Risk assessment of selected opportunisticpathogens in drinking water.

Item Type Dissertation-Reproduction (electronic); text

Authors Chaidez Quiroz, Cristobal,1969-

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this materialis made possible by the University Libraries, University of Arizona.Further transmission, reproduction or presentation (such aspublic display or performance) of protected items is prohibitedexcept with permission of the author.

Download date 05/05/2018 07:04:31

Link to Item http://hdl.handle.net/10150/191233

RISK ASSESSMENT OF SELECTED OPPORTUNISTIC PATHOGENS IN

DRINKING WATER

by

Cristobal Chaidez Quiroz

A Dissertation Submitted to the Faculty of the

GRADUATE INTERDISCIPLINARY PROGRAM IN NUTRITIONAL SCIENCES

In Partial Fulfillment of the Requirements For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

1999

1 h L. Price

(0 ,tA-0-n) V.Lightner

2

THE UNIVERSITY OF ARIZONAGRADUATE COLLEGE

As members of the Final Examination Committee, we certify that we have

read the dissertation prepared by Cristobal Chaidez Quiroz

entitled Risk Assessment of Selected Opportunistic Pathogens

in Drinking Water

and recommend that it be accepted as fulfilling the dissertation

requirement for the Degree of Doctor in Philosophy

rCha P. Ger ap

/.

142 49Date

I / ze Mc",Date

V 31-/c0 Date

Ag-/q9 Date

( 2 99DaLe

Final approval and acceptance of this dissertation is contingent uponthe candidate's submission of the final copy of the dissertation to theGraduate College.

I hereby certify that I have read this dissertation prepared under mydirection and recommend that it be accepted as fulfilling the dissertationrequirement.

,L$ 1,7Dissertation Director

Charles P. Gerba

1 /2-9Date

STATEMENT BY THE AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for anadvance degree at The University of Arizona and is deposited in the University Library tobe made available to borrowers under rules of the Library.

Brief quotation from this dissertation are allowable without special permission,provided that accurate acknowledgment of source is made. Request for permission forextended quotation from or reproduction of this manuscript in hole or in part may begranted by the head of the major department or the Dean of the Graduate College when,in his or her judgment, the proposed use of the material is in the interests of scholarship.In all other instances, however, permission must be obtained by the author.

3

Signed:

ACKNOWLEDGMENTS

I would like to express my sincere acknowledge to Dr. Charles P. Gerba, mymajor advisor, for the guidance and encouragement he provided throughout the course ofthis research.

I would also extend my thanks to Dr. Pat A. Rusin, Dr. Donald V. Lightner, Dr.Glenn J. Songer and Dr. Ralph Price for their service as the advisory committee as well astheir suggestions and insights toward the completion of this work.

Special thanks are for Mr. Jaime Naranjo. I would not be able to finish my work,without your help. Thanks for everything Jaime (my friend).Also my appreciation is for Dr. (Don) Carlos Enriquez. Your suggestion and friendshipmeant a lot to me.

My sincere thanks to my buddy Jorge Sandoval. I will never forget those days wespent talking about "the wonder years".

I would like to thank Dr. Adela Allen and Geraldine Olds (from the Graduatecollege) for being so helpful.

And finally, the Gerbais (Dr. Gerba's students): Thanks a lot to the Mexican crew:Dr. Maria Quifionez, Juan Antonio Vidales, Pablo Gortarez, Tomas Sepulveda y RicardoEnriquez. And also thanks to Pam, Patricia, Faezeh, Amy, Cherry, Jeanette, Denise , Luis.Mohammad, Absar, Frank and Hiroshi. And also for the old gang: Edlin, Manuella, Danaand Seema.

Thanks to all of you for being my friends all this time and making this task much easier.I will never forget those days at Gerba's lab.

4

DEDICATION

This work is dedicated to my parents, Cristobal Chaidez Medina and Leonor

Quiroz de Chaidez, to my sisters, Rosa Delia and Silvia Patricia, and for their kids

(Genessys, Daniel and Hansel), who were my constant inspiration during all those years.

I would also extend this dedication to my friend and lovely wife, Dr. Marcela

Vergara for your patience and support during the preparation of this work. I will love

you forever.

5

6

TABLE OF CONTENTSI. LIST OF FIGURES 8

II. LIST OF TABLES 9

ABSTRACT 11

III. INTRODUCTION 13

Problem Definition 13

Literature Review 14

The Risk Assessment 14

Microbial Risk Assessment 15

Opportunistic Pathogens in Water 19

Dissertation Format 45

IV. PRESENT STUDY 46

APPENDIX 1: MICROBIOLOGICAL QUALITYOF WATER VENDING MACHINES 48

APPENDIX 2: MICROBIOLOGICAL SURVEYOF PRIVATE ROOF WATER TANKS INCULIACAN, MEXICO 68

APPENDIX 3: MICROBIOLOGICALCOMPARISON OF THE QUALITY OFPOU-TREATED WATER, TAP WATER,AND BOTTLED WATER 85

TABLE OF CONTENTS-Continued

APPENDIX 4: AEROMONAS HYDROPHILAAND PSELTDOMONAS AERUGINOSA INDRINKING WATER FROM VARIOUSSOURCES: A RISK ASSESSMENT 113

APPENDIX 5: CONCLUSION 139

APPENDIX 6: BACTERIAL DATA 142

REFERENCES 154

7

LIST OF FIGURES

FIGURE 1.1, Percentage of positive samplesfor Pseudomonas aeruginosa (CFU/500 mL),total coliforms (CFU/100 mL) and heterotrophicplate count (HPC) bacteria greater than 500 CFU/mL 57

FIGURE 2.1, Percentage of samples positivefor Pseudomonas aeruginosa (CFU/500 mL),total and fecal coliforms (CFU/100 mL) andheterotrophic plate count (HPC) bacteria greaterthan 500 CFU/mL 77

FIGURE 3.1, Mean HPC bacteria densities (CFU/mL) 109

FIGURE 3.2, Mean Pseudomonas aeruginosadensities (CFU/500 mL) 110

FIGURE 3.3, Mean Aeromonas hydrophiladensities (CFU/500 mL) 111

FIGURE 3.4, Mean total coliforms densities (CFU/100 mL) 112

8

LIST OF TABLES

TABLE 1.1, Bacteriological quality of waterfrom vending machines 58

TABLE 1.2, Weekly sampling of three watervending machines 59

TABLE 1.3, Nozzle swab test results of fifteenwater vending machines 59

TABLE 1.4, Drain swab test results of fifteenwater vending machines 60

TABLE 1.5, Maximum and minimum valuesfor the physico-chemical parameters in analyzedwater from vending machines 61

TABLE 2.1, Bacteriological quality of waterfrom household roof tanks 78

TABLE 2.2, Average, maximum andminimum values for the physico-chemicalparameters in analyzed water from householdroof tanks 79

TABLE 3.1, General characteristics ofpoint-of-use (POU)-devices (faucet-mounted units) 91

TABLE 3.2, Percentage bacterial occurrencein POU-treated water, tap water (with POU),tap water (without POU), and bottled water 104

TABLE 3.3, Average, maximum, minimumvalues of bacteria in drinking water sources 105

9

TABLE 3.4, Occurrence of the HPC in drinkingwater sources 106

LIST OF TABLES-Continued

TABLE 3.5, Bacteria isolated and identified indrinking water samples 107

TABLE 3.6, Average (ave), maximum (max),and minimum (min) values for the physico-chemicalparameters in analyzed water (POU-treated water,tap water with POU, tap water without POU,and bottled water) 108

TABLE 4.1, Isolation of A. hydrophila fromdrinking water sources (CFU or %) 124

TABLE 4.2, Isolation of P. aeruginosa fromdrinking water sources (CFU or %) 125

TABLE 4.3, Risk of infection of Aeromonashydrophila and Pseudomonas aerugin osa inselected drinking water sources (2L/day/person) 127

TABLE 4.4, Risk of infection of Aeromonashydrophila and Pseudomonas aeruginosa inselected drinking water sources (4L/day/person) 128

TABLE 4.5, Risk of infection of A. hydrophilain bottled water excluding one high value 128

TABLE 4.6, Daily bacterial concentrations usedfor the risk of infection based on two and fourliter consumption 129

10

11

ABSTRACT

Water as a route of opportunistic bacterial disease transmission has not been well

established. The use of epidemiological evidence linking drinking water bacterial

contamination to health effects in a population is lacking and very costly to obtain. Also,

the significance of exposure to low-level contamination is difficult to determine

epidemiologically. This makes it difficult to estimate the impact on a community. The

use of risk assessment approach allows an understanding of low-level exposure; and to

define it in a more quantitative fashion. Microbial risk assessment was employed to

determine the risks associated with exposure to selected opportunistic bacterial pathogens

(Aeromonas hydrophila and Pseudomonas aeruginosa) present in drinking water from

various sources. An extensive analysis was conducted on drinking water obtained from

various sources including point-of-use (POU)-treated water, tap water with POU-

connection, tap water, bottled water, and water from vending machines and storage tanks.

Enumerated bacteria included: A. hydrophila, heterotrophic plate count (HPC) bacteria,

Mycobacteriuni spp., Plesiomonas shigelloides, P. aeruginosa, and total and fecal

coliforms. It was found that opportunistic pathogens were present in small numbers in

drinking water. Neither fecal coliforms nor P. shigelloides were found in the drinking

water samples. The annual risks of colonization based on the consumption of

2L/day/person for drinking water were determined to be as high as 7.9x10 - ' and 9.9x10 -4

for A. hydrophila and P. aeruginosa, respectively at exposure levels ranging from 90 to

12

10 CFliimL. respectively. The results obtained indicates that the risk of colonization is a

transient process, and the probability of infection may be very but could result in the most

vulnerable (very young, the elderly and immunocompromised).

More studies are needed on the occurrence of opportunistic pathogens in drinking

water from various sources and animal andjor human feeding studies to better define

dose-response in both healthy and immunocomprimised individuals. There is no doubt

that the greatest need for microbial risk assessment is the occurrence data. Therefore,

national surveys in drinking water from various sources will help in the developing of

microbial risk assessment for opportunistic bacterial pathogens. The use of conventional

methods as well as molecular approaches are recommended in order to obtain a more

accurate identification of waterborne bacterial pathogens.

CHAPTER 1

INTRODUCTION

The purpose of this dissertation was to assess the risks associated with the

exposure to selected opportunistic bacterial pathogens in drinking water from various

sources using a microbial risk assessment approach. This was accomplished by

determining the occurrence and concentration of selected opportunistic bacterial

pathogens in various drinking water sources.

Problem Definition

Opportunistic bacterial pathogens continues to occur in the United States despite

present technologies available for water treatment. This may be due to a breakdown in

treatment such as inability of the treatment process to remove all of the pathogenic

organisms present in the raw water, contamination after treatment, or bacterial re-growth

in the distribution system. Opportunistic bacterial pathogens are capable of surviving on

minimal nutrients, in pipes sediments, and becoming participants of biofilm formation.

Segments of the population particularly at risk of infection are newborn babies, the

elderly and immunocompromised individuals. However, their public health significance

13

14

with regard to the population at large is not well known. It is difficult. to determine the

human health significance of exposure to low-level contamination of water supplies.

Furthermore, data are limited on the occurrence of specific opportunistic pathogens in

drinking water. Better quantification of the occurrence of opportunistic bacteria is

necessary to understand the human health risk associated with exposure.

Literature Review

The Risk Assessment

Risk may be defined as the probability or likelihood of an adverse effect occurring

due to consequences set forth from a particular hazard (chemical, landfill, biologically-

contaminated food or water, or even a person's own behavior). The science of risk

assessment involves evaluating the risks posed to a society or to the environment in order

to better understand the scope of the problems that may result from the exposure to

particular hazards. Information learned through risk assessments can be used to help

policy-makers make informed decisions concerning the risks posed from a given hazard.

Risks analysis is the process of evaluating risks and risk issues. The process

involves risk assessment, risk management, and risk communication (NRC, 1994). The

risk assessment is the first step of the risk analyses process.

There are four fundamental steps in the risk assessment framework: 1) hazard

identification: 2) dose-response assessment; 3) exposure assessment; and, 4) risk

15

characterization (Rose et al.. 1991). The hazard identification step is the most easily

followed. Basically, a substance is determined as harmful or not based on laboratory and

field data as well as information obtained from epidemiological studies. The dose-

response assessment involves determining the relationship between the dose of the hazard

and the incidence of the adverse health effect. This step uses animal studies and requires

the extrapolation from high to low doses and from animals to humans (Macler and Regli,

1993).

The objective of the exposure assessment step is to measure the frequency and

intensity of the exposure, the route of exposure, and the population exposed. The

objective of the risk characterization step (which can be qualitative or quantitative) is to

estimate the risk of an adverse health effect occurring based on exposures determined in

the exposure assessment.

The use of risk assessment framework to estimate risks associated with

environmental exposure to microorganisms is of more recent development and is

discussed in the following section.

Microbial Risk Assessment

Risk assessment methodology has been developed to better understand the

significance of exposure to microorganisms in water. Epidemiological data obtained

from waterborne outbreaks provide information on the human health impacts of microbial

contaminated water. However, it is difficult to understand the impacts associated with

16

exposure to low levels of contamination (Rose et al., 1991). Risk assessment helps to

understand the significance of exposure to low-levels of microorganisms in water.

The four-tiered approach is used in the risk assessment framework. Hazard

identification, dose-response assessment, exposure assessment, and risk characterization

(NRC, 1983), can also be applied to assessing risks associated with exposure to

microorganisms. Hazard identification is the identification of the microbial agent as well

as the spectrum of the human illness and disease associated with the specific

microorganisms. Dose-response is the characterization of the relationship between the

dose administered and the probability of infection or disease on the exposed population.

Exposure assessment determines the size and the nature of the population exposed and

the route. Risk characterization is the integration of the three steps in order to estimate

the magnitude of the public health problem (NRC, 1983).

Two mathematical models have been shown to adequately describe the infection

process demonstrated in the dose-response studies of pathogenic microorganisms (Haas,

1983; Rusin et al., 1997). The exponential model:

Pi = 1 - e -(1/K) (N);

Where Pi = the probability of infection, 1/k= the fraction of ingested microorganisms

that survive to initiate infection, and N= number of microorganisms ingested or inhaled.

17

The beta-Poisson model.

Pi = 1 -(1 ± N/I3

Where Pi = the probability of infection, N= the number of microorganisms ingested, and

B represents the parameters of the host-virus interaction. Risk estimates illness and death

can also be computed from these models by incorporating morbidity and mortality ratios

of that particular microorganism (Haas et al., 1993).

Exposure assessment determines the amount of water consumed by an individual

as well as the concentration of microorganisms (e.g. bacteria). The USEPA uses two

liter/person/day for risk estimates of drinking water (Roseberry and Burmaster, 1992).

Risk assessment has been used recently to evaluate the human health impact of exposure

to microorganisms in food and various water supplies (Hass, 1983; Gerba and Haas,

1988; Rose et al., 1991; Regli et al., 1991; Rusin et al., 1997). The USEPA recommends

that microbial risks of infection should not exceed 1/10,000 (10) for drinking water for a

yearly exposure (Macler and Re2li, 1993).

The application of the risk assessment framework to viruses as a group has been

reviewed (Gerba and Haas, 1988; Regli et al., 1991; Haas et al., 1996). Rose et al.,

(1991) developed a risk assessment model to estimate the risk of infection after exposure

to treated waters contaminated with Giardia cysts. More recently, a risk assessment of

rotavirus has been conducted (Gerba et al., 1996). Rusin et al., (1997) estimated the daily

risk of bacterial (heterotrophic plate count) infection using the exponential model. Daily

18

risk were based on the consumption of 2 liters of water per day. Rusin et al., (1997)

summarized that the probability of infection is low and emphasized that further research

is needed to determine the occurrence of selected opportunistic pathogens in drinking

water.

