- Email: [email protected]
Microbial quality of drinking water from microfiltered water dispensers
G Model IJHEH-12698; No. of Pages 5
ARTICLE IN PRESS International Journal of Hygiene and Environmental Health xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
International Journal of Hygiene and Environmental Health journal homepage: www.elsevier.com/locate/ijheh
Microbial quality of drinking water from microfiltered water dispensers R. Sacchetti a,∗ , G. De Luca a , A. Dormi a , E. Guberti b , F. Zanetti a a b
Department of Medicine and Public Health, Division of Hygiene, University of Bologna, Via S. Giacomo, 12, 40126 Bologna, Italy Department of Public Health, UO Food Hygiene and Nutrition, Local Health Unit of Bologna, Via A. Gramsci 12, 40121 Bologna, Italy
a r t i c l e
i n f o
Article history: Received 2 July 2012 Received in revised form 2 May 2013 Accepted 4 June 2013 Keywords: Water dispenser Drinking water Heterotrophic plate count P. aeruginosa
a b s t r a c t A comparison was made between the microbial quality of drinking water obtained from Microfiltered Water Dispensers (MWDs) and that of municipal tap water. A total of 233 water samples were analyzed. Escherichia coli (EC), enterococci (ENT), total coliforms (TC), Staphylococcus aureus, Pseudomonas aeruginosa and heterotrophic plate count (HPC) at 22 ◦ C and 37 ◦ C were enumerated. In addition, information was collected about the principal structural and functional characteristics of each MWD in order to study the various factors that might influence the microbial quality of the water. EC and ENT were not detected in any of the samples. TC were never detected in the tap water but were found in 5 samples taken from 5 different MWDs. S. aureus was found in a single sample of microfiltered water. P. aeruginosa was found more frequently and at higher concentrations in the samples collected from MWDs. The mean HPCs at 22 ◦ C and 37 ◦ C were significantly higher in microfiltered water samples compared to those of the tap water. In conclusion, the use of MWDs may increase the number of bacteria originally present in tap water. It is therefore important to monitor the quality of the dispensed water over time, especially if it is destined for vulnerable users. © 2013 Elsevier GmbH. All rights reserved.
Introduction In Italy the quality of municipal tap water has reached a fairly high level (Anonymous, 2012a). Despite this, problems of an organoleptic nature or regarding the maintenance of the water supply have led many consumers to resort to bottled water and Italy in fact holds the record in Europe for the highest consumption of mineral water (Anonymous, 2012b). However, the elevated costs and the excessive amount of energy needed to produce bottled water have recently led to a general reassessment of tap water (Aqua Italia, 2012). In this context, the use of devices that treat the drinking water at the point of use is becoming increasingly more widespread. Such devices are marketed as being able to eliminate unpleasant odors and tastes and to remove any undesirable substances from the tap water. They often include systems for the addition of CO2 and for the cooling of the water. Compared to bottled water these devices offer the advantage of avoiding the need for the transport, storage and disposal of the bottles.
∗ Corresponding author. Tel.: +39 51 2094800; fax: +39 51 2094829. E-mail address: [email protected] (R. Sacchetti).
Numerous types of such devices are commercially available (Aqua Italia, 2012). One of the most commonly used in Italy is the microfiltered water dispenser (MWD). MWDs are devices directly attached to municipal drinking water supplies in private residences, offices, restaurants and hospitals. The treatment of the water by means of composite filters is carried out immediately before the water is dispensed. There is very little data in the literature about the quality of the water dispensed from MWDs. Baumgartner and Grand (2006), Charberny et al. (2006), Lèvesque et al. (1994), and Liguori et al. (2010) analyzed the quality of drinking water dispensed from other types of water dispensers (water coolers or soda fountains). In general, the water dispensed from these devices was found to be more contaminated than the water supplied to them. In previous studies (Sacchetti et al., 2009; Zanetti et al., 2009) we carried out laboratory tests on certain prototypes of MWDs to compare the ability of two disinfectants to ensure an adequate bacteriological quality of the dispensed water. In the present study we investigated the microbial quality of drinking water from MWDs used in collective restoration environments in an area of Northern Italy. In particular, we compared the bacteriological characteristics of microfiltered drinking water with that of municipal tap water and we studied the various factors that might influence the quality of the water dispensed by MWDs.
1438-4639/$ – see front matter © 2013 Elsevier GmbH. All rights reserved. http://dx.doi.org/10.1016/j.ijheh.2013.06.002
Please cite this article in press as: Sacchetti, R., et al., Microbial quality of drinking water from microfiltered water dispensers. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.06.002
G Model IJHEH-12698; No. of Pages 5
ARTICLE IN PRESS R. Sacchetti et al. / International Journal of Hygiene and Environmental Health xxx (2013) xxx–xxx
2
Materials and methods Microfiltered water dispensers The study focused on MWDs used in hospital and school canteens (n = 36) and a sample of MWDs in use in bars/restaurants (n = 34) in the area of Bologna, Northern Italy. The bars/restaurants were randomly selected from the list of collective restoration establishments in the territory. For each establishment selected for the study a phone call was made to verify that a MWD was in use. To be included in the study, the water dispensers had to be directly attached to municipal drinking water distribution systems, and had to use composite filters (of the EVERPURE type) for the treatment. The composite filters consist of a disposable cartridge containing a microfiltering membrane (0.5 micron pore size), made of polyethylene fibers, and powdered activated carbon. Some of them also contain a bacteriostatic element (copper or silver salts). For each MWD studied, information about the age of the device, the presence of a UV lamp, the presence of a bacteriostatic element in the filter, the amount of water consumed, the times of use/non use, the frequency and method of disinfection, the frequency of filter change, the interval between the last filter change and sampling and the last disinfection and sampling was collected by means of an ad hoc questionnaire. On the basis of this information, the continuous use of the device throughout the week was defined “daily use” and the compliance with the directions supplied by the manufacturer regarding the means and frequency of disinfection (at least 2 times/year or 4 times/year if the device was installed in a hospital or school) was defined “adequate disinfection”. The main structural and functional characteristics of the MWDs included in the study are given in Table 1. This information was not available in three devices. Water sampling From October 2010 to May 2011, 70 MWDs were examined. Samples were taken of all the various types of drinking water available (still unchilled and chilled water, or still chilled and carbonated Table 1 Structural and functional characteristics of the microfiltered water dispensers. Age in months (mean ± S.D.) Bacteriostatic element (%) Yes No UV lamp (%) Yes No Daily use (%) Yes No Liters of water consumed per day (mean ± S.D.) Frequency of filter change per year (%) 1 ≥2 Frequency of disinfection per year (%) ≤2 >2 Type of disinfectant (%) Chlorine Hydrogen peroxide Quaternary ammonium salts Adequate disinfectiona (%) Yes No Time since last filter change (days; mean ± S.D.) Time since last disinfection (days; mean ± S.D.)
