluents from the laundry of a 180-bed hospital, which generates 150 m³ of effluent daily, were evaluated. Of this volume, 33% is produced in the hospital laundry. The mass balance of the laundry products used in each step of the process was performed, obtaining the following values for consumption of 6.3 m3 of water: nonylphenol ethoxylate: 43.7 µg⋅L− 1; sodium hydroxide: 220 µg/L; hydrogen peroxide: 190 µg/L; quaternary de ammonium compounds: 63.5 µg⋅L− 1; and sodium bisulfite: 31 µg⋅L− 1. The acute ecotoxicity tests, with D. magna, revealed, for the raw effluent, extreme toxicity (EC50 24.5%). After undergoing effluent treatments, a 100% EC50 was obtained, indicating non-toxicity (Lutterbeck et al. 2020). In 2015, Kern and collaborators also analyzed the laundry effluent from the same hospital and identified the following surfactants: lauryl ethoxylate, benzoic acid, n-(4-aminophenyl)-n-methylacetamide, 2-(diethylamino)-n-(2,6-dimethylphenyl)-acetamide), and methyl-m-(1-methylbutyl) phenyl ester, by gas chromatography.
In the study of Lasek et al. (2019), effluents from a French university hospital, whose water consumption corresponds to 482 L/bed, were analyzed through the collection of samples carried out in campaigns between the years 2016 and 2017. The authors observed the concentrations of three substances that have detergent-disinfectant actions, the CHD, DDAC, and BAPLA, in 3 isolated collection points, as well as at the point of confluence between all effluents. The concentrations in mg⋅L− 1 observed for the sampling campaigns are expressed in Table 4. In the research by Kovalova et al. (2012) the authors analyzed the effluent from a Swiss hospital to evaluate the removal of micropollutants and disinfectants.
Table 4
Concentrations of surfactants detected in hospital effluents
Surfactants
|
Place sampled
|
Concentration (mg⋅L− 1)
|
Source
|
DDAC
|
Oncology, Laundry, cardiology, geriatrics, pulmonology, main pharmacy
|
1.52–3.25
|
(Lasek et al. 2019)
|
The main building of the hospital: medical clinic, operating rooms, endoscopy, and intensive care units
|
0.93–2.33
|
CHD
|
Oncology, Laundry, cardiology, geriatrics, pulmonology, main pharmacy
|
0.027–0.056
|
The main building of the hospital: medical clinic, operating rooms, endoscopy, and intensive care units
|
0.031–0.097
|
BAPLA
|
Oncology, Laundry, cardiology, geriatrics, pulmonology, the main pharmacy
|
0.031–0.097
|
The main building of the hospital: medical clinic, operating rooms, endoscopy, and intensive care units
|
0.018–0.093
|
Anionic
|
Median concentrations for the sampling campaigns conducted from 2013 to 2015
|
0.4
|
(Wiest et al. 2018)
|
Cationic
|
1.1
|
Nonionic
|
4.9
|
Anionic
|
Raw hospital wastewater
|
4.55 x 10− 4
|
(Buelow et al. 2020)
|
Treated hospital effluent
|
3.57 x 10 − 5
|
Cationic
|
Raw hospital wastewater
|
5.51 x 10 − 4
|
Treated hospital effluent
|
1.28 x 10 − 4
|
Nonionic
|
Raw hospital wastewater
|
6.27 x 10 − 3
|
Treated hospital effluent
|
3.155 x 10 − 5
|
Anionic
|
Average sample in 24h
|
< 0.01
|
(Boillot et al. 2008)
|
Cationic
|
5.0
|
Glutaraldehyde
|
0.0021
|
Nonionic
|
2.9
|
Anionic, upstream and downstream of hospital effluent discharge from February to May 2011
|
Arve River
|
< 0.10
|
(Perrodin et al. 2013)
|
Hospital Effluent
|
< 0.05
|
Cationic, upstream and downstream of hospital effluent discharge from February to May 2011
|
Arve River
Hospital Effluent
|
< 0.4
|
Arve River
Hospital Effluent
|
< 0.4 e 1
|
Nonionic, upstream and downstream of hospital effluent discharge from February to May 2011
|
Arve River
Hospital Effluent
|
< 2
|
Arve River
Hospital Effluent
|
16.8 e 17.9
|
Nonionic
|
Raw Hospital Effluent
|
4.24 a 19.95
|
(Perrodin et al. 2016)
|
Effluent after HWW
|
< 0.01
|
Anionic
|
Raw Hospital Effluent
|
< 0.05 a 0.24
|
Effluent after HWW
|
< 0.05
|
Cationic
|
Raw Hospital Effluent
|
< 0.4
|
Effluent after HWW
|
< 0.4
|
QAC BAC C12-18
|
-
|
0.049
|
(Kovalova et al. 2012)
|
BAC C12
|
-
|
0.034
|
DDAC-C10
|
-
|
0.102
|
Anionic
|
Average
|
1.96
|
(Top et al. 2020)
|
Through the statistical analyses developed, correlations were observed between the surfactants and other physicochemical parameters found in the monitored effluents of small, medium-sized, and large healthcare establishments (Table 5), in which the highest correlations observed originated in the effluents from the large establishments. The correlation of surfactants with oils was observed in all healthcare establishment sizes, except for medium-sized. The oils may originate from the effluents from the kitchen of the healthcare establishment's facilities. The correlation of surfactants with this parameter may be linked to their solubility in lipids, characterized by the hydrophobic part present in the structure of these organic compounds (Nakama 2017).