Dose-response studies using human volunteers have been conducted for several

microorganisms including Pseudomonas aeruginosa (Buck and Cooke, 1969),

Aeromonas hydrophila (Morgan et al., 1985). Mycobacterium avium (Murphey et al.,

1983), rotavirus (Ward et al., 1986), hepatitis A virus (Ward et al., 1958), poliovirus

(Koprowski. 1956), and Cryptosporidium (Dupont et al., 1995). More data are needed on

the quantitative occurrence of opportunistic bacterial pathogens in drinking water.

The application of risk assessment methodology to estimate risks posed by

specific bacteria would provide a better understanding of the significance of bacterial

waterborne disease. This methodology can be applied to drinking water in both

developed and developing countries.

An Extensive literature review on opportunistic bacterial pathogens is presented in

the next section. Opportunistic pathogens were selected for this study. The research

reported here is original, hence, no such information are available from previously

reported studies.

19

A eromonas hydrophila

The genus Aeromonas occur as straight cells that are rod-shaped with rounded

ends. Their size range is 0.3 to 1.0 i_tm in diameter and 1.0 to 3.5 lam in length and they

exist singly, in pairs or short chains. They are Gram-negative, facultative anaerobes and

generally motile by a single polar flagellum. Metabolism of glucose is both fermentative

and respiratory. They are oxidase and catalase-positive, reduce nitrates to nitrites, and

utilize carbohydrates with the production of acid with gas. They are resistant to the

vibriostatic agent 0/129 with optimum growth temperature of 22-28°C (Rippey and

CabeIli, 1979).

While Aeromonas are undoubtedly more commonly isolated from patients with

gastroenteritis, reports of Aeromonas sepsis, wound and ocular infections have appeared

increasingly in the literature (Janda and Duffey, 1988). Unlike gastroenteritis, these

infections are often reported to have serious debilitating outcomes (George et al., 1985).

The highest seasonal incidence of A. hvdrophilcz diarrhea in the Unites States

occurs during summer months, some studies indicate that A. hydrophila infection is the

second or third leading cause of bacterial gastroenteritis during these months (Moyer,

1987; LeChevallier et al., 1982). Seidler et al., (1980) found Aeromonas in distribution

water systems during the summer months. Evidence suggests that this species is an

enteric pathogen around the world. However, the incidence of Aeromonas associated

diarrhea is relatively low in developed countries (Havelaar and Vonk, 1988).

Gastroenteritis due to A. hydrophila resembles that of shigellosis, in which the stool

20

specimen is bloody in appearance and numerous fecal leukocytes are present (George et

al., 1985). Since 1968, A. hydrophda species have been recognized as an opportunistic

pathogen that affects young children, elderly people and the immunocompromised host

( Von Gravenitz and Mensh, 1968). The pathogenicity of A. hydrophda has been

associated with the production of exotoxins. Clinical and environmental strains of A.

hydrophda have been reported to produce both heat-labile and heat-stable cytotoxins that

have enterotoxic activities (Janda, 1991). Despite the production of extracellular enzymes

and toxins, the pathogenesis of A. hydrophila infections remains unclear (Handfield et al.,

1996).

The bacteria can be cultured on a wide range of agar media containing ampicillin.

Handfield et al., (1996) efficiently used ampicillin-dextrin agar with a standard

membrane filtration procedure for isolating A. hydrophda from drinking water. The

choice of a specific medium for isolation of Aeromonas spp. will depend on the type of

sample to be examined and whether a detection or quantitative determination is needed

(Jeppesen, 1995). Conventional biochemical tests help to identify Aeromonas (Janda et

al., 1995). Chromogenic substrates (API systems) and carbohydrate utilization (BIOLOG

GN microplates) are utilized (Janda et al., 1995). Aeromonas strains can also be identified

by fingerprinting (multilocus enzyme electrophoresis) (Altweg et al., 1991), ribotyping

(Martinetti and Altwegg, 1992), and protein electrophoresis (Kuijper et al., 1989).

Aeroinonas can survive standard chlorination and thus re-colonize the water

distribution networks after the chlorination process (Van der Kooij, 1988). Aeromonas

21

have been isolated from chlorinated water systems, where they persist as a component of

biofilms (Van der Kooij, 1991). They represent from 17 to 25% of the total indicator

bacteria recovered (Clark et al., 1982). The bacterium has been found to survive for up to

60 days and proliferate to levels >10 5 CFU/mL in water stored at room temperature

(Warburton, 1994b).

A. hydrophila is a ubiquitous bacterium frequently isolated from food, drinking

water, and aquatic environments (Holmberg et al., 1986). It has been recovered from 0.6-

18.2% of natural fresh water samples at concentration range of 0.1-3600 CFU/mL (Rusin

et al., 1997). Handfield et al., (1996) recovered 94 CFU of A. hydrophila from

chlorinated and unchlorinated drinking-water supplies. Small numbers of A. hydrophila

(6.6 to 12 °A) have been recovered from tap water (Chaidez and Gerba, 1998; Levesque et

al., 1994)

Chaidez and Gerba, (1998) and Tsai and Yu, (1997) recovered small numbers (1.4

to 6.9%) of the bacterium from bottled water. Two European studies from Spain and

Portugal have identified A. hydrophila (6.6 to 50%) in bottled mineral waters (Gonzalez

et al., 1987; Manaia et al., 1990). Hunter (1994) failed to recover the bacterium from

bottled water samples obtained from the United Kingdom. A. hydrophila as well as other

opportunistic pathogens have been encountered in plumbing system biofilms (Dennis et

al., 1989). Levesque et al., (1994) observed the bacterium in the water cooler biofilms

and recovered it from 22% of the samples tested.

22

Volunteer feeding studies with large oral doses (up to 10 10 CFU) have failed to

produce human illness (Morgan et al., 1985). Two subjects showed symptoms of

diarrhea with doses of 10' CFU, but the organism did not appear in the stools. There is

controversy as to whether the organism is a cause of human gastroenteritis. It possesses

several attributes that may act as a pathogen for humans. Also, the presence in the stools

of individuals with diarrhea suggests that the bacterium may be an infectious agent to

humans.

Coliform bacteria

The coli form group consists of several genera of bacteria belonging to the family

Enterobacteriaceae (Bitton, 1994). Traditionally these genera include Escherichia,

Citrobacter, Enterobacter and Klebsiella. However, using more modern taxonomical

criteria, the group is heterogenous and includes non-fecal lactose-fermenting bacteria as

well as other species which are rarely found in feces but are capable of multiplication in

water (Gleeson and Gray, 1997).

The coliform group is normally defined as compromising all aerobic and

facultative anaerobic, Gram-negative, non-spore forming, rod-shaped bacteria that

develop a red colony with a metallic sheen within 24 h at 35°C on an Endo type medium

containing lactose (APHA, 1995). They are also oxidase-negative, non-spore forming

and display beta-galactosidase activity. The coliform group also includes the

then-notolerant fecal coliforms. These are defined as being able to ferment lactose at

23

44.5 'C (WHO. 1993), and not only include E. cou, but also species of the Klebsiella,

Enterobacter and Citrobacter genera (Bitton, 1994). E. colt is considered to be the only

true fecal coliform as other thermotolerante coliforms can be derived from non-fecally

contaminated water (Gleeson and Gray, 1997).

The coliform index is still widely considered to be reliable indicator for potable

water (Gleeson and Gray, 1997). However, its use has been questioned as a measure of

water quality. The coliform group have several deficiencies as indicator organisms: re-

growth in aquatic environments and distribution systems, suppression by high

background bacterial growth, not indicative of health threat, a lack of correlation between

protozoan and viral numbers, and the occurrence of false-positive and false-negative

results (Gleeson and Gray, 1997; Bitton, 1994).

In the bacteriological examination of drinking water, emphasis is often placed on

frequent sampling and simpler, cheaper tests as opposed to occasional sampling by more

expensive methods (Gleeson and Gray, 1997). As both the multiple tube method and

membrane filtration (MF) techniques are recognized as being relatively simple and

inexpensive they have tended to become the leader of microbiological water assessment

worldwide. Both of these methods are based on the ability of coliforms to produce acid

and gas from lactose-based media (Gleeson and Gray, 1997). Despite their widespread

use, it is recognized that both methods suffer from inherent faults; the MPN is

fundamentally inaccurate, whereas the sensitivity of the MF procedure can be affected by

a number of variables (Bitton, 1994).

24

Rapid tests that can currently be performed for detecting coliforms have little

specificity and may include non-viable cells (PCR, Gene Probes), and false-positive

results (Colilert system) (Bitton, 1994).

Over the years. several groups of bacteria have been suggested as alternative of

fecal indicators. Organisms such as fecal streptococci and Clostridium perfringens have

widely used as indicator of fecal pollution (Geldreich, 1996; Payment et al., 1997). Many

proposed indicator organisms involve complex and expensive methodologies which are

beyond the means of capabilities of smaller utilities and developing countries and,

therefore, can not be applied as a routine basis (Gleeson and Gray, 1997). It would

therefore appear that currently used bacterial indicators will go on being used throughout

the world until an alternative is found that meets all the criteria established for indicator

organisms and that is acceptable in terms and simplicity (Bitton, 1994).

In conclusion, it would appear that there is still a place for the coliform as a

measure of treatment effectiveness; however, their presence in water should no longer

remain the sole source criterion on which the microbiological quality of water is based.

Heterotrophic plate count (HPC)

The first edition of Standard Methods for the Examination of Water and

Wastewater was published in 1905 and included methods for counting bacteria on agar

media, also known as standard plate count bacteria and currently referred to as the

heterotrophic plate count (HPC) bacteria (Rusin et al., 1997).

95

HPC bacteria in drinking water are composed of many transient organisms that

never colonize the distribution system, while other associate organisms are more

opportunistic, being capable of surviving on minimal nutrients, attachment to pipes

sediments, and becoming participants in the development of biofilms (Geldreich, 1996).

HPC bacteria represent the aerobic and facultative anaerobic bacteria that derive their

carbon and energy from organic compounds (Bitton, 1994). The number of recovered

bacteria depends on medium composition, period of incubation (1-7 days), and

temperature of incubation (20-35°C) (Reasoner, 1990). This group includes Gram-

negative bacteria belonging to the following genera: Pseudomonas, Aeromonas,

Mycobacterium, Klebsiella, Flavobacterium, Enterobacter, Citrobacter, Serratia,

Acinetobacter, Proteus, Alcaligenes, and Morallexa. Some members of this group are

opportunistic pathogens (Aeromonas, Flavobacterium, Pseudomonas and

Mycobacterium), but little is known about the effects of high numbers of HPC bacteria on

human health (Bitton, 1994). In drinking water, the number of HPC bacteria may vary

from less than one CFU/mL to more than 104 CFU/mL, and they are influenced mainly

by the temperature, presence of chlorine residual, and level of assimilable organic matter.

HPC bacteria levels should not be high due to the possible decrease in sensitivity of the

coliform test, the masking of health significant bacteria (EPA, 1989), and increased of

risk for exposure, especially to the infirm, the elderly and the young (Geldreich et al.,

1975; Payment et al., 1991).

26

Heterotrophic bacteria densities in most municipal water supplies are generally

below 100 organisms per mL except at static water locations in buildings where densities

are often one or two log higher because of warm ambient temperatures (Geldreich, 1986).

HPC bacteria are not of immediate public health significance, but upon amplification in a

protected habitat become the source of taste and odor complaints or emerge as an

opportunistic pathogen threat (Geldreich, 1996). Reasoner (1989), defined the use of

HPC bacteria in the water treatment process:

• Monitoring the efficiency of water treatment process (primarily disinfection).

• Evaluation of levels of HPC bacteria that may interfere with coliform

monitoring.

• Evaluation of finished water quality during storage and distribution.

• Evaluation of microbial growth on materials used in construction of potable

water treatment and distribution system.

• Evaluation of bacterial re-growth in treated drinking water.

• Evaluation of changes in the population of bacteria during changes in

treatment.

Bacteria in drinking water supplies are generally not physiologically vigorous.

Many of these organisms have been stressed by recent passage of surface water through

the disinfection process or are starved because of minimal nutrient concentrations of

available nutrients in the water (McFeters, 1990). Survivors that reach the distribution

system may adjust to low-nutrient conditions, and slowly colonize a biofilm (Reasoner,

27

1990). Therefore. cultivation requires careful consideration. The use of R2A agar in the

spread plate method is recommended as the medium of choice (Reasoner and Geldreich,

1985). The incubation time is interrelated with selection of temperature. As a general

rule, for the detection of a greater portion of HPC bacteria, it is necessary to use of 28°C

for at least 7 days (Lombardo et al., 1985).

A characteristic of some bacteria found in water supplies is the ability to form

brightly colored non-photosynthetic, non-diffusible pigments (Geldreich, 1996). The

pigmented bacteria may compromise 80-100% of the HPC bacteria not only in

distribution system, but also in bottled water and a variety of attachment devices

including drinking water fountains, water vending machines, ice machines, and point-of-

use-treated waters (Chaidez et al., 1997; Geldreich et al., 1985; Reasoner et al., 1989;

Ridgway and Olson, 1982). Pigmented bacteria appear to adapt well to the distribution

system environment, with at least two populations, the yellow and orange pigmented

groups, being always present in slow-flow and dead-end sections of the pipe network

(Geldreich, 1996).

Flavobacterium

The natural habitat of these bacteria are soil, plants, and water sources.

Flavobacterium form bright-yellow colonies. Flavobacterium can be opportunistic

pathogens and have been implicated in human infections including meningitis (infants),

pneumonia, endocarditis, and septicemia (Dooley et al., 1980). The most important

28

clinical species is F. meninç,Yosepticum (Rusin et al.. 1997). F. meningosepticum has been

recovered from a variety of water system attachment devices including nebulizers,

distilled water systems, sink faucets, and drinking water fountains (Werthamer and

Weiner, 1972). In the United States the cases of neonatal meningitis due to F.

meningosepticum are very rare, only 0.04% per year (Rusin et al., 1997).

A cinetobacter

Acinetobacter is isolated in very low numbers and it is considered to be of low

virulence. Humans carry the bacterium as a normal flora of the skin, being the site source

for most outbreaks of hospitals infections. Acinetobacter is involved in nosocomial and

community-acquired infections. Predisposing factors including malignancy, burns,

immunosuppression, major surgery are necessary to trigger the infection (Rusin et al.,

1997).

Acinetobacter can be isolated from many different water sources. LeChevallier et

al.. (1980) analyzed water from the distribution systems and found that from 5-97% of the

samples contained the bacterium. Acinetobacter has been recovered from bottled water

and water coolers (Gonzalez et al., 1987 and Levesque et al., 1994). Studies on the

infective dose are scarce. The lethal dose to mice is 10 6-10 8 when injected

intraperitoneally (Bergogne-Berezin, 1994).

29

VIoraxella

illoraxella appears as a normal flora of the respiratory tract as well as human skin.

Moraxella species cause upper respiratory infections and eye infections , they can also

produce septicemia and meningitis (Rusin et al., 1997). Moraxella species are present in

water sources, including distribution system, bottled water and water coolers (Payment et

al., 1988; Edberg et al.. 1996 and Levesque et al., 1994). No studies on the infective dose

have been done.

Xanthomonas

Most of the members of this group are non-pathogenic to humans, but may be

harmful to plants. X maltophila is the only member of this group that may cause disease

to humans. The diseases include septicemia, pneumonia, and wound infection.