44.2 (±34.9) 61.1 38.9 76.7 23.3 47.1 52.9 74.5 (±142.9) 63.2 36.8 52.6 47.4 10.0 44.0 46.0 55.0 45.0 160.0 (±155.5) 123.2 (±146.6)
a Adequate disinfection = compliance with the directions supplied by the manufacturer regarding type and frequency of disinfection.
chilled water, or still unchilled, still chilled and carbonated chilled water) in order to assess any differences in contamination and to highlight any critcal areas. At the same time a sample of the tap water entering the dispenser was also collected. The samples were always taken in the morning, after the devices had been working for about an hour. When different types of drinking water were collected from a single device, the same sampling sequence was always repeated. A total of 233 water samples were analyzed, including 70 samples of municipal tap water and 163 samples of water from MWDs (49 of still unchilled water, 63 of still chilled water, 51 of carbonated chilled water). To ensure that the samples were representative of the water consumed, flushing was not performed before sampling and the outer surfaces of the nozzles were not sterilized (Lèvesque et al., 1994; Liguori et al., 2010). Samples were collected in sterile 1 L plastic bottles containing 1 mL of sterile sodium thiosulphate solution (10%). They were kept at 4 ◦ C and analyzed within 2 h in our laboratory. Bacteriological analysis In accordance with Italian regulations for drinking water (D.Lgs 31/2001, application of EC directive 98/83), the following bacteriological parameters were quantified for each sample: Escherichia coli (EC), enterococci (ENT), indicator microorganisms of the quality of water (total coliforms – TC, heterotrophic plate count – HPC at 22 ◦ C), and supplementary microorganisms (Pseudomonas aeruginosa – PA and Staphylococcus aureus – SA). HPC at 37 ◦ C was also determined to obtain a more complete assessment of the bacteriological quality of the water. In Italy the measurement of the HPC at 37 ◦ C is required only for water sold in bottles or containers. The microbiological criteria for unbottled municipal drinking water are: absence in 100 mL per EC and ENT; absence in 100 mL per CT. The Italian regulations set no numerical value for HPC at 22 ◦ C but state that there should be no “abnormal change” compared to the values obtained during routine official checks. Also for PA no limit is set. PA is a supplementary parameter to be determined at the discretion of the local health authority. SA is a supplementary parameter, but its absence in 250 mL of water is required. All analyses followed the techniques proposed in the Standard Methods (APHA, 2005). HPCs at 37 ◦ C and 22 ◦ C were determined by the pour plate method using Plate Count Agar- Standard Methods Agar (Oxoid). The mean value of three replicates was calculated. The other microbial parameters were quantified by membrane filtration (0.45 micron pore-size sterile membrane, Millipore) in 100 mL of water (for EC, ENT, TC) and 250 mL (for PA and SA). The detection limit was 1 cfu per sample volume for all types of bacteria. EC: The filter was transferred to C-EC agar (Oxoid). After incubation at 44.5 ◦ C for 24 h, typical colonies (fluorescent green-blue under a Wood lamp and positive to indole test) were counted. Doubtful colonies underwent biochemical identification using the Enterotube II system (BBL). ENT: The filter was transferred to m-Enterococcus agar (Oxoid). After incubation at 35 ◦ C for 24–48 h, typical colonies (pink-brown in color and 0.3–2 mm in diameter) were confirmed by growth on Bile esculine agar (Oxoid) at 35 ◦ C for 48 h and by growth on Brainheart infusion broth (Oxoid) with 6.5% NaCl at 35 ◦ C for 48 h. TC: The filter was transferred to C-EC agar (Oxoid). After incubation at 37 ◦ C for 24 h, typical colonies (green-blue) were counted. Doubtful colonies underwent biochemical identification using the Enterotube II system (BBL). PA: The filter was placed on Pseudomonas CFC agar (Oxoid) and incubated at 30 ◦ C for 48 h. Colonies that were smooth, mucoid,
Please cite this article in press as: Sacchetti, R., et al., Microbial quality of drinking water from microfiltered water dispensers. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.06.002
G Model IJHEH-12698; No. of Pages 5
ARTICLE IN PRESS R. Sacchetti et al. / International Journal of Hygiene and Environmental Health xxx (2013) xxx–xxx
3
Table 2 Number (%) of positive samples and mean values of microbial counts in relation of water type. Bacteriological parameters
Municipal tap water (N = 70)
Still unchilled water (N = 49)
Still chilled water (N = 63)
Carbonated chilled water (N = 51)
Positive samples
Positive samples
Positive samples
Positive samples
No. (%) EC (Log10 cfu/100 mL) ENT (Log10 cfu/100 mL) TC (Log10 cfu/100 mL) SA (Log10 cfu/250 mL) PA (Log10 cfu/250 mL)
0 0 0 0 3 (4.3)
Mean
No. (%)
0.30
0 0 2 (4.2) 0 11 (22.9)
Mean
No. (%)
0.24
0 0 1 (1.6) 0 9 (14.3)
2.04
Mean
No. (%)
Mean
0.30
0 0 2 (3.6) 1 (2.0) 3 (5.9)
0.78 0.30 0.62
1.43
EC: Escherichia coli; ENT: enterococci; TC: total coliforms; SA: Staphylococcus aureus; PA: Pseudomonas aeruginosa.
fluorescent, blue-green or yellow-green in color, with diffuse pigmentation of the medium were presumed to be P. aeruginosa and were subsequently subcultured on TSA (Difco) and identified by the API 20NE system (bioMerieux). Atypical colonies were also counted and at least 5 colonies per plate, or all if less than 5, were subcultured on TSA (Difco) and identified by the API 20NE system (bioMerieux). SA: The filter membrane was incubated on Staph 110 medium (Oxoid) at 36 ◦ C for 40–48 h. To identify suspected colonies (dark orange in color) the API Staph System (BioMérieux) was used. Temperature and residual chlorine at the moment of sampling were also measured, with a mercury-filled Celsius thermometer and DPD colorimetric method (N,N-diethyl-p-phenylenediamine) respectively (APHA, 2005). Statistical analysis The values of microbial concentrations were converted into Log10 colony-forming units (Log10 cfu). The t test (paired) was applied to compare the logarithms of HPCs. Multivariate analysis was used to assess the interaction between the various factors that might influence the HPCs in the dispensed water. The following parameters were considered: HPC in the input water, temperature, residual chlorine, age of device, presence of bacteriostatic element in the filter, presence of UV lamp, liters of water dispensed every day, daily use, frequency of disinfection, adequate disinfection, type of disinfectant used, frequency of filter change, time since last filter change and time since last disinfection treatment. The percentage of variation in the heterotrophic plate counts of the input and output water (%) was calculated using the following formula: (log10 B − log10 A)/log10 A × 100 where A = HPC in input water and B = HPC in output water. The HPC values were expressed in Log10 (x + 1). The association between the % and the various structural and functional characteristics of the devices was calculated using simple correlation for continuous variables, Mann Whitney/Wilcoxon test for dichotomous variables and Kruskal–Wallis test for variables with more than two categories. The significance level chosen for all analyses was p < 0.05. Analyses were performed using the Statistical Package for Social Sciences software (SPSS, version 17). Results The results of the microbial analyses performed on samples of water from the MWDs and tap water are shown in Table 2. EC and ENT were not detected in any of the water samples. TC were never detected in the tap water, but were found in 5 samples of water taken from 5 different MWDs, in concentrations ranging from 0.30 to 1.56 Log10 cfu/100 mL. SA was recovered in only 1 water sample from an MWD.