Table 5
Correlations observed between anionic surfactants with other parameters monitored in small to medium-sized healthcare establishments effluents
Grouping/parameter
|
Aluminum
|
Ammonium
|
Boron
|
BOD
|
COD
|
Phenols
|
Nickel
|
Sulfide
|
Oils
|
All
|
-
|
-
|
-
|
|
-
|
-
|
-
|
-
|
***
|
Large
|
**
|
***
|
*
|
-
|
***
|
***
|
***
|
***
|
***
|
Medium
|
-
|
-
|
-
|
-
|
-
|
-
|
-
|
-
|
-
|
Small
|
-
|
-
|
-
|
**
|
**
|
-
|
-
|
-
|
***
|
Legend: BOD: biological oxygen demand; COD: chemical oxygen demand; ‘***’: 0.0001 (p-value); ‘**’: 0.001 (p-value); ‘*’: 0.01 (p-value). |
The current Brazilian legislation that dictates the conditions and standards for effluent discharge includes only the limits for anionic surfactants. The "Deliberação Normativa Conjunta" COPAM/CERH-MG nº 01, of May 5th, 2008 (Minas Gerais 2008), legislation in force in the State of Minas Gerais, Brazil, establishes the conditions and standards for effluent discharge into surface waters. For the organic parameters, which include surfactants that react with methylene blue, the limit set for discharge, except for public treatment systems, is 2.0 mg⋅L-1 of LAS, (Minas Gerais 2008). Regarding the quality standards for surface water, both in the state and federal legislation, CONAMA 357/2005, the maximum value to be maintained is 0.5 mg⋅L-1 of LAS. The "Companhia de Saneamento do Estado de Minas Gerais - COPASA”, defines the limits for the discharge of non-domestic effluents into the public collecting system, through Technical Standard T.187/6, in which the expected concentration allowed for surfactant substances is 5 mg⋅L-1 (COPASA 2018).
Cationic and nonionic surfactants do not yet have specific legislation for effluent discharge into sewage, consequently into surface waters (Brasil 2005; de Sousa et al. 2012; Jesus et al. 2013; Lasek et al. 2019). They have been detected in influent and effluent as well as in sludge samples from wastewater treatment plants. The WWTPs partially remove the QACs, and thus these compounds reach the environment. The action of microorganisms presents in the water body and natural photodegradation is slow, leading to an increase of these compounds in both waters and sediments (Pati and Arnold 2020). The attenuation mechanisms of these compounds in aquatic environments are slow photolysis and biodegradation, in addition to sorption on suspended materials, which can sediment (EPA 2006), causing persistence. They are also found adsorbed on sludge from aerobic and anaerobic processes. A total of 167 mg⋅kg-1 of total QACs was found in sewage sludge in Guangzhou, China (Li et al. 2014). The sludge from 52 WWTPs scattered across China was analyzed, seeking to identify the presence of adsorbed QACs. In the analyses of alkyl trimethyl ammonium, benzyl alkyl dimethyl ethyl ammonium, and dialkyl dimethyl ammonium homologues by LC MS/MS, the concentration range data found were 0.38-294 𝜇g⋅g-1, 0.94-191 𝜇g⋅g-1, and 0.64-343 𝜇g⋅g-1, respectively (Ruan et al. 2014). Chlorhexidine digluconate (CHD) showed toxicity in aquatic organisms, the lowest toxicity being to algae Pseudokirchneriella subcapitata (41.3 𝜇g⋅L-1, EC50, 24 h) and to Vibrio fischeri 3675.2 𝜇g⋅L-1, EC50 after 5 min assay (Jesus et al. 2013).