Antibiotic treatment and cancer therapy are the two risk factors associated with

this bacterium. Xanthomonas occur in the water environment at low numbers. It

compromises 5.7% of the HPC population found in raw surface water samples (Rusin et

al., 1997). The bacterium has been recovered from bottled waters and water coolers

(Manaia et al., 1990 and Levesque et al., 1994). High numbers (10 6-109) are required to

colonize the intestine of mice, but no signs of infection have been described (George et

al., 1989).

30

.11ycobacteria

It is common knowledge that some Mycobacteria are agents of chronic diseases in

humans and other animals, but it is not always recognized that most Mycobacteria are

free-living saprophytes which form part of the balanced microbial community in natural

habitats such as water or soils (Kubica and Wayne, 1984).

Mycobacteria are aerobic acid-alcohol that usually form slightly curved or straight

rods with a size range between 0.2-0.6 and 1.0-10 tm. Occasionally extensively

branched filaments may occur. The organisms are non-motile, and do not form

endospores, spores or capsules. Mycobacteria have cell walls with a high lipid content

that includes waxes (mycolic acids). They are not readily stained by the Gram method

but they are considered Gram-positive organisms (Kubica and Wayne, 1984).

Growth rates of Mycobacteria are slow with generation times varying by species

and ranging from 2 to >20h. Visible colonies may be produced after 2d to 8 weeks of

incubation under optimal conditions (depending the species). Optimal temperatures vary

widely among species, ranging from 30 to 45°C.

The most important pathogens are non-pigmented, including M tuberculosis, M

(MUM complex (avium and intracellulare), M asiaticum, and M. malmoense (Nolte and

Metchock, 1995).

31

The Mycobacterium avium complex (MAC) are composed of a serological group

consisting of 28 serovars of two species, M avium and M intracellulare. MAC are

ubiquitous organisms that cause disseminated disease, pulmonary disease, and cervical

lymphadenitis in the immunocompromised and to a lesser extent in normal individuals

(Korvick, 1996; Collins et al., 1984). Disseminated MAC infection is a progressive

illness, characterized by intermittent fever, sweating, weakness, anorexia, and weight

loss. Pulmonary disease due to MAC presents symptoms similar to those of tuberculosis,

with productive cough, fatigue, fever, weight loss, and night sweats. Pulmonary disease

caused by MAC was to be found as common as pulmonary tuberculosis in many areas of

the United States (Rosenweig and Schlueter, 1981). Cervical lymphadenitis usually

affects children less than 12 years of age, and is the leading cause of lymphadenitis in this

age group.

MAC becomes particularly invasive in susceptible individuals with predisposing

factors (elderly, newborns, burn cases, dialysis, AIDS patients, and individuals receiving

organ transplants). Furthermore, it has been recently shown that MAC infections also

affect normal individuals without predisposing factors (Glover et al., 1994). Studies

conducted by Prince et al., (1989) showed that M avium infection can also cause

osteomyelitis and septic arthritis in people with no known predisposing factors (Jones et

al., 1995; Prince et al., 1989). In the United States, many infections are asymptomatic

and occur early in life; 12% of the population has been infected by MAC (Von Reyn et

32

al., 1993). However, disease by MAC can be lethal and is difficult treat because it is

resistant to many antimycobacterial agents (Rusin et al., 1997).

M. avium has been frequently associated with disease in humans (Collins et al.,

1984). M avium is a waterborne opportunistic pathogen (Du Moulin et al., 1985).

However, the precise mode of transmission remains undetermined. Infection with M

avium is thought to occur from colonization of the gastrointestinal or respiratory tract

(Chin et al., 1994).

Several methods can be used to culture M avium from water sources. However,

the most popular and effective is membrane filtration, using 7H11 selective agar (Difco

Laboratories). Glover et al., (1994) successfully recovered MAC from Los Angeles tap

water following this method. Du Moulin and Stottmeier, (1978) concluded that treating

water samples with cetylpyridinium chloride (CPC) for 24 h reduced the contamination

and increased the recovery of Mycobacteria.

Mycobacteria are difficult to stain. Although they are classified as Gram-

positives organisms, the large amount of lipids present in their cell walls makes it

difficult to the dyes used in the Gram stain (Nolte and Metchock, 1995). Special staining

procedure is used to promote the uptake of dye in Mycobacteria (Auramine-O-acridine

orange, Ziehl-Neelsen and Kinyoun). Serology has been previously applied to the

identification of Mvcobacteria recovered from tap water. Graham et al., (1988) identified

M avium serovar 4 from one laboratory and four hospital wards, using the Shaeffer

seroagglutination method. PCR has been widely used for the differentiation of M. avium

33

and M intracellulare from clinical specimens (D,vadyk et al., 1994: Yamamoto et al.,

1993).

MAC can be isolated from numerous sources in the environment, including water,

aerosols, and soil (Kirschner et al., 1992). Evidence suggests a higher incidence of MAC

infections due to M avium than M intracellulare. For example, Korvick, (1996)

mentioned that over 98% of cases of MAC infections in AIDS patients are caused by M.

avium serotypes, particularly 1, 4, and 8, rather than M. intracellulare. Glover et al.,

(1994) found a high incidence of M. avium serotypes 1 and 8 from hospital water in the

Los Angeles area. These results coincide with those reported by Yakrus and Good,

(1990) in which it was found that serotype 8 was the predominant serotype among

clinical isolates in Los Angeles area. The environmental reservoirs and route of

acquisition of the organism are most likely water and aerosols. Hospital water systems

often harbor MAC (Horsburgh, 1991), and may be a source of nosocomial infections

(Von Reyn et al., 1994).

Glover et al., (1994) examined Los Angeles tap water and found M avium in 70%

of hospitals, 9% of dwellings, and 15% of reservoirs. Chaidez and Gerba, (1998) found

that 43% of the tap water analyzed had acid-fast organisms. It has also been found in up

to 50% of municipal and private drinking water samples at concentrations of 0.01-5.2

CFU/mL (Rusin et al., 1997). Collins et al., (1984) mentioned that MAC is particularly

resistant to chlorination of water. Surucu and Haas, (1976) and Goslee and Wolinsky,

(1976) reported that M. aviuin was resistant to the usual chlorine doses that inactivate

34

conforms and viruses. Mycobacteria were also more resistant than E. colt to inactivation

by chloramines and by ozone (Severin, 1976; Goslee and Wolinsky, 1976). The

resistance of those microorganisms to chlorine disinfection in water may be a particularly

important factor in the occurrence of MAC infections (Geldreich, 1996). MAC have been

isolated from water samples with temperatures ranging from 2 to 57.2°C, and residual

chlorine concentrations of 0.02 to 0.3 mg/L (Du Moulin et al., 1988; George et al., 1980).

Little information has been obtained about the presence of Mycobacteria in

bottled water. Caroli et al., (1985) reported the presence of Mycobacteria in 9 samples of

84 bottled water samples tested, but, none were M avium. The acid-fast bacilli isolated

were M gordonae, M flavescens, M phlei, and Nocardia sp. Similarly, Chaidez and

Gerba, (1998) identified acid-fast organisms from 13% of the bottled water samples

analyzed. Holtzman et al., (1997) examined 20 brands of bottled water and no acid-fast

organisms were found. The reason for this variation could be that not all bottled water

are processed by the same method and some could contain significant contaminants.

However further studies are required to elucidate such variation.

Point-of-Use (POU)-devices attached to building plumbing systems may even

enhance the growth of Mycobacteria (Geldreich, 1996). POU-devices might be expected

to remove Mycobacteria by physical methods, but it is also possible that the organism

might colonize and multiply in the filter's biofilm and, thereafter, appear in the product

water (Collins et al., 1984). Chaidez and Gerba, (1998) found acid-fast organisms in

33.3% of POU product water.

35

Chaidez et al., (1997) conducted a study on water obtained from vending

machines and found acid-fast organisms in the product water. Water vending machines

are connected to an approved local water supply system and processed the water by

reverse osmosis, carbon filters, and ultraviolet light (Lynkins et al., 1992). Studies have

shown that Mycobacteria are among the first organisms to colonize reverse osmosis

membranes (Geldreich, 1996). Ridgway et al., (1985) suggested that the hydrophobic

cell surface of Mycobacteria may play an important role in the bacterial adhesion to

reverse osmosis membranes.

It has been shown that tap water can harbor M. avium and that it may be the

vehicle responsible for transmission to AIDS patients, and to non-AIDS patients with

predisposing factors (Glover et al., 1994; Von Reyn et al., 1994; Mansfield and Lackner,

1997). Von Reyn et al., (1993) isolated M avium from 32% of the samples from water

supply systems in the United States. Peters et al., (1995) found M avium in 2% of the

samples obtained from tap water in Germany. Mycobacteria are important components

of biofilms in tap water distribution systems, being found in 90% of the biofilm samples

with densities ranging from 10 3 to 104 CFU/cm2 (Schulze-Robbecke et al.,

1992).

The infective dose of M. avium is quite high according to laboratory studies by

Bermudez et al., (1992). Five oral doses of 10 8 CFU of MAC resulted in bacteremia in

45% of the animals with 26% mortality. The use of alcohol (ingestion of ethanol) aided

to a significant increase the number of M avium recovered from the liver, spleen, and

36

appendix (Bermudez et al., 1992). Studies in human volunteers are scarce and no

publications are available that provide the infectious dose in humans.

Plesiomonas shigelloides

Plesiomonas shigelloides, a member of the family Vibronaceae, is a facultative

anaerobic, Gram-negative, nonspore-forming rod that measure 0.8 to 1.0 by 3.0 1AM and

can occur singly, in pairs, or in short chains (Janda et al., 1995). They are usually motile

by two or five lophotrichous polar flagella and are catalase, oxidase, and nitrate positive.

They ferment D-glucose and a few other carbohydrates but without production of gas

(Freund et al., 1988).

Infections attributed to P. shigelloides are almost exclusively restricted to two

clinical settings: The most common presentation is watery diarrheal illness often found

in individuals with a history of freshwater contact, seafood consumption, exposure to

amphibia or reptiles, or travel to developing countries (Janda et al., 1995); The

gastroenteritis is usually a mild self-limiting disease with fever, chills, abdominal pain,

nausea, or vomiting (Holmberg et al., 1986). The peak season of P. shigelloides

associated diarrheal disease appear to be similar to that of Aeromonas-associated diarrhea

(warner months). The second well-recognized syndrome associated with P. shigelloides

is septicemia, often accompanied by meningitis (Janda et al., 1995). Most published

cases of Plesiomonas-meningitis have occurred in newborn babies whose deliveries

which been complicated by various medical conditions, including prolonged rupture of

37

the mother's membranes. The fatality rate in such instances approaches 70% (Janda et

al., 1995).

The link between P. shigelloides and diarrhea appears to be rather weak.

Liesenfeld et al., (1993) found P. shigelloides in stool samples in only 0.014% of

diarrheal patients in Germany. This frequency is not much higher than the carrier rate of

0.0078% described by Arai et al., (1980) in healthy humans. Rautelin et al., (1995)

isolated 20/13,027 (0.15%) from stool samples in Helsinki. All except 2/20

Plesiomonas-positive patients had diarrhea; 13 patients had onset of illness after foreign

travel and 5 patients had chronic diarrhea with symptoms lasting 2 months. Morbidity

and Mortality Weekly Report (MMWR) described an outbreak in associated with

contaminated water supply in Livignstone County, NY. Thirty persons became ill

(diarrhea), due to the presence of P. shigelloides and Salmonella in the drinking water

source (MMWR, 1998).

P. shigelloides has been isolated using routine clinical procedures for the

Enterobactericeae which includes MacConkeys, Salmonella-Shigella, Deoxycholate,

Hektoen, and XLD agars (Cooper and Brown, 1968). The organism grows well on

enteric agars and can easily be mistaken as a member of the Enterobactericeae (Miller

and Koburger, 1985). There are several selective media proposed for the isolation of the

bacterium (Jeppesen. 1995). However, for routine analysis of environmental and food

samples inositol brilliant green bile salts (IBB) and Plesiomonas (PL) agar are

recommended for best results (Jeppesen, 1995).

38

P. shigelloides is widespread in nature and has been isolated in many different

countries (Miller and Koburger, 1985). Hernandez and Rodriguez de Garcia, (1997)

found a prevalence of 59% in surface water from Venezuela during the warmer months.

Tsukamoto et al., (1978) surveyed water samples in Japan and found that P. shigelloides

was present in 60% of pond water and 35.7% of river water samples. In Nigeria, well

waters were sampled for the presence of P. shigelloides and 7.4% of the wells were

positive (Kwaga et al., 1988). Medema and Schets, (1993) found P. shigelloides in 30 of

42 samples in fresh water in the Netherlands. P. shigelloides showed a positive

correlation with the presence of fecal pollution (E. cou). Others authors have reported the

occurrence of P. shigelloides in surface water around the world including Australia

(Cooper and Brown, 1968), and the United States (Rutala et al., 1982). P. shigelloides is

considered as a waterborne opportunistic pathogen (Tsukamoto, 1978). It is found most

often in fresh surface waters, but there are some reports of its presence in seawater

(Cabelli, 1978). Chaidez and Gerba, (1998) analyzed various water sources including

bottled water, tap water, and point-of-use treated water and found that P. shigelloides was

absent from those sources.

Animal and human feeding studies have been conducted for P. shigelloides.

Herrington et al., (1987), conducted human feeding studies using young healthy humans.

None of the volunteers developed diarrhea or any other disease symptoms. Pretreatment

with ampicillin did not enhance colonization. In rabbit feeding studies, it was found that

only one rabbit developed diarrhea, but when re-tested, none of the five strains could

39

induce diarrhea. Oral doses of P. shigelloides produced disease in a semi-consistent

manner only when administered to gnotobiotic piglets. Oral doses of 109 CFU produced

diarrhea within 6 days in 3 of 3 piglets while a higher dose of 10 10 CFU induced shock in

2 of 2 piglets (Rusin et al., 1997).

Pseudomonas aerugin osa

Pseudomonas species are Gram-negative, straight or slightly curved rods, motile

with polar flagella (Hoadley, 1977). They are oxidase and catalase-positive. The size

ranges between 1 to 5 im long and 0.5 to 1.0 p.m wide. They are obligate aerobes, with a

respiratory metabolism in which oxygen is the terminal electron acceptor (Gilligan,

1995). Some isolates may grow under anaerobic conditions if nitrate is present (Hoadley,

1977). The nutritional versatility of Pseudomonas aid to utilize carbohydrates, alcohols,

and aminoacids as a carbon source (Gilligan, 1995). Certain species have distinctive

colony morphologies or pigmentation. Some species can multiply at 4°C, but most are

mesophilic, with optimal growth temperatures between 30 and 37°C (Palleroni, 1984). P.

aeruginosa is capable of growing at 41°C (Hoadley, 1977).

The genus Pseudomonas contains several well-recognized pathogens, including

Pseudomonas mallei, P. pseudomallei, P. cepacia, and P. aeruginosa. The Pseudomonas

most commonly associated with food and water is P. aeruginosa (Stiles, 1989). P.

40

aeruginosa represents special hazard in patients requiring instrumentation (i.e. catheter,

aerators. respiratory therapy equipment) which can introduce bacteria to susceptible

tissues, also in lesions containing necrotic tissues (burns) or fluids (urine) in which the

organism can grow rapidly, from small inocula to potentially invasive numbers (Fick,

1993). In addition to its nosocomial sources, the bacterium may be found in swimming

pools, hot tubs, and contact lens solutions (Gilligan, 1995).