Contamination by PA at very low levels (0–0.60 Log10 cfu/250 mL) was detected in 3 tap water samples (4.3%); in 16 MWDs (22.9%) at least one sample of dispensed water was found to be contaminated by PA. The highest levels of PA were detected in still unchilled and still chilled water, where the values reached respectively 3.26 Log10 cfu/250 mL and 3.68 Log10 cfu/250 mL. The differences between the mean HPC values in the various types of microfiltered water vs those in the tap water and the relative statistical significance are shown in Fig. 1. The mean HPCs at 22 ◦ C and 37 ◦ C were significantly higher in microfiltered water samples compared to those of the input water. The highest HPC values were observed in the carbonated water, where they reached 3.70 Log10 cfu/mL at 37 ◦ C and 3.46 Log10 cfu/mL at 22 ◦ C. The mean HPCs at 22 ◦ C and 37 ◦ C were significantly higher in carbonated water compared to chilled and unchilled still water. The mean temperature was 19.5 ◦ C for tap water, 23.5 ◦ C for still unchilled water, 9.8 ◦ C for still chilled water and 9.7 ◦ C for carbonated chilled water. The mean residual chlorine content was 0.18 mg/L for tap water, 0.10 mg/L for still unchilled water and 0.08 for still chilled water. In the multivariate analysis no parameter was found to be significant vs HPC in the dispensed water either at 22 or 37 ◦ C. The % at 22 ◦ C was seen to be directly and significantly (p < 0.01) correlated with the age of the device in the still unchilled water, the still chilled water and in the carbonated chilled water. When the disinfection was adequate the % at 22 ◦ C in the still unchilled water was lower and statistically significant (p < 0.01) compared to when the disinfection was not adequate. A negative trend was observed between the increase in HPC values and the residual chlorine present in the water dispensed. Discussion The aim of this study was to assess the quality of the water dispensed from the MWDs used for collective restoration in our territory. The exact number of MWDs used throughout the area is unknown, since the local health authority is often not notified of the installation of the devices. Although destined for human consumption, this type of water is rarely monitored and very little is known about its bacteriological quality. No specific microbial criteria exist but the water dispensed must nevertheless meet the required legal standards for drinking water. The results of our study revealed no contamination from EC and ENT in any of the water samples examined. In accordance with Italian norms, the absence of these microorganisms, considered to be indicators of fecal pollution, confirms the safety of the water with regards enteropathogenic bacterial agents. TC were detected in only 5 samples of microfiltered water. The low levels detected and the absence of TC in the input water would suggest that the presence of these microorganisms in the microfiltered water derives from a contamination of the distribution points of the devices. This
Please cite this article in press as: Sacchetti, R., et al., Microbial quality of drinking water from microfiltered water dispensers. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.06.002
G Model IJHEH-12698; No. of Pages 5 4
ARTICLE IN PRESS R. Sacchetti et al. / International Journal of Hygiene and Environmental Health xxx (2013) xxx–xxx
Fig. 1. Mean HPC values in the various types of microfiltered water vs tap water and relative statistical significance.
can easily occur when the nozzles are protruding and the appliances are used by a large number of people. TC are among the parameters considered to be indicators of water quality; given that some species of TC are of confirmed environmental origin, the health risk associated with the occasional presence of these microorganisms in the water is assessed, case by case, by the local health authority. While the absence of EC and ENT must always be rigorously respected, the occasional failure to conform to the norms set for TC can be tolerated, at least in the first instance. As far as the supplementary parameters are concerned, the presence of SA found in a single sample of microfiltered water can equally be explained by human contamination of the distribution point. The greater frequency and higher concentrations of PA detected in the dispensed water as opposed to the input water, confirm the
microorganism’s ability to colonize the circuits of MWDs (Sacchetti et al., 2009). It is well known that PA naturally present in the environment may be the source of disease in vulnerable subpopulations (the elderly or the very young, immunocompromised patients) (Aumeran et al., 2007; Trautmann et al., 2001, 2005). Moreover, the presence of high numbers of PA in drinking water may be associated with complaints about taste, odor and turbidity (WHO, 2011). We found that the HPCs in the microfiltered water were statistically higher than those in the tap water. However, since no scientific evidence exists to suggest that, in the absence of fecal contamination, HPC values are directly related to a health risk in a healthy population (WHO, 2011), no limit has been established for HPCs in current Italian regulations. Nevertheless, an increase in the level of HPCs is an indication of bacterial regrowth (Bonadonna
Please cite this article in press as: Sacchetti, R., et al., Microbial quality of drinking water from microfiltered water dispensers. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.06.002
G Model IJHEH-12698; No. of Pages 5
ARTICLE IN PRESS R. Sacchetti et al. / International Journal of Hygiene and Environmental Health xxx (2013) xxx–xxx
et al., 2008) and can lead to a deterioration in the esthetic quality of the water (e.g. taste, odor and discoloration) (Chowdhury, 2011). It cannot be ruled out, though, that high HPC values may represent a risk for the vulnerable users, since HPCs include many opportunistic pathogenic bacteria in water habitats (Achromobacter, Pseudomonas, Moraxella, Stenotrophomonas) (Allen et al., 2004; Pavlov et al., 2004). The highest HPC values were observed in the carbonated water. This could be explained by the presence of a small carbonization reservoir in the MWDs, where the water tends to stagnate during periods of non use. In addition, as highlighted in the information collected for the study, this reservoir is often not disinfected so as to avoid technical problems and organoleptic alterations to the water. In our study the HPC values were used to make a statistical analysis of the influence of specific characteristics of the MWDs on the quality of the dispensed water. In the multivariate analysis no parameter was found to be significant vs HPC in the dispensed water. This may be due in part to the small sample size and the considerable variability of the characteristics considered. The direct statistically significant correlation between the % at 22 ◦ C and the age of the device, can probably be explained by the progressive formation of biofilm in the water circuits, providing a favorable habitat for psychrophilic bacterial flora. The gradual release of planktonic bacteria from the biofilm leads to an increase in the HPC values at 22 ◦ C in the dispensed water. The water lines of MDWs are made of a plastic material and are a few millimeters in diameter, while their length may reach several meters. It is well known that in devices where water flows through a network of narrow bore tubing made of plastic material, bacteria can form multispecies adherent biofilm on the inside of the waterlines (Sacchetti et al., 2006, 2007; Szymanska, 2005; Walker and Marsh, 2007). A negative trend was observed between the increase in HPC values and the residual chlorine present in the water dispensed. As expected, the reduction in residual chlorine observed in the microfiltered water compared to the tap water is a factor that favors bacterial regrowth. Also periods of stagnation are well documented as favoring bacterial regrowth; however, in our study, given the extreme variability of the conditions of use/non use of the devices, it was not possible to observe any type of association. Finally, the increase in the HPC levels at 22 ◦ C was lower when the disinfection conformed to the manufacturer’s directions given in the operating manual of the device. This confirms the importance of an adequate disinfection of the devices. The measurement of HPC levels could represent the most simple means of monitoring the bacterial regrowth in the water circuits and the possible presence of opportunistic pathogenic bacteria. Acceptable HPC levels could be those set by the municipal drinking water regulations and the guidelines of other countries (e.g. 100 cfu/mL in Germany and the Netherlands or 500 cfu/mL in the USA and Canada) (Chowdhury, 2011), or the limits set by the Italian norms for municipal drinking water on sale in bottles or containers (100 cfu/mL at 22 ◦ C and 20 cfu/mL at 37 ◦ C). Moreover, the measurement of HPC may be useful to assess the efficacy of the disinfection procedures. However, specific testing for the presence of PA may also be advisable, especially in situations at risk.