3.3. Available analytical methodologies to identify and quantify different surfactants
The analytical methodologies for the determination of surfactants in environmental samples have grown over the years in the world, but their implementation in the control of their release into the environment is still slow. This can be attributed to the process and the investments in equipment and standards for certification. The great diversity of compounds and metabolites can contribute to the complexity of this problem.
For environmental samples, anionic surfactants have well-established methods, such as the spectrophotometric method, which determines the active substances that react with methylene blue, the compounds derived from LAS, which was used to determine surfactants in the samples of this study. For the determination of nonionic surfactants in environmental samples, the cobalt thiocyanate reacting substances method (SMWW 5540:2018), and the method using Dragendorff reagent (ISO 7875-2:1984) (ISO 1984) have been used. Regarding cationic surfactants, instrumental analytical methodologies have been used, such as HPLC, LC MS, LC MS/MS, with higher costs involved in the acquisition of equipment and supplies. Moreover, analytical standards are required in addition to the metabolites formed, which are not always available for purchase. It also involves the pre-treatment of liquid and solid samples, from the collection and preservation until the solid phase extraction, its recovery, and the validation of the methodology (Paijens et al. 2020).
3.4. Physicochemical characteristic of healthcare establishments effluents and their contribution to the discharge of surfactants in the sewage system and related impacts
The effluents from the laundry of the Regional University Hospital of Maringá, in the state of Paraná (Brazil) showed 130 mg⋅L− 1 and 240 mg⋅L− 1 of BOD and COD, respectively. Products are used in the different stages of the process, being basically cationic surfactants (Furtado et al. 2020). In that study, the effects of cytotoxicity were investigated with Allium cepa L., with the effluents from each step of the process, and after treatment by a physicochemical process, and this one, associated with UV/H2O2 processes. That effluent showed to be cytotoxic by inhibiting the cell division of Allium cepa L. (Furtado et al. 2020). When investigating the efficiency of hospital effluent treatment using UV photocatalysis and ozonation, Souza et al. (2018) were able to remove surfactants from the effluent generated by an 840-bed hospital in Porto Alegre, Brazil. The initial concentration of anionic surfactants was at 4.96 mg⋅L− 1, reduced to 4.25 mg⋅L− 1 after the coagulation process, and after the application of O3/UV treatment, they found 0.218 mg⋅L− 1 and a reduction of total organic carbon by 54.7% (Souza et al. 2018).
Surfactants and their degradation products have been found in several concentrations in effluents, treated sludge, surface water, solids, sediments, among others, offering risks to the environment (Ying 2006; Olkowska et al. 2014). In Brazil, detergents can still contain phosphorus in their composition, a nutrient that contributes to the eutrophication phenomenon. Von Sperling (2014), explains that up to 50% of the phosphorus found in domestic sewage may originate from detergents (Von Sperling 2014). According to Panouillères and collaborators (2007), detergents and disinfectants are among the main substances that cause toxicity to aquatic species. The toxic effects to the biotic community have been discussed in the literature and differ according to the organism and the concentration to which it is exposed.
Wang et al. (2015) studied the effects caused by anionic, cationic, and nonionic surfactants on embryos and larval stage of zebrafish. The authors identified results as delayed embryo development and changes in larval locomotion activities. Lechuga et al. (2016) reported the toxicity of surfactants on Daphnia magna, luminescent bacteria, Vibrio fischeri, and freshwater and saltwater microalgae. The results show that Vibrio fischeri exhibited the highest sensitivity to the surfactants tested, although the other organisms were also affected to a lesser extent. Boillot and Perrodin (2008) showed the combination of glutaraldehyde with surfactants (anionic, cationic, and nonionic), both used in hospital routines, and their toxic effects related to the mobility of Daphnia Magna. Other authors have studied the toxic effects of surfactants and combinations with these compounds, to Daphnia Magna (Emmanuel et al. 2005; Hodges et al. 2006; Lechuga et al. 2016). Nałęcz-Jawecki et al. (2003) analyzed the toxicity of QACs on Vibrio fischeri, ciliated protozoa, and Artemia franciscana. The authors concluded that these compounds showed toxicity to all organisms tested.
Finally, as presented by Lutterbeck and colleagues (2020a), hospital laundry wastewater can pose negative impacts to human and environmental health, and this topic is rarely discussed in the scientific literature, when discussing the treatment of these effluents. Also, treatment alternatives did not show a complete solution to the toxic effects of this effluent. In this context, concepts as cleaner production and zero waste are important to reduce the quantity of surfactants in the wastewater and, consequently, reducing the toxicity of effluents generated by healthcare facilities. Changing the process with the same hygiene security procedure represents an opportunity in this context.