It rarely causes problems in healthy persons, but it has been implicated in

epidemic diarrhea in infants (Hunter and Ensign, 1947). In hospitalized and

immunocompromised patients the organism has the ability to produce serious diseases,

including endocarditis, meningitis, pneumonia, and septicemia (Bodey et al., 1983). The

infection causes a sudden onset of fever, abdominal distention, and pain that gradually

increases (Stiles, 1989). It also causes nosocomial urinary tract infections, wound

infections, and bacteremia (Favero et al., 1971). Other serious infections involves eyes,

ear, skin and nose (Geldreich, 1996). The CDC reported an outbreak in which

individuals who were bathed in a hot tub developed folliculitis and otitis due to P.

aeruginosa (CDC, 1982).

P. aeruginosa is a common pathogen in the U.S., this bacterium alone has

accumulated some impressive figures: Up to two thirds of hospitalized patients,

particularly the critically ill, become colonized with P. aerugin osa, this often presages

more invasive disease. Two thirds of pneumonias are hospital-acquired, and P.

aeruginosa is listed as the leading cause of nosocomial respiratory tract infection (Fick,

41

1993). The mortality of nosocomial pneumonias remains as high as 70% (Celis et al.,

1988). Cystic fibrosis (CF) is the most common lethal genetic disease of Caucasians, and

chronic Pseudonionas respiratory infections complicate 90% of CF patients leading to

respiratory failure and death (Gilligan, 1991).

In 1946, an epidemic of acute P. aeruginosa gastroenteritis was reported in which

milk was implicated. There were 409 cases with symptoms of diarrhea, cramps, nausea,

and vomiting. Symptoms were more severe in infants and children, and there were nine

deaths in infants (Hunter and Ensign, 1947). In 1974, an outbreak of P. aeruginosa

foodborne illness was reported in a school; 7.7% of the infected developed weakness,

dizziness, and arthralgia, but without diarrhea (Meitert et al., 1984).

Bert et al., (1998) reported an outbreak of P. aeruginosa in hospital tap water;

nine patients were diagnosed with urinary tract infections, pneumonia and sinusitis. An

outbreak of funicular sepsis (infection of the umbilical cord resulting in bacteremia) in 10

newborns in a hospital of Germany was reported with P. aerugionsa the organism

isolated (Weber et al., 1971). There is some evidence that such infections are caused by

consumption of contaminated water. However, the information in this study was scarce

and the evidence was circumstantial (Rusin et al., 1997).

The presence of P. aeruginosa in surface waters has been associated with sewage

discharges (Alonso et al., 1989). Counts of the bacterium in sewage often exceed 10 5 per

100 mL (Hoadley, 1977). Its presence may be considered as an indicator of such

pollution (Warburton et al., 1994; Geldreich, 1986). Up to 15% of the normal human

42

population are fecal carriers of P. aeruginosa. The isolation rate from feces of

hospitalized patients is 20% or greater, and recovery rates occur in patients with longer

times of hospitalization (Sloodley and Thom, 1970). Colonized patients are at greater

risk of infection (Bodey et al., 1983). The most important mode of transmission is

person-to-person, but it can also occur from common environmental sources, including

water, soil, and foods (Shooter et al., 1969).

Several methods have been employed to enumerate or confirm the presence of this

bacterium in waters. However, the membrane filtration technique has been the most

widely applied for the isolation of P. aeruginosa (Hoadley, 1977). P. aeruginosa isolates

are easily recognized on primary isolation media on the basis of colony morphology,

production of pigments and grape-like odor (Gilligan, 1995). Culture media facilitate the

isolation from environmental sources. Generally, bacterial growth showing blue-green

fluorescence after 24h at 37°C on cetrimide is virtually always P. aeruginosa (Hoadley,

1977). The bacterium can be identified on the basis of a positive oxidase test, and triple

sugar iron (TSI) agar reaction (Gilligan, 1995). Most of the commercial kit systems can

identify P. aeruginosa (Gilligan, 1995) However, such kits are expensive and the results

are not 100% accurate (Gilligan, 1995).

P. aeruginosa is able to multiply in low nutrient water and can therefore, colonize

a variety of drinking water sources (bottled water, water vending machines, mineral

water) (Gonzalez et al., 1987; Chaidez et al.. 1997; and Manaia et al., 1990).

43

P. aeruginosa can be recovered in low numbers in drinking water (Hardalo and

Edberg, 1997). Hoadley, (19 7 ') found that the occurrence of this microorganism is

usually low in water, especially in treated (chlorinated) drinking water (3.7%). Similarly,

Edberg et al., (1996) found that 3% of drinking water samples (bottled water, water

coolers units, and tap water) contained P. aeruginosa. Five isolates from the green

pigmented colonies were satisfactorily identified as strains of P. aeruginosa from bottled

water from Nigeria (Ogan, 1992). Manaia et al., (1990) identified P. aeruginosa in 29%

of Portuguese bottled water and Fewtrell et al., (1997) found the bacterium in 1.2% of

bottled water from England. Chaidez and Gerba (1998) found P. aeruginosa in 8.3% of

bottled water collected from Costa Rica, Mexico, U.S., Japan, Korea, Greece, and

Panama. On the other hand, Hernandez-Duquino and Rosenberg, (1987) did not isolate

P. aeruginosa from both German and American bottled waters nor did Hunter, (1994) in

bottled water obtained from England.

Levesque et al., (1994) did not detect P. aeruginosa in water cooler samples. In

contrast, P. aeruginosa was found in 23% of the water vending machines (WVM) water

samples (Chaidez et al., 1997). Point-of-Use (POU) home devices may be very effective

initially for some specific contaminants, however their usefulness may become limited.

Chaidez and Gerba, (1998) found P. aeruginosa in POU devices during a six weeks

sampling. This indicate that colonization is not transient and may be due to the ability of

P. aeruginosa to form biofilms in water pipes (Brown and Gauthier, 1993; Geldreich,

1996). P. aeruginosa has the ability to multiply in process units such as sand and

44

activated carbon filters (Grabow et al., 1980). P. aeruginosa use trace nutrients present

either in the water or attached to the carbon filters to grow (TamaE,rini and Gonzalez,

1997). P. aeruginosa can be also recovered often in high numbers, in common food,

especially vegetables. Tomato salad may contain 5x10 3 organisms per serving size

(Kominos et al., 1972).

It is apparent that P. aerugin osa can cause both foodborne and waterborne illness

(Meitert et al., 1984; Bert et al., 1998). Contaminated water supply might be a more

significant source of infection than foods, especially for infants (Weber et al., 1971). The

presence of P. aeruginosa in water does not justify specific attention for healthy people,

but in the hospital environment, especially for immunocompromised individuals, special

precautions would be justified (Stiles, 1989). Although acquired immunodeficiency

syndrome (AIDS) patients are immunocompromised, P. aeruginosa is not predominant

opportunistic pathogen for this population (Rusin et al., 1997).

Larger numbers are required to colonize the respiratory tract of healthy

individuals. The infective dose for P. aeruginosa is quite high according to Buck and

Cooke, (1969). They found that oral doses of >1.5x106 CFU are required to colonize

healthy volunteers. Volunteers with ampicillin treatment excrete 2x10 8 of P. aeruginosa

for up to 14 days. The selective pressure of antibiotic treatment could extend the

incidence and time of colonization. Excretion was limited to 6 days by volunteers not

taking ampicillin. Colonization by P. aeruginosa in hospitalized individuals may occur

45

by ingesting less than 10 3 to 104 of the organism (Pollack, 1990). Although P.

aerugin osa is sometimes found as a normal flora in man, the prevalence of colonization

in healthy adults outside the hospital are low (2.6-24%), and the probability of infection

in healthy individuals appears to be very low (Rusin et al., 1997). Hospitalization leads

to greatly increased rates of carriage, especially in patients with serious burns.

Colonization of a hospital patient often presages disease (Rusin et al., 1997).

Dissertation Format

The research presented in the appendices of this dissertation consists of four

related studies designed and undertaken by the candidate: 1) The microbiological quality

of water from vending machines; 2) A survey of the microbial quality of water from

storage tanks; 3) Microbiological comparison of the water quality of point-of-use treated

water, tap water, and bottled water; and 4) Aeromonas hydrophila and Pseudomonas

aeruginosa in drinking water from various sources: a risk assessment.

Dr. Charles P. Gerba is a co-author on all of the papers and served as an advisor to

the candidate's research. The Department of Nutritional Sciences requires that each

candidate submit their original research to peer-reviewed scientific journals for

publication. By using this dissertation format, the candidate's research is presented as

four separate papers.

46

CHAPTER 2

PRESENT STUDY

The methods. results and conclusions of this study are presented in the articles

appended to this dissertation. The following is a summary of the most important

findings.

The first article is a study conducted on the microbiological quality of drinking

water from vending machines. The information obtained from this study can be used to

identify the bacterial quality of the water dispensed from such machines. Opportunistic

bacterial pathogens data were used for the elaboration of microbial risk assessment.

The second article presents a survey of the microbial drinking water quality of

storage tanks. The information obtained here helped to inform and persuade the

community to the importance of maintenance of water storage tanks (WST). Also,

Pseudomonas aerugmosa data obtained in this study was used to conduct a microbial

risks assessment.

The third article is an extensive study on the bacterial quality of the drinking

water obtained from point-of-use (POU) devices, tap water with POU-connection, and

bottled water. The information was utilized to perform the microbial risk assessment.

A formal microbial risk assessment was conducted using the data obtained from

the three articles appended to this dissertation. Aeromonas hydrophila and P. aeruginosa

47

data were selected to estimate risk associated with exposure to opportunistic bacterial

pathogens in drinking water. The results and the conclusions are presented in the fourth

article. Estimated annual risks of colonization for drinking water from selected sources

were determined to be as high as 10 and 10' for A. hydrophda and P. aeruginosa,

respectively.

APPENDIX 1:

MICROBIOLOGICAL QUALITY OF WATER VENDING MACHINES

Cristobal Chaidez Quiroz 1 , Pat Rusin', Jaime Naranjo', and Charles P. Gerba* I '"

Department of Nutritional Sciences', Department of Soil, Water, and EnvironmentalSciences',

The University of Arizona, Tucson, AZ., 85721.

48

*Corresponding author

Abstract

Consumption of water from vending machines has recently increased in the

United States. However, studies describing the bacteriological quality of these machines

are scarce. In this study, bacteriological analyses were performed on samples from 30

water vending machines (WVM), three of which were sampled weekly for three weeks.

Bacteriological analyses were also conducted on the nozzle dispensers and the drains of

fifteen WVM. Heterotrophic bacteria, total and fecal coliforms, and Pseudomonas

aeruginosa were enumerated. Physico-chemical parameters such as pH, temperature,

turbidity, and residual chlorine were also examined. P. aeruginosa was found in 23% of

the water samples and coliform bacteria in 20%. Heterotrophic plate count (HPC)

bacteria were found in all samples and 73% had numbers greater than 500 colony forming

units (CFU)/mL. The HPC bacteria ranged from 9 to 48,000 CFU/sampled area on the

dispensing nozzle. Total coliforms and Escherichia coli were detected in the drain

samples with HPC bacterial concentrations from 1,000 to 56,000 CFU/sampled area. No

significant correlation was found between the physico-chemical and bacteriological

parameters. Regular cleaning and maintenance procedures should be implemented to

reduce bacterial concentrations in the product water.

Keywords: Water vending machines; bacteria; coliform; Pseudomonas aeruginosa;

heterotrophic bacteria.

49

Introduction

The water vending machine (WVM) industry has experienced an increase in growth over

the past few years (Lynkins et al., 1992). Likewise public awareness of water quality has

increased significantly. People use WVM as an alternative drinking water supply

because they believe it tastes better, is safer, and free of contaminants. There are many

factors that lead consumers to obtain drinking water from sources other than the tap.

Health risks, both real and perceived, are often the main reason why people use WVM as

a source of drinking water. Tucson, Arizona has large numbers of WVM located at high

traffic locations such as supermarkets, drugstores, and variety stores. Water is delivered,

by these machines, in volumes (3.7 liters) to the consumer's containers. The Food and

Drug Administration (FDA) considers WVM as food vendors and regulates them

according to the Food and Beverage Model Ordinance. However, the state of Arizona

has no specific regulations regarding WVM unless a complaint is received. Since water

from vending machines must come from an approved municipal source, the assumption is

that the water should meet Environmental Protection Agency (EPA) standards (Gelt,

1996). According to Lynkins et al., (1992) WVMs follow a typical POU/POE (point-of-

use/point-of-entry) technology that require little technical service. The carbon filters are

changed as required, the ultraviolet lamp is replaced annually, and the reverse osmosis

(RO) membrane is replaced every 2-3 years. The WVMs surveyed in this study are

50

51

connected to an approved local water supply system and processed the water by reverse

osmosis, carbon filtration, and ultraviolet light.

Drinking water contains heterotrophic bacteria including Acinetobacter,

Flavobacterium, Moraxella, Çvtophoga, Achromobacter, Pseudomonas, and Alcaligenes

spp. (McFeters, 1990). Some of these genera include opportunistic pathogens such as

Pseudomonas aeruginosa, Alcaligenes faecalis, Acinetobacter baumanii, and

Flavobacterium meningosepticum species that can cause nosocomial and community-

acquired infections (Bitton, 1994). Of these, P. aeruginosa is the pathogen most common

associated with documented illnesses (Rusin et al., 1997). Favero et al., (1971) described

P. aeruginosa as an important nosocomial pathogen that is associated with hospital

equipment contamination. Contaminated foods and water supplies have also been

implicated as routes of transmission in the hospital environment (Shooter et al., 1969),

although never conclusively shown to be a source. P. aeruginosa is responsible for 9%

of all nosocomial infections (Hoadley, 1977), being the leading cause of nosocomial

respiratory tract infections. It also causes nosocomial urinary tract infections, wound

infections, and bacteremia (Murray et al., 1995).

Other factors affecting the microbial quality of WVM product water are the

carbon filters and RO membranes. While carbon filters may be very effective initially

removing specific contaminants, studies have shown (Taylor et al., 1979) that

heterotrophic plate count (HPC) bacteria can colonize the carbon filters in a short period,

contributing high levels of bacteria to the final product water. According to Taylor et al.,

52

(1979) carbon filters in point-of-use devices resulted in a one to two log increase of

bacterial concentrations over those found in public tap water samples. Wallis et al.,

(1974) warned against the incorporation of charcoal filters into water systems after

observing that bacterial densities increased in treated (chlorinated) waters following an

overnight period of non-use. In the WVM, initial counts of HPC are low, but can

increase rapidly if there is inadequate cleaning of the carbon filters, RO-membrane, or

ultraviolet lamp (Hunter, 1994). Payment et al., (1991) found that water treated by

reverse osmosis units may produce an increase in the HPC bacterial counts of

approximately 1,000 to 10,000 CFU/mL in the product water.

The microbial quality of water produced by point-of-use devices is extremely

variable, being a function of the contamination of the water, the service life of the filter,

water temperature, and the frequency of static water conditions (Geldreich, 1989). In

England, Hunter and Burge, (1986) conducted a study on the bacteriological quality of

water and soft drinks from automatic vending machines and showed that 44% of 25

drinking water samples and 6% of the soft drinks contained coliforms and 84% and 39%

had HPC bacterial levels greater than 1,000 organisms per mL, respectively. Automatic

vending machines are described as units that dispense water or soft drinks into disposable

cups. Some machines contain water filtration system to improve the taste of the product

water (Hunter, 1992). The presence of coliforms should not be ignored because these

bacteria are consistently present in high numbers in feces (human and wildlife) and are a

significant warning of a potential risk for pathogen exposure from a contaminated water

53

source (Geldreich. 1996).

To our knowledge no previous bacteriological studies have been conducted on

water vending machines. The purpose of the study was to determine the occurrence P.

aeruginosa, total and fecal coliforms, and HPC bacteria in WVM product water, and their

correlation to each other in the WVM product water. Others factors such as pH,

temperature, chlorine, and turbidity were measured and correlated with the presence these

of bacteria, and are also reported here.