5
In conclusion, the use of MWDs may increase the number of bacteria originally present in tap water. It is therefore important to monitor the quality of the dispensed water over time, especially if the MWDs are quite old and/or installed in establishments for vulnerable users. Acknowledgments The authors would like to thank Mara Veronesi and Morena Bertelli, Mauro Daolio, Giorgia Mistroni, Giampaolo Murtas, Dimitri Zuffa for their valuable assistance in collecting samples. They also thank Giovanni Lorusso for the technical assistance in culturing water samples. References Aqua Italia, 2012. Consumo di acqua potabile presso la popolazione italiana, www.aquaitalia.it Allen, M.J., Edberg, C.S., Reasoner, D.J., 2004. Heterotrophic plate count bacteriaWhat is their significance in drinking water? Int. J. Food Microbiol. 92, 265–274. Anonymous, 2012a. Sai cosa bere? Altroconsumo 261, 10–15. Anonymous, 2012b. Annuario Bevitalia 2011–12. Beverfood ed. APHA, 2005. Standard Methods for the Examination of Water and Wastewater, 21st ed. APHA-AWWA-WEF, Washington, DC. Aumeran, C., Paillard, C., Robin, F., Kanold, J., Baud, O., Bonnet, R., Souweine, B., Traore, O., 2007. Pseudomonas aeruginosa and Pseudomonas putida outbreak associated with contaminated water outlets in an oncohaematology paediatric unit. J. Hosp. Infect. 65, 47–53. Baumgartner, A., Grand, M., 2006. Bacteriological quality of drinking water from dispensers (coolers) and possibile control measures. J. Food Prot. 69, 3043–3046. Bonadonna, L., Memoli, G., Chiaretti, G., 2008. Formazione Di Biofilm Sul Materiale A Contatto Con Acqua: Aspetti Sanitari E Tecnologici Rappoti ISTISAN 08/19. Charberny, I.F., Kaiser, P., Sonntag, H.G., 2006. Can soda fountains be recommended in hospitals? Int. J. Hyg. Environ. Health 209, 471–475. Chowdhury, S., 2011. Heterotrophic bacteria in drinking water distribution system: a review. Environ. Monit. Assess. 184 (10), 6087–6137. Lgs, D., 31/2001. Attuazione della Direttiva 98/83/CE relativa alle acque destinate al consumo umano. G.U. della Repubblica Italiana. Supplemento n◦ 52 del 3 marzo. Lèvesque, B., Simard, P., Gauvin, D., Gingras, S., Dewailly, E., Letarte, R., 1994. Comparison of the microbiological quality of water coolers and that of municipal water systems. Appl. Environ. Microbiol. 60, 1174–1178. Liguori, G., Cavallotti, I., Arnese, A., Amiranda, C., Anastasi, D., Angelillo, F., 2010. Microbiological quality of drinking water from dispenser in Italy. BMC Microbiol. 10, 1–5. Pavlov, D., de Wet, C.M.E., Grabow, W.O.K., Ehlers, M.M., 2004. Potentially pathogenic features of heterotrophic plate count bacteria isolated from treated and untreated drinking water. Int. J. Food Microbiol. 92, 275–287. Sacchetti, R., Baldissarri, A., De Luca, G., Lucca, P., Stampi, S., Zanetti, F., 2006. Microbial contamination in dental unit waterlines: comparison between er:yag laser and turbine lines. Ann. Agric. Environ. Med. 13, 275–279. Sacchetti, R., De Luca, G., Zanetti, F., 2007. Influence of material and tube size on DUWL contamination in a pilot plant. New Microbiol. 30, 29–34. Sacchetti, R., De Luca, G., Zanetti, F., 2009. Control of Pseudomonas aeruginosa and Stenotrophomonas maltophilia contamination of microfiltered water dispensers with peracetic acid and hydrogen peroxide. Int. J. Food Microbiol. 132, 162–166. Szymanska, J., 2005. Electron microscopic examination of dental unit waterlines biofilm. Ann. Agric. Environ. Med. 12, 295–298. Trautmann, M., Michalsky, T., Wiedeck, H., Radosavljevic, V., Ruhnke, M., 2001. Tap water colonization with Pseudomonas aeruginosa in a surgical intensive care unit (ICU) and relation to Pseudomonas infections of ICU patients. Infect. Control Hosp. Epidemiol. 22, 49–52. Trautmann, M., Lepper, P.M., Haller, M., 2005. Ecology of Pseudomonas aeruginosa in the intensive care unit and the evolving role of water outlets as a reservoir of the organism. Am. J. Infect. Control 33, S41–S49. Walker, J.T., Marsh, P.D., 2007. Microbial biofilm formation in DUWS and their control using disinfectants. J. Dent. 35, 721–730. WHO, 2011. Guidelines for drinking water quality, fourth ed. Geneva. Zanetti, F., De Luca, G., Sacchetti, R., 2009. Control of bacterial contamination in microfiltered water dispensers (MWDs) by disinfection. Int. J. Food Microbiol. 128, 446–452.
Please cite this article in press as: Sacchetti, R., et al., Microbial quality of drinking water from microfiltered water dispensers. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.06.002
ARTICLE IN PRESS International Journal of Hygiene and Environmental Health xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
International Journal of Hygiene and Environmental Health journal homepage: www.elsevier.com/locate/ijheh
Microbial quality of drinking water from microfiltered water dispensers R. Sacchetti a,∗ , G. De Luca a , A. Dormi a , E. Guberti b , F. Zanetti a a b
Department of Medicine and Public Health, Division of Hygiene, University of Bologna, Via S. Giacomo, 12, 40126 Bologna, Italy Department of Public Health, UO Food Hygiene and Nutrition, Local Health Unit of Bologna, Via A. Gramsci 12, 40121 Bologna, Italy
a r t i c l e
i n f o
Article history: Received 2 July 2012 Received in revised form 2 May 2013 Accepted 4 June 2013 Keywords: Water dispenser Drinking water Heterotrophic plate count P. aeruginosa
a b s t r a c t A comparison was made between the microbial quality of drinking water obtained from Microfiltered Water Dispensers (MWDs) and that of municipal tap water. A total of 233 water samples were analyzed. Escherichia coli (EC), enterococci (ENT), total coliforms (TC), Staphylococcus aureus, Pseudomonas aeruginosa and heterotrophic plate count (HPC) at 22 ◦ C and 37 ◦ C were enumerated. In addition, information was collected about the principal structural and functional characteristics of each MWD in order to study the various factors that might influence the microbial quality of the water. EC and ENT were not detected in any of the samples. TC were never detected in the tap water but were found in 5 samples taken from 5 different MWDs. S. aureus was found in a single sample of microfiltered water. P. aeruginosa was found more frequently and at higher concentrations in the samples collected from MWDs. The mean HPCs at 22 ◦ C and 37 ◦ C were significantly higher in microfiltered water samples compared to those of the tap water. In conclusion, the use of MWDs may increase the number of bacteria originally present in tap water. It is therefore important to monitor the quality of the dispensed water over time, especially if it is destined for vulnerable users. © 2013 Elsevier GmbH. All rights reserved.
Introduction In Italy the quality of municipal tap water has reached a fairly high level (Anonymous, 2012a). Despite this, problems of an organoleptic nature or regarding the maintenance of the water supply have led many consumers to resort to bottled water and Italy in fact holds the record in Europe for the highest consumption of mineral water (Anonymous, 2012b). However, the elevated costs and the excessive amount of energy needed to produce bottled water have recently led to a general reassessment of tap water (Aqua Italia, 2012). In this context, the use of devices that treat the drinking water at the point of use is becoming increasingly more widespread. Such devices are marketed as being able to eliminate unpleasant odors and tastes and to remove any undesirable substances from the tap water. They often include systems for the addition of CO2 and for the cooling of the water. Compared to bottled water these devices offer the advantage of avoiding the need for the transport, storage and disposal of the bottles.
∗ Corresponding author. Tel.: +39 51 2094800; fax: +39 51 2094829. E-mail address: [email protected] (R. Sacchetti).
Numerous types of such devices are commercially available (Aqua Italia, 2012). One of the most commonly used in Italy is the microfiltered water dispenser (MWD). MWDs are devices directly attached to municipal drinking water supplies in private residences, offices, restaurants and hospitals. The treatment of the water by means of composite filters is carried out immediately before the water is dispensed. There is very little data in the literature about the quality of the water dispensed from MWDs. Baumgartner and Grand (2006), Charberny et al. (2006), Lèvesque et al. (1994), and Liguori et al. (2010) analyzed the quality of drinking water dispensed from other types of water dispensers (water coolers or soda fountains). In general, the water dispensed from these devices was found to be more contaminated than the water supplied to them. In previous studies (Sacchetti et al., 2009; Zanetti et al., 2009) we carried out laboratory tests on certain prototypes of MWDs to compare the ability of two disinfectants to ensure an adequate bacteriological quality of the dispensed water. In the present study we investigated the microbial quality of drinking water from MWDs used in collective restoration environments in an area of Northern Italy. In particular, we compared the bacteriological characteristics of microfiltered drinking water with that of municipal tap water and we studied the various factors that might influence the quality of the water dispensed by MWDs.
1438-4639/$ – see front matter © 2013 Elsevier GmbH. All rights reserved. http://dx.doi.org/10.1016/j.ijheh.2013.06.002
Please cite this article in press as: Sacchetti, R., et al., Microbial quality of drinking water from microfiltered water dispensers. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.06.002
G Model IJHEH-12698; No. of Pages 5
ARTICLE IN PRESS R. Sacchetti et al. / International Journal of Hygiene and Environmental Health xxx (2013) xxx–xxx
2
Materials and methods Microfiltered water dispensers The study focused on MWDs used in hospital and school canteens (n = 36) and a sample of MWDs in use in bars/restaurants (n = 34) in the area of Bologna, Northern Italy. The bars/restaurants were randomly selected from the list of collective restoration establishments in the territory. For each establishment selected for the study a phone call was made to verify that a MWD was in use. To be included in the study, the water dispensers had to be directly attached to municipal drinking water distribution systems, and had to use composite filters (of the EVERPURE type) for the treatment. The composite filters consist of a disposable cartridge containing a microfiltering membrane (0.5 micron pore size), made of polyethylene fibers, and powdered activated carbon. Some of them also contain a bacteriostatic element (copper or silver salts). For each MWD studied, information about the age of the device, the presence of a UV lamp, the presence of a bacteriostatic element in the filter, the amount of water consumed, the times of use/non use, the frequency and method of disinfection, the frequency of filter change, the interval between the last filter change and sampling and the last disinfection and sampling was collected by means of an ad hoc questionnaire. On the basis of this information, the continuous use of the device throughout the week was defined “daily use” and the compliance with the directions supplied by the manufacturer regarding the means and frequency of disinfection (at least 2 times/year or 4 times/year if the device was installed in a hospital or school) was defined “adequate disinfection”. The main structural and functional characteristics of the MWDs included in the study are given in Table 1. This information was not available in three devices. Water sampling From October 2010 to May 2011, 70 MWDs were examined. Samples were taken of all the various types of drinking water available (still unchilled and chilled water, or still chilled and carbonated Table 1 Structural and functional characteristics of the microfiltered water dispensers. Age in months (mean ± S.D.) Bacteriostatic element (%) Yes No UV lamp (%) Yes No Daily use (%) Yes No Liters of water consumed per day (mean ± S.D.) Frequency of filter change per year (%) 1 ≥2 Frequency of disinfection per year (%) ≤2 >2 Type of disinfectant (%) Chlorine Hydrogen peroxide Quaternary ammonium salts Adequate disinfectiona (%) Yes No Time since last filter change (days; mean ± S.D.) Time since last disinfection (days; mean ± S.D.)