Materials and Methods

Thirty water samples were collected from WVM, three of which were sampled for

three consecutive weeks. Locations throughout the Tucson area were selected with the

majority of these outside (against the entrance side of the building) of supermarkets.

Water samples were collected in sterile one-liter plastic bottles containing 2 mL of sterile

10% sodium thiosulfate to neutralize any residual chlorine. To ensure that the samples

were representative of the water consumed, the outer surface of the WVMs' nozzle were

not sanitized. The samples were transported refrigerated to the laboratory for processing

within one hour. Biofilms were observed in some of the machines outer nozzle and drain.

Therefore, biofilm samples were collected from the nozzle and drain of fifteen WVM.

Sterile cotton swabs were used to collect biofilm samples and placed in 2 mL of sterile

universal neutralizer (polysorbate 80, 0.3%, sodium thiosulfate, 0.5%. L-histidine, 0.1%,

and phosphate buffer, 0.25N) for transport to the laboratory, according to The Official

54

Methods of Analysis of the Association of Official Analytical Chemist (1984).

P. aeruginosa, total and fecal coliforms in water samples were enumerated by the

membrane filtration method, according to the Standard Methods for the Examination of

Water and Wastewater (1992) (0.45 um, Gelman Science, Ann Arbor, MI), using

Pseudomonas agar base (Oxoid, Hampshire, England), m-Endo, and m-FC agar (Difco

laboratories, Detroit, MI), respectively. P. aeruginosa samples were incubated a 37°C for

24 h; total and fecal coliforms were incubated at 37°C and 44.5°C for 24 h, respectively.

HPC bacteria were enumerated by the spread plate method using R2A agar (Difco

laboratories), and incubation at 25°C for 7 days, according to the Standard Methods for

the Examination of Water and Wastewater (1992). Presumptive identification of P.

aerugin osa was based on the production of green pigmentation, a positive oxidase test,

the ability to grow at 42°C and 4°C, and the ability to oxidize dextrose. Confirmation was

conducted using the API2OE system (Biomurex Vitek, Hazelwood, MO). Confirmation

of total and fecal coliform bacteria were conducted using the Colilert test (IDEXX,

Westbrook, ME).

Biofilm swab samples from drain and nozzle sites were placed in sterile universal

neutralizer, vortexed and immediately plated on m-Endo, m-FC, and R2A agar for the

detection of total coliforms, fecal coliforms, and HPC bacteria, respectively. The Colilert

test was utilized to assess the presence of total coliforms and Escherichia coll. The

approximate sampled area for nozzle and drain surfaces was 3 and 20 cm 2 , respectively.

Factors such as pH, temperature, turbidity, and residual chlorine that may

55

influence the quality of the water dispensed by the VvvVM were also measured. These

procedures were performed according to Standard Methods for the Examination of Water

and Wastewater (1992).

Finally, the relationship between physico-chemical and bacteriological parameters

were analyzed by regression analysis using STAT101 computer software (1993).

Results

Microbiological analyses

Bacteriological results of water samples from 30 WVM are shown in Table 1. P.

aeruginosa was found in 23% of the samples and total coliforms in 20%. HPC bacteria

were found in all samples with 73% having numbers greater than 500 organisms per mL.

The percentage of samples that were positive are shown in Figure 1. During three

consecutive weeks, three WVM were concurrently tested. P. aeruginosa was present in

most of the samples except one (Table 2). Total coliforms were found in only one

sample. Concentrations of HPC bacteria varied with 50% of the samples containing

greater than 500 CFU/mL. The results from the nozzle swab test are shown in Table 3.

Total coliforms were not detected using either m-Endo agar or the Colilert test in the

nozzle samples. HPC concentrations varied from 9 to 48,000 CFU/ sampled area.

Total and fecal coliforms were detected in the drain biofilms using Colilert test

56

and m-Endo agar in ten of the fifteen samples (Table 4). HPC counts were found to

range from 1,000 to 56.000 CFU/sampled area.

Physical and Chemical analyses

The maximum and minimum levels of physico-chemical parameters of 30 WVM

are shown in Table 5. The temperature of the WVM water samples varied from 25 to

32°C. This variation may have been due to location of the WVM as some were exposed

directly to sunlight whereas others were located in shaded areas and the product water

remained at a lower temperature. The pH range was between 6.0 and 7.4, with no

obvious explanation for such variation. Turbidity varied from 0.15 to 3.00 NTU. Such

variation might be explained by the presence of dust and biofilm in some of the WVMs'

nozzles; however, further studies are required to elucidate the source of turbidity in the

water. Finally, total and free chlorine was absent in all of the machines except one,

which contained 0.07 mg/L of free chlorine.

Analysis by regression method

Regression analysis revealed that there were no statistical correlation between

physico-chemical and microbiological parameters (p<0.05) (STAT computer software

1993).

P. aeruginosa Total coliforms HPC bacteria

20

10

57

FIGURE 1. Percent of samples positives for Pseudoznonas aerzzginosa, total coliformsand heterotrophic plate count (HPC) bacteria greater than 500 CFU/mL

TABLE 1. Bacteriological quality of water from vending machines

Vending machine P. aeruginosce Total coliformsb HPC bacteria'

1 0 0 4,0001 25 0 6,5003 0 0 3104 0 0 1,3005 7 0 4,0006 0 0 1,3007 0 2 2,1008 0 0 1,3009 0 3 1,80010 0 5 1,80011 0 3 2012 0 0 L00013 0 0 3,00014 0 0 2,00015 0 0 13016 21 0 8,60017 0 0 20018 0 0 3,00019 0 0 80/0 0 0 81021 0 3 6,00022 22 0 5,20023 0 0 8,00024 2 0 <1025 79 0 6,00026 0 1 35027 0 0 1,00028 0 0 2029 0 0 5,20030 2 0 1,000

' CFU/500mLb CFU/100mL

CFU/mL

58

TABLE 2. Weekly sampling of three water vending machines

Vendingmachine

Weekl Weelc2 Week3

P.aer. TC HPC P. aer. IC HPC P. aer TC HPCA 0 5 300 0 0 1,800 0 0 60B 1 0 40 20 0 8,600 1 0 300C 6 0 430 16 0 5,200 3 0 1,200

P. aer. = P. aeruginosa, CFU/500 m.LIC = Total coliforms, CFU/100 mLHPC = Heterotrophic plate count, CFU/mL

TABLE 3. Nozzle swab test results of fifteen water vending machines

NOZZLE Total coliforma Total coliformb Fecal coliformb

(m-Endo agar) (Colilert test) (Colilert test)HPC bacteria'

(R2A agar)

1 <10 102 <10 4,6003 <10 94 <10 4,6505 <10 5,0006 <10 8007 <10 46,4008 <10 1009 <10 36,00010 <10 48,00011 <10 10,00012 <10 1,00013 <10 12,00014 <10 6,00015 <10 1,000

CFU/sampled areapresence or absence

( - ) = Non-detected

59

60

TABLE 4. Drain swab test results of fifteen water vending machines

DRAIN Total conform'(m-Endo agar)

Total conform'(Colilert test)

Fecal conform'(Colilert test)

HPC bacteria'(R2A agar)

1 10 — — 3,9002 500 4_ 4,2503 20 + 2,8004 <10 _ 4,3505 6,000 + 4,6506 120 + 56,0007 2,800 + + 36,0008 <10 6,0009 8,000 + 480,00010 3,000 + 400,00011 1.000 — 480,00012 10 + + 32,00013 <10 - 16,00014 <10 1,60015 <10 1,000

CFU/sampled areab presence or absence

CFU/sampled area( + ) = Detected( - ) = Non-detected

61

TABLE 5. Maximum and minimum values for the physico-chemical parameters inanalyzed water from vending machines

Conditions

Maximum MinimumResidual chlorine (mg/L) 0.07 0pH 7.4 6.0Temperature (°C) 32 25Turbidity (NTU) 3.00 0.08

Discussion

The WVM tested in the Tucson area use ultraviolet light to disinfect the dispensed

water. However, this study demonstrates that high numbers of bacteria may colonize the

product water. This may be partially explained by the fact that U.V. treatment leaves no

residual in the water. The EPA recommends that HPC bacteria should not exceed 500

CFU/mL in drinking water in order to reduce interference with the detection of coliform

bacteria and as a general measure of water quality (Bitton, 1994). The findings in this

study indicate that 73% of water samples from WVM did not meet that recommendation

(Figure 1). These results coincide with those found by Hunter and Burge, (1986) in their

study on the bacteriological quality of water and soft drinks from automatic vending

machines, in which HPC bacteria concentrations greater than 1,000 CFU/mL were found

in 84% and 39% of the product samples respectively.

Total coliforms are often used as indicator organisms of the presence of human

and animal feces although, some species occur naturally in the aquatic environment,

62

being common to soil and vegetation (Geldreich, 1996). Total coliform bacteria have

also been found in insects (Burgess et al., 1973). However, the presence of total

coliforms is a significant warning of a potential risk for pathogens from a contaminated

water source. Their presence should never be ignored (Mack, 1977). In this study, 20%

of the product water contained total coliforms (Figure 1). These findings are similar to a

study done by Levesque et al., (1994) in which coliforms were detected in 30% of water

samples from water coolers located in residences and workplaces.

P. aeruginosa is an important opportunistic pathogen and agent of nosocomial

disease. Hoadley, (1977) identified in his study that P. aeruginosa is a potential hazard

in water, but the occurrence of this microorganism is usually low, especially in treated

(chlorinated) drinking water (3.7%). Similarly, Edberg et al., (1996) found that only 2 to

3% of drinking water samples (bottled water, water coolers units, and tap water)

contained P. aeruginosa. Levesque et al., (1994) did not detect P. aeruginosa in water

cooler samples. In contrast, P. aeruginosa was found in 23% of the WVM water samples

in our study (Figure 1). P. aeruginosa was detected in 2 of 3 WVM's during three

consecutive weeks (Table 2), indicating that colonization is not transient (Geldreich,

1996). This may be due to the ability of P. aeruginosa to form biofilms in water pipes

(Brown and Gauthier, 1993). The nutritional versatility of P. aeruginosa enables it to

grow using trace nutrients present either in the water or attached to the carbon filters

(Tamagini and Gonzalez, 1997). Larger numbers are required (>1.5x106 CPU) to

63

colonize the respiratory tract of healthy individuals (Buck and Cooke, 1969), and

colonization by P. aeruginosa is a presage of disease (Pollack. 1990). However, the

probability of infection to healthy individuals is very low (Rusin et al., 1997).

The results obtained from nozzle and drain biofilm samples indicate that

contamination of these sites could play an important role in the quality of the water from

these machines. Previous studies (Levesque et al., 1994), showed that nozzles of water

coolers were colonized by total coliforms and fecal coliforms 44 and 20% of the time,

respectively. Neither total nor fecal coliforms were detected in the WVM nozzles in this

study (Table 3). However, only a limited number of nozzles were sampled. Numerous

HPC bacteria were found in the nozzle biofilm scrapings, which were consistent with

high numbers found in many of the product water samples. The numbers of HPC bacteria

in the drain biofilm (Table 4) were higher than those found in the nozzle. In addition, 10

of the 15 samples were positive for both total and fecal coliforms in the latter case.

Kneller et al., (1990) conducted a study done on the drain of automatic vending machines

and showed HPC counts ranging from 8 to 10 6 CFU/sampled area, and also reported the

isolation of E. coui from 23% of the samples.

The bacterial biofilm population in the present study may be due to inappropriate

cleaning and/or contamination of the nozzle area by consumers. Hunter and Burge,

(1986) indicated that inadequate cleaning of vending machines dispensing soft drinks and

drinking water may result in biofilm formation contributing to the bacterial population in

the water samples. The biofilm formation may also be exacerbated by the exposure of the

64

drain to fomites (i.e. containers used by the consumer to collect the water from the

vending machines) transmitting bacteria that may proliferate on the moist surfaces.

Burgess et al., (1972) mentioned that coliforms may be carried to the drain surfaces via

insects (cockroaches and flies), and demonstrated that E. coli is harbored and transmitted

by cockroaches. Khalil et al., (1994) showed that flies also serves as reservoir of E. cou.

Cockroaches also have been shown to carry P. aeruginosa and to excrete the organism in

high numbers (Fotedar et al., 1993). The presence of coliforms in the drain area might not

pose a risk to the WVM product water, however, inadequate cleaning practices by trained

personnel may increase the risk of contamination of the WVM product water (Hunter,

1992).

The question remains whether any of these findings have any relevance to public

health. The authors' knowledge there has not been an outbreak of disease linked to

WVM product water. However, there have been no studies that have attempted to

identify any adverse effects.

In conclusion, these results indicate that one should be aware of the

microbiological quality of the water dispensed from the WVM. Although the WVM is

well-designed and poses effective treatment for the water dispensed (reverse osmosis,

carbon filters, and UV light), the manufacturer should be aware of the need for regular

maintenance of the WVM and emphasize the importance of controlling bacterial growth

in such units.

65

References to Appendix 1

American Public Health Association. 1992. Standard Methods for the Examination ofWater and Wastewater. 18' ed. Washington, DC. American Public Health Association.

Bitton, G. 1994. Microbiological Aspects of Drinking Water Treatment and Distribution.p. 261-292. In Wiley-Liss, (ed.) Wastewater Microbiology. New York, NY.

Brown, L.M. and J.J. Gautheir. 1993. Testing of home use carbon filters. J. Amer. WaterWorks Assoc. 71:577-579.

Burgess, N.R.H., Mcdermott, S.N. and J. Whiting. 1973. Laboratory transmission ofenterobacteriaceae by the oriental cockroach, Blatta orientalis. J. Hyg. 71:9-14.

Edberg, S.C., Gallo, P. and C. Kontnick. 1996. Analysis of the virulence characteristicsof bacteria isolated from bottled water, and tap water. Microbial Ecol. Hlth. Dis. 9:67-77.

Favero, M.S, Carson, L.A., Bond, W.W. and N.J. Petersen. 1971. Pseudomonasaeruginosa: growth in distilled water from hospitals. Science. 173:836-838.

Fotedar, R., Banerjee, U. and Shriniwas. 1993. Vector potential of the German cockroachin dissemination of Pseudomonas aeruginosa. J. Hosp. Infect. 23:55-59.

Geldreich, E.E. 1996. Creating microbial quality in drinking water. p 80-90. In CRC (ed.)Microbial Quality of Water Supply in Distribution Systems. Boca Raton, FL.

Geldreich, E.E. 1989. Drinking water microbiology-new directions toward water qualityenhancement. hit. J. Food Microbiol. 9:295-312.

Gelt, J. 1996. Consumers increasingly use bottled water, home water system, watervending machines to avoid direct tap water. Arroyo. 9:1-12.

Hoadley, A.W. 1977. Potentially health hazards associated with Pseudomonasaeruginosa. p. 80-114. In Hoadley and Dutka (ed.) American Society for Testing andMaterials.

Hunter, P.R. 1994. The microbiology of drinks' vending machines. Microbiol. Europe.2:8-10.

66

Hunter, P.R. 1992. Bacteriological, hygienic, and public health aspects of food and drinkfrom vending machines. Critical Rev. Environ. Control. 22:151-167.

Hunter, P.R. and S.H. Burge. 1986. Bacteriological quality of drinks from vendingmachines. J. Hyg. Camb. 97:497-500.

Khalil, K., Lindblom, G.B., Mazhae, K. and B. Kaijser. 1994. Flies and water asreservoirs for bacterial enteropathogens in urban and rural areas in and around Lahore,Pakistan. Epidemiol. Infect. 133:435-444.

Kneller, P.B., Jayasawal, R.K. and L.M. Eils. 1990. Sanitation control for cold cup softdrink vending machines . Diary Food Environ. Sanitation. 10:449.

Levesque, B., Simard, P., Gauvin. D., Gringas, S., Dewally, E. and R. Letarte. 1994.Comparison of the microbiological quality of the water coolers and that of municipalwater systems. Appl. Environ. Microbiol. 60:1174-1178.

Lynkins, B.W., Clark, R.M. and J.A. Goodrich. 1992. Point-Of-Use/Point-Of-Entry forDrinking Water Treatment. Lewis, eds. Chelsea, MI.

Mack, W.N. 1977. Total Coliform Bacteria. Bacterial Indicators/health hazards associatedwith water. p. 59-64. In Hoadley and B.J. Dutka, (ed.) American Society for Testing andMaterials.

McFeters, G.A. 1990. Microbiology of drinking water distribution. p. 40-45. In SpringerVerlag (ed.) Drinking Water Microbiology. New York, NY.

Murray, P.R., Baron, E.J., Pfaller, M.A., Tenover, F.C. and R.H. Yolken. 1995.Pseudomonas and Burkhordelia. p. 510-511. In ASM press (ed.) Manual of ClinicalMicrobiology. Washington DC.

Payment, P., Franco, E., Richardson, L and J. Siemiatycki. 1991. Gastrointestinal healtheffects associated with the consumption of drinking water produced by point-of-usedomestic reverse osmosis filtration units. Appl. Environ. Microbiol. 57:945-948.

Rusin, P.A., Rose, J.B., Haas, C.H. and C.P. Gerba. 1997. Risk assessment ofopportunistic bacterial pathogens in drinking water. Rev. Environ. Contam. Toxicol.152:57-83.

Shooter, R.A., Cooke, E.M., Gaya, H., Kumar, P., Patel, M.T., Thom, B.T. and D.R.France. 1969. Food and medicaments as possible sources of hospital strains ofPseudomonas aeruginosa. LANCET. 1:1221-1229.

67

Statistics Software for Today's Students. Release 1.1. Copyright 1993. log Minitab Inc.

Tamagini, L.M. and R.D. Gonzalez. 1997. Bacteriological stability and growth kinetics ofP. aeruginosa in bottled water. J. Appl. Microbiol. 83:91-94.

Taylor, R.H., Allen, M.J. and E.E. Geldreich. 1979. Testing home use carbon filters. J.Amer. Water Works Assoc. 71:577-579.

The Official Methods of Analysis of the Association of Official Analytical Chemist. 14 th

ed. (1984).

Wallis, C., Stagg, H. and J.L. Melnick. 1974. The hazards of incorporating charcoalfilters into domestic water systems. Wat. Res. 8:111-113.

APPENDIX 2:

MICROBIOLOGICAL SURVEY OF PRIVATE ROOF WATER TANKS INCULIACAN, MEXICO

Cristobal Chaidez Quiroz l* , Aurora Candi12 , and Charles P. Gerba'

Department of Nutritional Sciences', Facultad de Ciencias Quimico-Biologicas', andDepartment of Soil, Water and Environmental Sciences'.

Universidad Autonoma de Sinaloa, MexicoThe University of Arizona, Tucson, AZ 85721

68

*Corresponding author

Abstract

A survey was conducted in Culiacan, Mexico on the water quality of private roof

water tanks (PRWT). In Culiacan, the house water connection is provided to nearly

100% of the urban population; however, the scarcity of water, especially in the summer,

obligates the water authority to ration potable water. As a result, people store water from

the public distribution system in PRWT. Little is known about the microbiological water

quality of the PRWT therefore, this study assessed the quality of the water after storage in

PRWT. Heterotrophic bacteria, total and fecal coliforms, Pseudomonas aeruginosa,

enteroviruses, and protozoa (Cryptosporidium and Giardia) were enumerated. Physico-

chemical parameters such as pH, temperature, turbidity, and residual chlorine were also

examined. P. aeruginosa was found in 28% of the samples; 32% also contained total

coliforms, and 20% Escherichia co/i. Heterotrophic plate count (HPC) bacteria were

found in all of the samples with numbers between 10 2 and 10 5 CFU/mL. No enteroviruses

or protozoan parasites were detected suggesting adequate protection from human or large

animal contamination. Inappropriate cleaning of the private roof water tanks may

account for high levels of bacteria.

69

Introduction

The purpose of a water supply distribution system is to deliver safe drinking water

that is adequate in quantity and acceptable in terms of taste, odor, and appearance (WHO,

1993). The Water supply released into the distribution system becomes altered during its

passage through pipes , open reservoirs, standpipes, and storage tanks (Geldreich, 1996).

Every effort should be made to achieve a drinking water quality as high as

practicable. Failure to do so exposes the population to waterborne diseases. Those at

greatest risk are the very old, the very young, and patients in hospitals (burn and

immunosuppressed patients) (WHO, 1993). Risk of disease must be controlled at the

community level by provision of uncontaminated water and basic sanitation (Ford and

Colwell, 1996). In developing countries, fast growing populations combined with poor

living conditions in rural areas have forced many people to migrate to cities in search of

better living conditions (Mintz et al., 1995). Urbanization affects water availability and

quality, this growing demand for urban water has placed much emphasis on storing water

from the public distribution system into private water storage tanks (PWST).

In Mexico, tap water is stored in private roof water tanks (PRWT) as a backup for

household use, including bathing, cooking, and drinking purposes. In Culiacan, Mexico,

the water distribution system reaches almost 100% of the urban population. However,

because of the intermittent delivery during summer, nearly all households store water in

70

71

PRWT. The urban population of Culiacan reaches almost 700,000 inhabitants (INTEGI,

census 1990). A study was conducted on the microbiological quality of the municipal

reservoirs and distribution system and showed that potable water in Culiacan is of

satisfactory quality (Candil et al., 1990). However, that study did not cover the bacterial

quality of water stored in PRWT. Storing water may contribute dramatically to the water

quality deterioration due to water stagnation (Mintz et al., 1995). Municipal water has

been shown to be further contaminated when the water is stored at home (Myo Han et al.,

1989). In Peru a study was conducted in which fecal coliform counts from water

household's storage containers were 10-fold higher than levels in the tap from which they

had been drawn, indicating that water was being contaminated further during storage

(Swerdlow et al., 1992). A sanitary survey was conducted in drinking water reservoirs

from Jordan, and 20% were contaminated with E. coli (Rabi and Abo-Shehada, 1995).

Reservoirs should be covered to avoid contamination of the supply with bird excrement

and surface run-off. Birds may be infected with intestinal pathogens such as Salmonella

and E. coli which may be pathogenic to man (Geldreich, 1996).

Organic and inorganic suspended particles may protect microorganisms from

disinfectants and may provide a suitable substratum for bacterial proliferation (Schreiber

and Schoenen, 1994), which could result in a colonization of the stored water (Silverman

et al., 1993). In many developing countries, municipal water is unsafe because of

inadequately maintained pipes, low pressure, intermittent delivery, and clandestine

connections (Ford and Colwell, 1996), implying possible consequences to the

79

microbiological quality of the drinking water itself. The risk of diarrheal disease due to

contamination of drinking water during household storage was noted in surveys

conducted by the World Health Organization (WHO) in the 1960s (Van Zilj, 1996).

Since PRWT are widely used and few studies on the microbiological quality have

been conducted, the objectives of the study were to determine the sanitary condition of

the PRWT in Culiacan, and to determine the types of microorganisms (bacteria, parasites,

and enteroviruses) colonizing them.

Materials and Methods

Sample collection

Drinking water samples from storage tanks were collected in the city of Culiacan,

Mexico. Residences located within the urban area were chosen as the only criteria to

participate in the study. Water in Culiacan is obtained from wells along the river shore

extracted by a pumping system to be stored in five municipal reservoirs. Public access is

strictly prohibited, and run off is channeled away from the reservoirs. Water is treated by

conventional processes that include flocculation, sedimentation, filtration, and

chlorination. For sampling purposes, the city was divided into five sectors and

randomized sampling was carried out in each of the sectors. For each sector, five

representative houses were chosen to obtain a total of 25 water samples for bacteria

analyses and 5 samples for virus and parasite analyses.

73

Roof water tanks made of concrete or polyvinyl chloride (PVC), and had a

capacity of 300 to 500 L. Tanks are usually covered and constructed with an inlet and

outlet pipe, floating valve, and drainage pipe. Water is drawn from the main water supply

source provided to the house. The stored water serves as a reserve water source when

interruptions to the municipal system occur.

Bacteriological analyses

Water samples were collected in sterile one-liter plastic bottles containing 2mL of

10% sodium thiosulfate to neutrilize any residual chlorine. Water samples were placed

on ice and analyzed within 2 to 3 h after sampling. All bacteriological analyses were

carried out according to the Standard Method for the Examination of Water and

Wastewater (APHA, 1995). Pseudomonas aeruginosa and total and fecal coliforms were

enumerated by the membrane filtration method (0.45 1.tm, Gelman Science, Ann Arbor,

MI) using Pseudomonas agar base (Oxoid, Hampshire, England), m-Endo and mFC agar

(Difco laboratories, Detroit. MI), respectively. P. aeruginosa samples were incubated at

37°C for 36 h. total and fecal coliforms were incubated at 37°C and 44.5°C for 24 h,

respectively. The heterotrophic plate count (HPC) bacteria were enumerated by the

spread plate method using R2A agar (Difco laboratories), and incubated at 25°C for seven

days. Identi fi cation of P. aeruginosa and coliform bacteria were conducted by the

API2OE system (Biomurex Vitek, Hazelwood, MO).

Virological and parasitological analyses

Storage water samples were analyzed for the presence of human enteroviruses and

Giardia cysts and Cryptosporidium oocysts, according to the Standard Method for the

Examination of Water and Wastewater (APHA, 1995). For viruses, a volume of 378 L

(100 gallons) of stored water was concentrated by passage through a MK filter (Zeta plus,

CUNO, Meriden, CT). The filter was eluted with 900 mL of 1.5% beef extract V (Becton

Dickinson, Cockeysville, MD), buffered with 0.05 M glycine, pH 9.5 (Ma et al., 1994).

Reconcentration of the eluate was via flocculation (Katzenelson et al., 1976). The eluate

was resuspended on 0.1M NaJIPO 4 , and adjusted to pH 7.2-7.3 with 1N HC1 and 1N

NaOH. Concentrates were treated with Freon 113 to reduce toxicity and assayed on the

BGM cell line. The cells were cultivated on minimal essential medium (MEM)

supplemented with 10% fetal calf serum in 75 cm2 tissue culture flasks. The flasks were

incubated at 37°C for 7-10 days before being frozen at -20°C and small amount of

supernatant fluid from each flask was transferred to a new monolayer of BGM cells.

There were observed for 10-14 days for the development of viral cytopatic effect (CPE).

For parasite samples, the same volume was processed (378 L), by passing it

through yarn-wound polypropylene cartridge filters (CUNO, Meriden, CT). The filters

were cut in half, teased and washed in a 1% Tween 80 solution. Eluents were

reconcentrated by centrifugation, and purified by sucrose-Percoll flotation (Sigma

-74

75

Chemical, St. Louis. MO), collected on membrane filters and analyzed for Giardia cysts

and CiTyptosporidium oocysts by immunofluorescence using specific mouse monoclonal

antibodies (Meridian, Cincinnati. OH), and subsequently stained with goat anti-mouse

antibodies conjugated to a fluorescent dye (Kyrkegaard and Perry Laboratories Inc.,

Gaithersburg,, MD). Membranes were mounted on glass slides, and examined with the

aid of an epifluorescence microscope. The cysts and oocyts were identified by their size,

shape, and green fluorescence around the walls, and confirmed by observation for internal

structures by differential phase contrast microspcopy (Rose et al., 1988).

Physico-chemical analyses

Physico-chemical analyses were performed according to the Standard Method for

the Examination of Water and Wastewater (APHA, 1995). The pH, temperature,

turbidity, and residual chlorine were determined. A database was prepared using the

Excel 4.0 computer program software. Regression analyses were performed on physico-

chemical and microbiological parameters.

Results

Bacteriological results are summarized in Table 1. P. aeruginosa was found in

28% of the samples, total and fecal coliforms in 32% and 20% , respectively (Figure 1).

76

HPC bacteria were found in all of the samples with numbers between 10 : and 10'

CR.,7mL. P. aeruginosa isolates were confirmed by the API2ONE system (Biomurex),

and selected colonies for m-Endo were confirmed by the API2OE system (Biomurex).

The predominant coliform bacteria were E. coui and Serratia marcencens. Enteroviruses,

Giardia cysts, and Cryptosporidium oocysts were not detected in the water obtained from

any PRWT.

Results of physico-chemical parameters are summarized in Table 3. The

temperature of the water varied from 17 to 35°C. An explanation of such variation may

be the different types of roof storage tanks tested (concrete and PVC), some roof tanks

tested were also protected from the sunlight by trees and therefore the water remained at

lower temperatures. The pH ranged from between 6.9 and 9.5. Such variation is

probably influenced by the materials used in construction. Turbidity varied from 0.38 to

135 NTU, such variation might be due to dust and biofilm accumulation in the storage

tank, and some tanks also had broken or unfitted covers allowing the accumulation of

large particles. Total and free chlorine were present in all of the samples with values

from 0.1 to 4.5 and from 0.1 to 4 mg/L, respectively.

Regression analyses revealed that there were no statistical significance among physico-

chemical and microbiological parameters (P<0.05).

0 % of samples positive

77

P. aerziginosa

Total conforms Fecal coliforms HPC bacteria

FIGURE 1. Percentage of samples positive for Pseudomonas aeruginosa (CFU/100 mL),total and fecal coliforms (CFU/100 mL) and heterotrophic plate count (HPC) bacteria(>500 CFU/mL)

TABLE 1. Bacteriological quality of water from household roof tanks

Roof tank P. aeruginosaCFU/100mL

Total coliformsCFU/100mL

Fecal coliformsCFU/100mL

HPC bacteriaCFU/mL

Zone-A1 0 1 l) 625007 0 7 0 195003 0 0 0 725004 100 1 1 2240005 3 0 0 17000

Zone-B1 ') 0 0 500000) 0 20 20 2000003 10 10 10 2180004 3 0 0 4000005 5 120 0 6300

Zone-C1 0 0 0 230002 0 0 0 200003 0 0 0 12504 0 0 0 1420005 0 0 0 1550

Zone-D1 0 0 0 3000-,- 1 0 0 1520003 0 15 15 1000004 0 0 0 3020005 0 15 3 6000

Zone-E1 0 0 0 27107 0 0 0 5603 0 0 0 88004 0 0 0 80005 0 0 0 2500

78

79

TABLE 2. Average, maximun and minimun values for the physico-chemical parametersin analyzed water from household roof tanks

Average Maximun Minimun

pH 7.60 9.50 6.90Temperature (°C) 26.00 35.0 17.00Turbidity (NTU) 10.32 135.0 0.38

Total chlorine (mg/L) 0.50 4.50 0.10Free chlorine (mg/L) 0.40 4.00 0.10

Discussion

Water tanks are normally used for storing water to cope with periods of maximum

demand on the water supply system. However, storage tanks can be colonized by

microorganisms if there are no proper protection against external contamination (WHO,

1993). Although, it was found that potable water in Culiacan, Mexico, is of good

microbiological condition before reaching consumers, it was found that PRWT may cause

water quality deterioration during storage (Candil et al., 1990). Drinking water becomes

contaminated after storage in home by dipping objects (i.e. bucket) into the tank resulting

in contamination of the water (Mintz et al., 1995). Our study indicates that providing

water to residences with residual chlorine may not to be enough to prevent deterioration

of its bacteriological quality. The United States Environmental Protection Agency

(USEPA) recommends that HPC bacteria should not exceed 500 CFU/mL in drinking

water in order to reduce interference with the detection of coliform bacteria (EPA, 1986).