44.2 (±34.9) 61.1 38.9 76.7 23.3 47.1 52.9 74.5 (±142.9) 63.2 36.8 52.6 47.4 10.0 44.0 46.0 55.0 45.0 160.0 (±155.5) 123.2 (±146.6)
a Adequate disinfection = compliance with the directions supplied by the manufacturer regarding type and frequency of disinfection.
chilled water, or still unchilled, still chilled and carbonated chilled water) in order to assess any differences in contamination and to highlight any critcal areas. At the same time a sample of the tap water entering the dispenser was also collected. The samples were always taken in the morning, after the devices had been working for about an hour. When different types of drinking water were collected from a single device, the same sampling sequence was always repeated. A total of 233 water samples were analyzed, including 70 samples of municipal tap water and 163 samples of water from MWDs (49 of still unchilled water, 63 of still chilled water, 51 of carbonated chilled water). To ensure that the samples were representative of the water consumed, flushing was not performed before sampling and the outer surfaces of the nozzles were not sterilized (Lèvesque et al., 1994; Liguori et al., 2010). Samples were collected in sterile 1 L plastic bottles containing 1 mL of sterile sodium thiosulphate solution (10%). They were kept at 4 ◦ C and analyzed within 2 h in our laboratory. Bacteriological analysis In accordance with Italian regulations for drinking water (D.Lgs 31/2001, application of EC directive 98/83), the following bacteriological parameters were quantified for each sample: Escherichia coli (EC), enterococci (ENT), indicator microorganisms of the quality of water (total coliforms – TC, heterotrophic plate count – HPC at 22 ◦ C), and supplementary microorganisms (Pseudomonas aeruginosa – PA and Staphylococcus aureus – SA). HPC at 37 ◦ C was also determined to obtain a more complete assessment of the bacteriological quality of the water. In Italy the measurement of the HPC at 37 ◦ C is required only for water sold in bottles or containers. The microbiological criteria for unbottled municipal drinking water are: absence in 100 mL per EC and ENT; absence in 100 mL per CT. The Italian regulations set no numerical value for HPC at 22 ◦ C but state that there should be no “abnormal change” compared to the values obtained during routine official checks. Also for PA no limit is set. PA is a supplementary parameter to be determined at the discretion of the local health authority. SA is a supplementary parameter, but its absence in 250 mL of water is required. All analyses followed the techniques proposed in the Standard Methods (APHA, 2005). HPCs at 37 ◦ C and 22 ◦ C were determined by the pour plate method using Plate Count Agar- Standard Methods Agar (Oxoid). The mean value of three replicates was calculated. The other microbial parameters were quantified by membrane filtration (0.45 micron pore-size sterile membrane, Millipore) in 100 mL of water (for EC, ENT, TC) and 250 mL (for PA and SA). The detection limit was 1 cfu per sample volume for all types of bacteria. EC: The filter was transferred to C-EC agar (Oxoid). After incubation at 44.5 ◦ C for 24 h, typical colonies (fluorescent green-blue under a Wood lamp and positive to indole test) were counted. Doubtful colonies underwent biochemical identification using the Enterotube II system (BBL). ENT: The filter was transferred to m-Enterococcus agar (Oxoid). After incubation at 35 ◦ C for 24–48 h, typical colonies (pink-brown in color and 0.3–2 mm in diameter) were confirmed by growth on Bile esculine agar (Oxoid) at 35 ◦ C for 48 h and by growth on Brainheart infusion broth (Oxoid) with 6.5% NaCl at 35 ◦ C for 48 h. TC: The filter was transferred to C-EC agar (Oxoid). After incubation at 37 ◦ C for 24 h, typical colonies (green-blue) were counted. Doubtful colonies underwent biochemical identification using the Enterotube II system (BBL). PA: The filter was placed on Pseudomonas CFC agar (Oxoid) and incubated at 30 ◦ C for 48 h. Colonies that were smooth, mucoid,
Please cite this article in press as: Sacchetti, R., et al., Microbial quality of drinking water from microfiltered water dispensers. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.06.002
G Model IJHEH-12698; No. of Pages 5
ARTICLE IN PRESS R. Sacchetti et al. / International Journal of Hygiene and Environmental Health xxx (2013) xxx–xxx
3
Table 2 Number (%) of positive samples and mean values of microbial counts in relation of water type. Bacteriological parameters
Municipal tap water (N = 70)
Still unchilled water (N = 49)
Still chilled water (N = 63)
Carbonated chilled water (N = 51)
Positive samples
Positive samples
Positive samples
Positive samples
No. (%) EC (Log10 cfu/100 mL) ENT (Log10 cfu/100 mL) TC (Log10 cfu/100 mL) SA (Log10 cfu/250 mL) PA (Log10 cfu/250 mL)
0 0 0 0 3 (4.3)
Mean
No. (%)
0.30
0 0 2 (4.2) 0 11 (22.9)
Mean
No. (%)
0.24
0 0 1 (1.6) 0 9 (14.3)
2.04
Mean
No. (%)
Mean
0.30
0 0 2 (3.6) 1 (2.0) 3 (5.9)
0.78 0.30 0.62
1.43
EC: Escherichia coli; ENT: enterococci; TC: total coliforms; SA: Staphylococcus aureus; PA: Pseudomonas aeruginosa.
fluorescent, blue-green or yellow-green in color, with diffuse pigmentation of the medium were presumed to be P. aeruginosa and were subsequently subcultured on TSA (Difco) and identified by the API 20NE system (bioMerieux). Atypical colonies were also counted and at least 5 colonies per plate, or all if less than 5, were subcultured on TSA (Difco) and identified by the API 20NE system (bioMerieux). SA: The filter membrane was incubated on Staph 110 medium (Oxoid) at 36 ◦ C for 40–48 h. To identify suspected colonies (dark orange in color) the API Staph System (BioMérieux) was used. Temperature and residual chlorine at the moment of sampling were also measured, with a mercury-filled Celsius thermometer and DPD colorimetric method (N,N-diethyl-p-phenylenediamine) respectively (APHA, 2005). Statistical analysis The values of microbial concentrations were converted into Log10 colony-forming units (Log10 cfu). The t test (paired) was applied to compare the logarithms of HPCs. Multivariate analysis was used to assess the interaction between the various factors that might influence the HPCs in the dispensed water. The following parameters were considered: HPC in the input water, temperature, residual chlorine, age of device, presence of bacteriostatic element in the filter, presence of UV lamp, liters of water dispensed every day, daily use, frequency of disinfection, adequate disinfection, type of disinfectant used, frequency of filter change, time since last filter change and time since last disinfection treatment. The percentage of variation in the heterotrophic plate counts of the input and output water (%) was calculated using the following formula: (log10 B − log10 A)/log10 A × 100 where A = HPC in input water and B = HPC in output water. The HPC values were expressed in Log10 (x + 1). The association between the % and the various structural and functional characteristics of the devices was calculated using simple correlation for continuous variables, Mann Whitney/Wilcoxon test for dichotomous variables and Kruskal–Wallis test for variables with more than two categories. The significance level chosen for all analyses was p < 0.05. Analyses were performed using the Statistical Package for Social Sciences software (SPSS, version 17). Results The results of the microbial analyses performed on samples of water from the MWDs and tap water are shown in Table 2. EC and ENT were not detected in any of the water samples. TC were never detected in the tap water, but were found in 5 samples of water taken from 5 different MWDs, in concentrations ranging from 0.30 to 1.56 Log10 cfu/100 mL. SA was recovered in only 1 water sample from an MWD.