80

Higher numbers are often the results of bacterial regrowth. particularly in distribution

systems and storage water containers (Geldreich, 1996). HPC bacteria are capable of

surviving on minimal nutrients, attach to pipe sediments, and participate in the biofilm

formation (Geldreich, 1996). There is a generalized concern regarding the public health

risk of some members of the HPC bacteria (LeChevallier et al., 1985; Payment et al.,

1988). However, health risk with HPC bacteria have not been established.

Sediments in drinking water reservoirs have been evaluated and colony counts of

up to 3x10 5 CF1i/mL using standard plate agar have been reported (Schreiber and

Schoenen, 1994). In our study we did not attempt to analyze the sediments in the

household tanks, however the water itself had between 10 2 and 10 CFU/mL. The purpose

of quantifying indicator organisms is to predict the possible occurrence of specific

pathogens and assess the sanitary quality of the water (Ford and Colwell, 1996). Total

coliforms are often used as indicator organisms for the presence of human and animal

excreta. However, indicator organisms such as E. co/i can grow in the absence of fecal

contamination , and therefore their usefulness is sometimes limited (Rivera et al., 1988).

Some strains are common in soil and vegetation (Geldreich, 1996), and insects may be

carriers of E. coli (Burgess et al., 1973). In this study, 32 and 20% of the water samples

contained total and fecal coliforms, respectively (Figure 1), E. co/i and Serratia

marcescens being the predominant coliform bacteria identified in this study.

Enterobacter cloacae and Citrobacter fretindii were detected in the outlet of open

81

holding reservoirs containing chlorinated water (Paspaliaris and Hodgson, 1994). In Iraq,

a study was conducted on tap water and

storage tanks which found that Enterobacter cloacae and Klebsiella pneumoniae were the

predominant organisms, respectively (Sameer et al., 1988).

P. aerugin osa is the most prevalent pathogen in the genus Pseudomonas. P.

aeruginosa is predominantly a nosocomial pathogen (Pollack, 1990), but it is present in

many environments including swimming pools, whirpools, hot tubs, contact lenses

solutions and other water-related reservoirs (Rusin et al., 1997). In this study, P.

aeruginosa was found in 28% of the water storage tanks (Figure 1). It has also been

isolated from water distribution systems, and 3% of the time in tap water (Geldreich and

Allen, 1975). It is able to form biofilms on several surfaces including PVC and carbon

filters (Geldreich and Allen, 1975). The nutritional versatility of P. aeruginosa enables it

to grow using trace nutrients presents in water (Tamagini and Gonzalez, 1997). The

presence of P. aeruginosa in storage tanks did not coincide with a study in which P.

aeruginosa was present in low numbers (3.7%), especially in chlorinated tap water

(Hoadley, 1977).

The absence of enteroviruses and protozoan parasites suggests that the coliforms

and fecal coliforms are not from human fecal material. It also suggests that the enteric

bacteria which were detected may have originated from birds, other small animals, or

debris that gained access into the tanks. The PRWT examined had covers but only a few

firmly fitted which may contribute to the high contamination level.

82

Our bacteriological results indicate that water from storage tanks in Culiacan,

Mexico, did not meet drinking water standards for bacterial indicators. However, the

health risk associated with the presence of bacteria in the storage tank is probably low as

no human parasite or viral pathogens were detected.

References to Appendix 2:

Allen, M.J. and E.E. Geldreich. 1975. Bacteriological criteria for ground waterquality. Ground Water. 13:45-52.

American Public Health Association. 1995. Standard Methods for the Examination ofWater and Wastewater. le ed. Washington, DC.

Burgess, N.R.H., Mcdermott, S.N., and J. Whiting. 1973. Laboratory transmission ofEnterobacteriaceae by oriental cockroach, Blatta orientalis. J Hyg. 71:9-14.

Candil-Ruiz, A.. Uribe, M., Diaz, S.P., De Leon, R., Naranjo, J., and C.P. Gerba. 1990.Calidad microbiologica del agua de la zona centro de Sinaloa. Sociedad Mexicana deIngieneria Sanitaria Ambiental, A. C. Congreso Nacional. 25-26. Oaxaca, Oax.

Environmental Protection Agency (EPA). 1986. Ambient water quality criteria forbacteria. EPA 44015-84-002. Washington, DC.

Ford, T.E. and R.R. Colwell. 1996. A global decline in microbiological safety of water: acall for action. American Academy of Microbiology. Washington, DC.

Geldreich, E.E. 1996. Microbial Quality of Water Supply in Distribution Systems.Boca Raton, FL.

Hoadley, A.W. Potentially health hazards associated with Pseudomonasaeruginosa, bacterial indicator/health hazards associated with water. American Societyfor Testing and Materials. Hoadley and Dutka (ed.) Atlanta, GA. 1977.

83

Instituto Nacional de Estadistica, Geografia e Infounatica (INEGI). 1990. XI CensoGeneral de Poblacion y Vivienda. Mexico.

Katzenelson, E., Fattal, B., and T. Hostovesky. 1976. Organic flocculation: an efficientsecond-step concentration method for the detection of viruses in tap water. App!. Environ.Microbiol. 32:638-639.

LeChevallier, M.W.. Singh, A., Schiemann, D.A., and G.A. McFeters. 1985. Changes invirulence of waterborne enteropathogens with chlorine injury. App!. Environ. Microbiol.50:412-419.

Ma, J., Naranjo, J., and C.P Gerba. 1994. Evaluation of MK filters for recovery ofenteroviruses from tap water. App!. Environ. Microbiol. 60:1974-1977.

Mintz, E.D., Reiff, E.M., and R.V. Tauxe. 1995. Safe water treatment and storage in thehome: A practical new strategy to prevent waterborne disease. J Amer. Med. Assoc.273:848-953.

Myo Han, A., Nwe 00, K., Midorikawa, Y., and S. Shwe. 1987. Contamination ofdrinking water during collection and storage. Trop. Geograph. Med.. 41:138-140.

Paspaliaris, P. and B. Hodgson. 1994. The role of cypress leaves in promoting growth ofcoliform organisms in holding reservoirs. Wat. Res. 28:2147-2151.

Payment, P., Gamache, F. and G. Paquette. 1988. Microbiological and virologicalanalysis of water from two water filtration plants and their distribution systems. Can. J.Microbiol. 34:1304-1309.

Pollack, M. 1990. Pseudomonas aeruginosa. p. 1673-1687. In Mandell, G.H., DouglasR.G. Jr., Benett, J.E. (ed.) Principles and practice of infectious diseases. Livingstone, NY.

Rabi, A.Z. and M.N. Abo-Shehada. 1995. Sanitary survey of private drinking waterreservoirs in Northern Jordan. Int. J. Environ. Hlth. Res. 5:255-261.

Rivera, S.C., Hazen, T.C. and G.A. Toranzos. 1988. Isolation of fecal coliforms frompristine sites in a tropical rain forest. App!. Environ. Microbiol. 54:513-517.

Rose, J.B., Darbin, H. and C.P. Gerba. 1988. Correlations of the protozoa,Cryptosporidium and Giardia with water quality variables in a watershed. Wat. Sci. Tech.20:271-276.

84

Rusin, P.A., Rose, J.B., Haas, C.H. and C.P Gerba. 1997. Risk assessment ofopportunistic bacterial pathogens in drinking water. Rev. Environ. Con tam. Toxicol.152:57-83.

Sameer, F.J., Zainab, A., Al, D. and T.A. Haddad. 1988. Antibiotic resistance coliformand fecal coliform bacteria in drinking water. Wat. Air Soil Poll. 39:377-382.

Schreiber, H. and D. Schoenen. 1994. Chemical, bacteriological and biologicalexamination and evaluation of sediments from drinking water reservoirs. ZbL Hyg.196:153-169.

Silverman, S., Nagy, L.A. and B.H. Olson. 1983. Variations in particulate matter, algae,and bacteria in uncovered finished drinking water reservoirs. J Amer. Water WorksAssoc. 75:191-195.

Swerdlow, D.L., Mintz, E.D., Rodriguez, M., Tejeda, E., Ocampo, C., Espejo, L.,Greene, K.D., Saldana, W., Seminario, L., Tauxe, R.V., Wells, J.G., Bean, N.H., Ries,A.A., Pollack, M., Vertiz, B. and P.A. Blake. 1992. Waterborne transmission of epidemiccholera in Trujillo, Peru: lessons for a continent at risk. LANCET. 340:28-32.

Tamagini, L.M. and R.D. Gonzalez. 1997. Bacteriological stability and growth kinetics ofP. aeruginosa in bottled water. J. App!. Microbiol. 83:91-94.

Van Zilj, W.J. 1966. Studies on diarrhoea! in seven countries by the WHO diarrhoealdisease study team. Bull. Wld. Organ. 35:249-261.

World Health Organization Guidelines for drinking water quality. 2" d ed. Geneva,Switzerland. 1993.

APPENDIX 3:

MICROBIOLOGICAL COMPARISON OF THE QUALITY OF POINT-OF-USE(POU)-TREATED WATER, TAP WATER, AND BOTTLED WATER

Cristobal Chaidez Quiroe and Charles P. Gerba'

Department of Nutritional Sciences and Department of Soil, Water, and EnvironmentalSciences'.

The University of Arizona, Tucson, AZ. 85721

85

*Corresponding author

86

Abstract

The water quality of point-of-use (POU) water treatment devices, and that of tap

water with POU-connections, tap water, and bottled water were compared. Heterotrophic

plate count (HPC) bacteria, total and fecal coliforms, and acid-fast organisms

(Itlycobacteria spp.) as well as the opportunistic bacterial pathogens Aeromonas

hydrophila, Plesiomonas shigelloides, and Pseudomonas aeruginosa were enumerated.

The highest concentration of bacteria was found in POU-treated water. P. aerziginosa,

acid-fast organisms, and total coliforms were present in 38.5, 43.8, and 82.4% of the

samples, respectively. HPC bacteria were present in all of the POU-treated water

samples, with concentrations ranging from 10 2 to 10 CFU/mL. Neither fecal coliforms

nor P. shigelloides were recovered from any samples. Tap water with a POU-connection

also had higher numbers of bacteria than tap and bottled water samples. It was concluded

that tap and bottled water had lower numbers of A. hydrophila, acid-fast organisms, HPC

bacteria, P. aeruginosa and coliforms than POU-treated water, and tap water with a POU-

connection. The use of POU-devices may amplify the bacteria present in the distribution

system by promoting biofilm formation. Regular maintenance of POU-devices is

recommended to minimize the numbers of bacteria.

87

Introduction

In the past decade there has been a growing concern by public health officials,

regulatory agencies, and the general public about the safety of drinking water supplies

(Geldreich, 1996). A growing segment of the public is convinced that many health risks

may emerge from the water distribution system. An available alternative to this threat is

bottled water, or point-of-use (POU)-treated water (Geldreich, 1996; McFeters, 1990).

In North America, the market for bottled water is large, with an annual growth

rate estimated at 25% in recent years. The consumption of bottled water in the United

States has risen from four gallons per person per year in 1984 to 9.7 gallons per year in

1994. In 1995, Americans bought 2.8 billion gallons of bottled water, and in 1996 they

spent $2 billion on drinking water (Swanson, 1996; Holtzman et al., 1997). Overall, 10

to 15% of the U.S. households drink bottled water on a regular basis (Symons, 1997).

Bottled water is intended to be used as a beverage, however, it has also been

marketed as ideal for infant formula preparation, or nursery drinking water, for use in

reconstituting other foods, for cleaning contact lenses, for skin care, and for humidifiers

(Warburton et al., 1992). Thus, bottled water should be of better microbiological quality

than most foods, especially if intended for use by vulnerable populations (pregnant

women, the infirm, the elderly, and the very young) (Warburton, 1993).

Bottled water may contain bacteria that enters as contaminants, including a wide

range of saprophytic species and human pathogens (Warburton et al., 1993). It has been

88

shown that many shallow and deep aquifiers contain microorganisms, sometimes at levels

as high as 10 5 -10 CFU/mL (Warburton, 1993). Sometimes, bottled water industry uses

potable water as a source water, and indigenous bacteria from the municipal source may

appear in the product water (Warburton, 1993).

The point-of-use (POU) devices industry has experienced a growth increase over

the past years. In the United States, the Water Quality Association has reported that there

are at least 325 manufacturers of POU home water treatment devices. The annual sales in

1990 rose to nearly $1 billion and these products were serving more than 5 million

families nationwide (McFeters, 1990).

Water purifiers are treatment devices designed to remove contaminats from the

water (Lynkins et al., 1992). POU-devices use a wide range of treatments to control

contaminants in drinking water. Treatments include: adsorption, ion-exchange, and

disinfection such as chlorination, ozonation, silver salts, and ultraviolet light. In some

designs, the treatment package consists of several processes in series to achieve better

removal and control of water contaminants (Lynkins et al., 1992; McFeters, 1990).

However, studies have shown that these units can deteriorate the microbial water quality

(Geldreich, 1996). The microbial quality of the water produced by these devices is

influenced by input water flora, service life of the filter, physico-chemical water quality

(i.e. temperature), and frequency of static water conditions (Geldreich et al., 1985).

Amplification of the heterotrophic bacterial population in POU-devices is of concern.

Opportunistic pathogens can be found as part of the heterotrophic plate count (HPC)

89

bacteria in drinking water (Rusin et al., 1997a). Therefore, their growth may threat

vulnerable segments of the population (very young, pregnant women, and

immunocompromised).

A variety of bacteria occur in drinking water by entering during water line repairs,

in open finished water reservoirs, in blackflows from pipelining projects, and new pipe

network (Payment et al., 1997; Geldreich, 1996). Most of these bacteria are not of

immediate public health significance, but some can represent a health risk (Geldreich,

1996). Higher numbers are often the result of bacterial re-growth, particularly in

distribution systems (Geldreich, 1996; Olson and Nagy, 1984), and in water treatment

devices (Geldreich et al., 1985; Payment, 1989; Reasoner et al., 1987; Snyder et al.,

1995). Bacterial densities in water effluents from POU-devices can be expected to

increase by one or two logs over the number detected in water supplies (Taylor et al.,

1979). Wallis et al., (1974) observed that bacterial densities increased in filters devices

after an overnight period of non-use. This proliferation depends on certain conditions

such as seasonal changes, temperature and free chlorine concentration (Geldreich, 1996).

Public water supplies were never intended to provide a sterile product, but to

supply safer drinking water. However, before reaching the consumer, the microbial

quality of potable water is modified in composition through treatment processes,

colonization in the distribution system, and selectively amplification in attachment

devices (Geldreich, 1986). These changes are caused to a large extent by a variety of

90

microbiological activities that occur as water passes through the distribution network

(Bitton, 1994).

Studies on the bacteriological quality of the POU-devices under laboratory

settings have been extensively conducted (Naranjo et al., 1997; Geldreich et al., 1985;

Geldreich and Reasoner, 1989; Tobin et al., 1981). However, field studies are limited.

Previous field studies have been conducted (Snyder et al., 1995), but differentiation and

concentration of selected opportunistic pathogens have not been performed. The

purposes of the study reported here were : 1) to evaluate the microbiological quality of

the water dispensed by POU-devices, tap water with POU-connection tap water, and

bottled water, 2) to compare the quality of each of the drinking water sources, and 3) to

determine the occurrence of selected opportunistic pathogens.

Materials and Methods

Sample collection

Point-of-use (POU) devices were installed in 10 private homes in January of

1998, in Tucson, AZ. The units were connected to the faucet that delivers cold water.