Contamination by PA at very low levels (0–0.60 Log10 cfu/250 mL) was detected in 3 tap water samples (4.3%); in 16 MWDs (22.9%) at least one sample of dispensed water was found to be contaminated by PA. The highest levels of PA were detected in still unchilled and still chilled water, where the values reached respectively 3.26 Log10 cfu/250 mL and 3.68 Log10 cfu/250 mL. The differences between the mean HPC values in the various types of microfiltered water vs those in the tap water and the relative statistical significance are shown in Fig. 1. The mean HPCs at 22 ◦ C and 37 ◦ C were significantly higher in microfiltered water samples compared to those of the input water. The highest HPC values were observed in the carbonated water, where they reached 3.70 Log10 cfu/mL at 37 ◦ C and 3.46 Log10 cfu/mL at 22 ◦ C. The mean HPCs at 22 ◦ C and 37 ◦ C were significantly higher in carbonated water compared to chilled and unchilled still water. The mean temperature was 19.5 ◦ C for tap water, 23.5 ◦ C for still unchilled water, 9.8 ◦ C for still chilled water and 9.7 ◦ C for carbonated chilled water. The mean residual chlorine content was 0.18 mg/L for tap water, 0.10 mg/L for still unchilled water and 0.08 for still chilled water. In the multivariate analysis no parameter was found to be significant vs HPC in the dispensed water either at 22 or 37 ◦ C. The % at 22 ◦ C was seen to be directly and significantly (p < 0.01) correlated with the age of the device in the still unchilled water, the still chilled water and in the carbonated chilled water. When the disinfection was adequate the % at 22 ◦ C in the still unchilled water was lower and statistically significant (p < 0.01) compared to when the disinfection was not adequate. A negative trend was observed between the increase in HPC values and the residual chlorine present in the water dispensed. Discussion The aim of this study was to assess the quality of the water dispensed from the MWDs used for collective restoration in our territory. The exact number of MWDs used throughout the area is unknown, since the local health authority is often not notified of the installation of the devices. Although destined for human consumption, this type of water is rarely monitored and very little is known about its bacteriological quality. No specific microbial criteria exist but the water dispensed must nevertheless meet the required legal standards for drinking water. The results of our study revealed no contamination from EC and ENT in any of the water samples examined. In accordance with Italian norms, the absence of these microorganisms, considered to be indicators of fecal pollution, confirms the safety of the water with regards enteropathogenic bacterial agents. TC were detected in only 5 samples of microfiltered water. The low levels detected and the absence of TC in the input water would suggest that the presence of these microorganisms in the microfiltered water derives from a contamination of the distribution points of the devices. This
Please cite this article in press as: Sacchetti, R., et al., Microbial quality of drinking water from microfiltered water dispensers. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.06.002
G Model IJHEH-12698; No. of Pages 5 4
ARTICLE IN PRESS R. Sacchetti et al. / International Journal of Hygiene and Environmental Health xxx (2013) xxx–xxx
Fig. 1. Mean HPC values in the various types of microfiltered water vs tap water and relative statistical significance.
can easily occur when the nozzles are protruding and the appliances are used by a large number of people. TC are among the parameters considered to be indicators of water quality; given that some species of TC are of confirmed environmental origin, the health risk associated with the occasional presence of these microorganisms in the water is assessed, case by case, by the local health authority. While the absence of EC and ENT must always be rigorously respected, the occasional failure to conform to the norms set for TC can be tolerated, at least in the first instance. As far as the supplementary parameters are concerned, the presence of SA found in a single sample of microfiltered water can equally be explained by human contamination of the distribution point. The greater frequency and higher concentrations of PA detected in the dispensed water as opposed to the input water, confirm the
microorganism’s ability to colonize the circuits of MWDs (Sacchetti et al., 2009). It is well known that PA naturally present in the environment may be the source of disease in vulnerable subpopulations (the elderly or the very young, immunocompromised patients) (Aumeran et al., 2007; Trautmann et al., 2001, 2005). Moreover, the presence of high numbers of PA in drinking water may be associated with complaints about taste, odor and turbidity (WHO, 2011). We found that the HPCs in the microfiltered water were statistically higher than those in the tap water. However, since no scientific evidence exists to suggest that, in the absence of fecal contamination, HPC values are directly related to a health risk in a healthy population (WHO, 2011), no limit has been established for HPCs in current Italian regulations. Nevertheless, an increase in the level of HPCs is an indication of bacterial regrowth (Bonadonna
Please cite this article in press as: Sacchetti, R., et al., Microbial quality of drinking water from microfiltered water dispensers. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.06.002
G Model IJHEH-12698; No. of Pages 5
ARTICLE IN PRESS R. Sacchetti et al. / International Journal of Hygiene and Environmental Health xxx (2013) xxx–xxx
et al., 2008) and can lead to a deterioration in the esthetic quality of the water (e.g. taste, odor and discoloration) (Chowdhury, 2011). It cannot be ruled out, though, that high HPC values may represent a risk for the vulnerable users, since HPCs include many opportunistic pathogenic bacteria in water habitats (Achromobacter, Pseudomonas, Moraxella, Stenotrophomonas) (Allen et al., 2004; Pavlov et al., 2004). The highest HPC values were observed in the carbonated water. This could be explained by the presence of a small carbonization reservoir in the MWDs, where the water tends to stagnate during periods of non use. In addition, as highlighted in the information collected for the study, this reservoir is often not disinfected so as to avoid technical problems and organoleptic alterations to the water. In our study the HPC values were used to make a statistical analysis of the influence of specific characteristics of the MWDs on the quality of the dispensed water. In the multivariate analysis no parameter was found to be significant vs HPC in the dispensed water. This may be due in part to the small sample size and the considerable variability of the characteristics considered. The direct statistically significant correlation between the % at 22 ◦ C and the age of the device, can probably be explained by the progressive formation of biofilm in the water circuits, providing a favorable habitat for psychrophilic bacterial flora. The gradual release of planktonic bacteria from the biofilm leads to an increase in the HPC values at 22 ◦ C in the dispensed water. The water lines of MDWs are made of a plastic material and are a few millimeters in diameter, while their length may reach several meters. It is well known that in devices where water flows through a network of narrow bore tubing made of plastic material, bacteria can form multispecies adherent biofilm on the inside of the waterlines (Sacchetti et al., 2006, 2007; Szymanska, 2005; Walker and Marsh, 2007). A negative trend was observed between the increase in HPC values and the residual chlorine present in the water dispensed. As expected, the reduction in residual chlorine observed in the microfiltered water compared to the tap water is a factor that favors bacterial regrowth. Also periods of stagnation are well documented as favoring bacterial regrowth; however, in our study, given the extreme variability of the conditions of use/non use of the devices, it was not possible to observe any type of association. Finally, the increase in the HPC levels at 22 ◦ C was lower when the disinfection conformed to the manufacturer’s directions given in the operating manual of the device. This confirms the importance of an adequate disinfection of the devices. The measurement of HPC levels could represent the most simple means of monitoring the bacterial regrowth in the water circuits and the possible presence of opportunistic pathogenic bacteria. Acceptable HPC levels could be those set by the municipal drinking water regulations and the guidelines of other countries (e.g. 100 cfu/mL in Germany and the Netherlands or 500 cfu/mL in the USA and Canada) (Chowdhury, 2011), or the limits set by the Italian norms for municipal drinking water on sale in bottles or containers (100 cfu/mL at 22 ◦ C and 20 cfu/mL at 37 ◦ C). Moreover, the measurement of HPC may be useful to assess the efficacy of the disinfection procedures. However, specific testing for the presence of PA may also be advisable, especially in situations at risk.
5
In conclusion, the use of MWDs may increase the number of bacteria originally present in tap water. It is therefore important to monitor the quality of the dispensed water over time, especially if the MWDs are quite old and/or installed in establishments for vulnerable users. Acknowledgments The authors would like to thank Mara Veronesi and Morena Bertelli, Mauro Daolio, Giorgia Mistroni, Giampaolo Murtas, Dimitri Zuffa for their valuable assistance in collecting samples. They also thank Giovanni Lorusso for the technical assistance in culturing water samples. References Aqua Italia, 2012. Consumo di acqua potabile presso la popolazione italiana, www.aquaitalia.it Allen, M.J., Edberg, C.S., Reasoner, D.J., 2004. Heterotrophic plate count bacteriaWhat is their significance in drinking water? Int. J. Food Microbiol. 92, 265–274. Anonymous, 2012a. Sai cosa bere? Altroconsumo 261, 10–15. Anonymous, 2012b. Annuario Bevitalia 2011–12. Beverfood ed. APHA, 2005. Standard Methods for the Examination of Water and Wastewater, 21st ed. APHA-AWWA-WEF, Washington, DC. Aumeran, C., Paillard, C., Robin, F., Kanold, J., Baud, O., Bonnet, R., Souweine, B., Traore, O., 2007. Pseudomonas aeruginosa and Pseudomonas putida outbreak associated with contaminated water outlets in an oncohaematology paediatric unit. J. Hosp. Infect. 65, 47–53. Baumgartner, A., Grand, M., 2006. Bacteriological quality of drinking water from dispensers (coolers) and possibile control measures. J. Food Prot. 69, 3043–3046. Bonadonna, L., Memoli, G., Chiaretti, G., 2008. Formazione Di Biofilm Sul Materiale A Contatto Con Acqua: Aspetti Sanitari E Tecnologici Rappoti ISTISAN 08/19. Charberny, I.F., Kaiser, P., Sonntag, H.G., 2006. Can soda fountains be recommended in hospitals? Int. J. Hyg. Environ. Health 209, 471–475. Chowdhury, S., 2011. Heterotrophic bacteria in drinking water distribution system: a review. Environ. Monit. Assess. 184 (10), 6087–6137. Lgs, D., 31/2001. Attuazione della Direttiva 98/83/CE relativa alle acque destinate al consumo umano. G.U. della Repubblica Italiana. Supplemento n◦ 52 del 3 marzo. Lèvesque, B., Simard, P., Gauvin, D., Gingras, S., Dewailly, E., Letarte, R., 1994. Comparison of the microbiological quality of water coolers and that of municipal water systems. Appl. Environ. Microbiol. 60, 1174–1178. Liguori, G., Cavallotti, I., Arnese, A., Amiranda, C., Anastasi, D., Angelillo, F., 2010. Microbiological quality of drinking water from dispenser in Italy. BMC Microbiol. 10, 1–5. Pavlov, D., de Wet, C.M.E., Grabow, W.O.K., Ehlers, M.M., 2004. Potentially pathogenic features of heterotrophic plate count bacteria isolated from treated and untreated drinking water. Int. J. Food Microbiol. 92, 275–287. Sacchetti, R., Baldissarri, A., De Luca, G., Lucca, P., Stampi, S., Zanetti, F., 2006. Microbial contamination in dental unit waterlines: comparison between er:yag laser and turbine lines. Ann. Agric. Environ. Med. 13, 275–279. Sacchetti, R., De Luca, G., Zanetti, F., 2007. Influence of material and tube size on DUWL contamination in a pilot plant. New Microbiol. 30, 29–34. Sacchetti, R., De Luca, G., Zanetti, F., 2009. Control of Pseudomonas aeruginosa and Stenotrophomonas maltophilia contamination of microfiltered water dispensers with peracetic acid and hydrogen peroxide. Int. J. Food Microbiol. 132, 162–166. Szymanska, J., 2005. Electron microscopic examination of dental unit waterlines biofilm. Ann. Agric. Environ. Med. 12, 295–298. Trautmann, M., Michalsky, T., Wiedeck, H., Radosavljevic, V., Ruhnke, M., 2001. Tap water colonization with Pseudomonas aeruginosa in a surgical intensive care unit (ICU) and relation to Pseudomonas infections of ICU patients. Infect. Control Hosp. Epidemiol. 22, 49–52. Trautmann, M., Lepper, P.M., Haller, M., 2005. Ecology of Pseudomonas aeruginosa in the intensive care unit and the evolving role of water outlets as a reservoir of the organism. Am. J. Infect. Control 33, S41–S49. Walker, J.T., Marsh, P.D., 2007. Microbial biofilm formation in DUWS and their control using disinfectants. J. Dent. 35, 721–730. WHO, 2011. Guidelines for drinking water quality, fourth ed. Geneva. Zanetti, F., De Luca, G., Sacchetti, R., 2009. Control of bacterial contamination in microfiltered water dispensers (MWDs) by disinfection. Int. J. Food Microbiol. 128, 446–452.
Please cite this article in press as: Sacchetti, R., et al., Microbial quality of drinking water from microfiltered water dispensers. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.06.002