The faucet-mounted units provide an easy access to both filtered and unfiltered (tap)

water by a simply twist from the vertical (filtered water) to the horizontal position

(unfiltered water). Thus, an influent (unfiltered) and an effluent (filtered) water sample

91

were collected from the installed POU-devices. Table 1 describes the general

characteristics of the POU-devices used in this study.

Table 1. General characteristics of point-of-use (POU) devices (faucet mounted units)

Unit Model Filtration system Service life(gallons)

PUR FM-2000 Micron carbon 300PUR FM-2000 Micron-carbon 300PUR FM-2000 Micron-carbon 300PUR FM-2000 Micron-carbon 300PUR FM-2000 Micron-carbon 300

PUR-plus RF-30520 Micron-carbon 1000Teledyne pik FM-1 Carbon 300

Culligan FM-15 Carbon 200Culligan FM-15 Carbon 200

Nordic ware NW-78100 Carbon 500

Sample collection procedures are critical to obtain a representation of water

quality within the POU-device. Therefore, to make sure that the samples were

representative of the water consumed, neither the POU-treated water nor the tap water

were flushed before sampling and there was no attempt to sterilize the outer surface of

both sites. First-draw water samples were collected aseptically in sterile 1-liter plastic

bottles from both POU-device and the tap water. A total of four liters of water were

obtained from each site. Each bottle sample contained 2 mL of sterile sodium thiosulfate

to neutralize any residual chlorine. Samples were stored in a plastic cooler and packed

with ice for transport to the laboratory for immediate processing.

9?

The sampling was carried out over a period of six weeks. Every week a sample

was obtained from both POU-treated water and tap water. Also, samples from other tap

home-owners (without POU-device connection) were obtained to determine if the POU-

device exert an influence on the bacteriological quality of the tap water. Samples were

collected every week to obtain a total of thirty water samples. They were collected in

sterile plastic bottles containing 2 mi, of sterile sodium thiosulfate. Tap water samples

from homes without POU-devices were used as the baseline value.

Bacteriological analysis were conducted on 72 brands of retail bottled water from

7 countries (Costa Rica, Greece, Japan, Korea, Mexico, Panama, and US). Container

sizes ranges from 300 to 3785 mL. The types of water were: distilled, spring, and

purified water. Samples were placed at room temperature before analysis.

Bacteriological analyses

Bacterial analyses were performed on POU-treated water, tap water with POU-

connection, tap water, and bottled water according to the Standard Methods for the

Examination of Water and Wastewater (APHA, 1995), unless otherwise indicated. HPC

bacteria were determined using the spread plate method by placing 0.1 mL of the sample

onto R2A agar plates (Difco Laboratories, Detroit, MI). These plates were incubated at

25°C for 7 days. Membrane (0.45 pm pore size) filtration technique (Gelman Science

Laboratories, Ann Arbor, MI) was used for the enumeration of Aeromonas hydrophila,

acid-fast organisms, Pseudomonas aeruginosa, Plesionzonas shigelloides, and total and

93

fecal coliforms. Five-hundred mL of water samples were used for A. hydrophila, P.

aerugin osa, and P. shigelloides, the membranes (Gelman Sciences Laboratories) were

placed onto Aeromonas agar (Difco Laboratories), Pseudomonas agar (Oxoid

Laboratories, Hampshire, England), and inositol brilliant green bile salts (IBB) agar

(Jeppensen, 1995), respectively. Membranes were incubated at 37°C for 36 h. One-

thousand mL samples were collected for the enumeration of acid-fast organisms. The

membranes were decontaminated with 0.04% of Cetylpiridinium chloride (CPC) for 2h

and placed onto 7H11 agar (Difco Laboratories) at 37°C for four weeks at 5% CO,

environment. One-hundred mL samples were obtained to enumerate total and fecal

coliforms. Membranes were placed onto m-Endo at 37°C and mFc agar at 44.5°C for 24

h, respectively.

API2OE and API2ONE (Biomurex Vitek, Hazelwood, MO) systems were utilized

for the identification of fermentative and non-fermentative bacteria. The acid-fast

(Kinyoun) stain was used for the identification of the acid-fast organisms.

Physico-chemical analyses

Physico-chemical analyses were performed according to the Standard Method for

the Examination of Water and Wastewater (APHA, 1995). Turbidity, pH, temperature

and residual chlorine were monitored in the POU-treated water, tap water with POU-

connection and tap water, and bottled water.

94

Statistical method

Water quality data were analyzed by the Pearson correlation and the linear

regression with the aid EXCEL computer software (Microsoft Co., Redmond, WA).

Results

Results of bacteriological analyses from the POU-treated water, tap water with

POU-connection, tap water, and bottled water are shown in Table 2. The average

concentration of A. hydrophila, P. aeruginosa, and total coliforms are presented in Table

3. The results of the HPC bacteria for POU-treated water, tap water with POU-

connection, tap water, and bottled water are presented in Table 4. Table 5 shows the

identification of selected bacterial isolates from the drinking water sources. Physico-

chemical parameters are summarized in Table 6.

The numbers of HPC bacteria were compared among POU-treated water, tap

water with POU-connection, and tap water (Figure 1). Data were collected over a period

of six weeks. POU-devices were colonized by bacteria shortly after installation, and the

numbers were much higher than those found in tap water with POU-connection and tap

water. P. aeruginosa densities of POU-treated water, tap water with POU-connection,

and tap water are shown on Figure 2. The average number of A. hydrophila was also

compared in Figure 3. The average of total coliforms is summarized and compared in

Figure 4.

95

Correlation and regression analyses were perfolined on bacteriological and

physico-chemical data, however, no significant differences were found.

Discussion

Bacterial colonization of POU-devices occurred shortly after installation. POU-

treated water showed the highest bacterial contamination when compared with other

drinking water sources (Table 2). The average concentration of A. hydrophila was low in

POU-treated water (29.5/mL) (Table 3). This might be due to the fact that A. hydrophila

is not able to grow extensively in nutrient-poor waters, and is not able to compete

efficiently with the heterotrophic bacteria (Holmes et al., 1996). Moreover, it has been

suggested that heterotrophic bacteria might be inhibitory to A. hydrophila (Hunter, 1993).

In contrast, P. aeruginosa and total coliforms were recovered at much higher

concentrations (102 and 139 CFU/mL, respectively) in POU-treated water than other

drinking water sources (Table 3). The explanation of this is that P. aeruginosa and total

coliforms participate in biofilm formation and they are able to co-exist in the biofilm

surfaces (Geldreich, 1996). Geldreich, (1996) mentioned that organisms that reside in

biofilms are expected to be more consistently found in the product water. POU-treated

water had higher concentrations of acid-fast bacilli (based on the acid-fast stain),

followed by tap water with POU-connection, tap water and bottled water (Table 2).

Acid-fast organisms are also able to participate in biofilm formation and have been

96

recovered from chlorinated tap water and bottled water (Bullin et al., 1970; Geldreich,

1996; Caroli et al., 1985).

The greatest concentration of HPC bacteria were found in POU-treated water,

followed by tap water with POU-connection, tap water, and bottled water (Table 4). The

microbial flora of bottled water is altered as water is extracted from the source and passes

through the bottling plant. Warburton et al., (1998) found that 5.5% of the bottled water

tested exceeded 1x104 CFU/mL. The HPC bacteria found in POU-treated water, tap

water with POU-connection and tap water are compared in Figure 1. POU-treated water

and tap water with POU-connection maintained high constant numbers of HPC bacteria

during the six weeks of the sampling. Tap water and bottled water had similar

concentrations of HPC bacteria, indicating that the use of POU-devices may promote the

proliferation of heterotrophic bacteria and might then have had an influence on the water

source (tap water connection). This might be explained by the fact that water filters may

support biofilm formation. These results coincide with those found by Levesque et al.,

(1994), in their study of the bacteriological quality of tap water and water coolers. They

found that the water taken from water coolers were much more contaminated than the tap

water. They suggested that water coolers might affect the microbial quality of the bottled

water obtained from the faucets. In contrast, Snyder et al., (1995) analyzed the

bacteriological quality of 24 domestic ground water supplies treated with powdered

activated carbon filters (POU-PAC). They found that POU-PAC did not have an effect

on the bacteriological quality of drinking obtained from underground sources.

97

Hunter, (1994) states that bottled water must be free of coliforms (0/100 mL) and

P. aeruginosa (0/250 mL). In the current study P. aeruginosa was present in 8.3% and

coliforms in 15.2% of the bottled water tested at concentrations of 0.6 and 4.4 CFU/mL,

respectively (Table 2). A. hydrophda and acid-fast organisms (based on acid-fast stain)

must be absent as well (Geldreich, 1996; Warburton et al., 1994), but they were identified

in 8.7% and 11.4% of the samples, respectively (Table 2). Caroli et al., (1985) reported

the isolation of acid-fast organisms in nine bottle water samples. Slade et al., (1986)

isolated A. hydrophila from 41 of 95 (43%) bottles of mineral water in Saudi Arabia.

Opportunistic bacterial pathogens, total coliforrns, and HPC bacteria were

detected in the bottled waters (Table 2). The results are similar to those reported by

others authors (Edberg et al., 1996; Ogan, 1992; Fewtrell et al., 1997; Gonzalez et al.,

1987; Tsai and Yu, 1997; Hunter, 1994; Manaia et al., 1990; Caroli et al., 1985). Low

numbers of opportunistic pathogens were detected in bottled water, except one sample

obtained from Costa Rica that contained 90 CFIJ/mL of A. hydrophda. Warburton et al.,

(1994) mentioned that this organism is able to survive and proliferate to levels of 1000

CFU/mL in bottled water stored at room temperature. Bacterial contamination can be

attributed to several factors including contamination during bottling and storage

(Geldreich, 1986). Prolonged storage at room and refrigeration temperatures allow

multiplication of bacteria to numbers >1x10 5 (Bischofberger et al., 1990; Geldreich et al.,

1975; Warburton et al., 1992).

98

The presence of P. aeruginosa , A. hydrophila, and total coliforms may indicate

contamination by pollution and human fecal matter, and is an indication of a poor

manufacturing practices (Hoadley, 1977; Warburton et al., 1992).

Overall, opportunistic pathogens were present in all of the drinking water sources.

Although, the numbers of cells necessary for an infective dose is relatively high for

opportunistic pathogens (ranging from 1 x10 6 to 1x10' ° CFU), the volume of water

consumed could conceivably deliver an infective dose during serial exposures

(Warburton et al., 1992). Adverse health effects is dependent upon, the bacteria involved,

the number of bacteria ingested and the individual's general health including resistant to a

particular organisms (McFeters, 1990).

The United States Environmental Protection Agency (EPA) has suggested that

HPC bacteria counts in drinking water should not exceed 500 CFU/mL, primarily

because of the interference of coliforms detection (EPA, 1989). In this study, drinking

water often failed to meet such a recommendation. HPC bacteria were often present in

numbers that reached lx 1 0 3 to 1x10 7 CFU/mL. Caution is needed when interpreting the

public health significance of HPC bacteria in drinking water. Although, it has been

suggested that these bacteria may be potentially pathogenic to the most vulnerable

individuals including infants and immunocompromised (McFeters, 1990), the potential

for adverse health effects appear to be low (Rusin et al., 1997a)

Various studies have shown that home treatment devices are effective in the

removal of contaminants in drinking water (microbiological and chemical), but the

99

effectiveness might be short-lived (McFeters, 1990). In this study, POU devices

efficiently removed chlorine from the tap water source (Table 6). However, the microbial

quality of the water produced from these devices was extremely variable. This may be a

function of service life of the filter, water static conditions in distribution systems, and

water line repairs rather than the municipal water treatment (Geldreich et al., 1986).

On the basis of the results obtained in this study, bottled water and municipal tap

water were superior in bacterial quality when compared with the water dispensed by

POU-devices and the tap water with POU-connection. This results indicated that the use

of POU-devices might increase the numbers of bacteria in drinking water, promote

biofilm formation, and might have an influence on the bacterial quality of the connected

tap water.

The results obtained in this study suggest that microbial contamination might be

derived from the plumbing system (sediment deposition, static water zones, and warm

water), and POU-devices may allow bacteria to proliferate in the filter device. Therefore,

the larger the water standing in the distribution system, the greater the potential there is

for bacteria to accumulate in the POU-device. Regular maintenance of the filter

(changing filter cartridge at least as frequently as recommended by the manufacturers)

may enhance the water quality of the water dispensed from POU-device.

100

References to Appendix 3:

American Public Health Association (APHA). 1995. Standard Methods for theExamination of Water and Wastewater, 1 8th ed. Washington, DC.

Bischofberger, T. Cha, S.K., Schmitt, R., Konig, B. and W. Schmidt-Lorenz. 1990. Thebacterial flora of non-carbonated, natural mineral water from springs to reservoir andglass and plastic bottles. Int. I Food Microbiol. 11:51-72.

Bitton, G. 1994. Wastewater Microbiology. Wiley-Liss. NY.

Bullin, C.H., Tanner, E.I. and C.H. Collins. 1970. Isolation of Mycobacterium xenopeifrom water taps. I Hyg. Camb. 68:97-100.

Caroli, G., Levre, E., Armani, G., Biffi-gentili, S. and G. Molonari. 1985. Search for acid-fast in bottled water. J. App!. Bacteriol. 58:461-464.

Collins, C.H., Grange, J.M. and M.D. Yates. 1984. Mycobacteria in water. J. Appl.Bacteriol. 57:193-211.

Difco laboratories. 1984. Difco manual: dehydrated culture media and reagentsfor microbiology, 10 6' ed. Difco laboratories, Detroit, MI.

Edberg, S.C., Gallo, P. and C. Kontnick. 1996. Analysis of the virulence characteristicsof bacteria isolated from bottled water, water cooler, and tap water. Microbial Ecol. Hlth.Dis. 9:67-77.

Environmental Protection Agency (EPA). 1989. Drinking water health effects task force.Health effects of drinking water treatment technologies. Lewis Pubs. Chelsea, MI.

Fewtrell, L., Kay, D., Wyer, M., Godfree, A. and G. O'neall. 1997. Microbiologicalquality of bottled water. Wat. Sci. Tech. 35:47-53.

Geldreich, E.E. 1996. Microbial quality of water supply in distribution systems. CRC,Boca Raton, FL.

Geldreich, E.E. and D.J. Reasoner. 1989. Home water treatment devices and waterquality, p. 147-167. In G.A McFeters, Springer-verlag (ed.). Drinking watermicrobiology. New York, NY.

101

Geldreich, E.E. 1986. Potable water: new directions in microbiological regulations. ASM

News. 52:530-534.

Geldreich, E.E., Taylor, R.H., Blannon, J.C. and D.J. Reasoner. 1985. Bacterial

colonization of point-of-use water treatment devices. J. Amer. Water Works. Assoc.

77:72-80.

Geldreich, E.E., Nash, H.D., Reasoner, D.J. and R.H. Taylor. 1975. The necessity of

controlling bacterial population in potable water-bottled water, and emergency water

supplies. J. Amer. Water Works Assoc. 67:117-124.

Gonzalez, C. Gutierrez, C. and T. Grande. 1987. Bacterial flora in bottled uncarbonated

mineral drinking water. Can. J. Microbiol. 33:1120-1125.

Hoadley, A. W. 1977. Pseudomonas aeruginosa in surface waters. p. 31-57. In young,

V.M. Raven press (ed.) Pseudomonas aeruginosa: ecological aspects and patients

colonization., New York, NY.

Holmes, P., Niccolls, L.M. and D.P. Sartory. 1996. The ecology of mesophilic

Aeromonas in aquatic environment. p. 126-150. In Wiley (ed.) The genus Aeromonas.

New York, NY.

Holtzman, A.E., Aronson, T.W., Glover, N., Froman, S., Stelma, G.N., Sebata, S.N.,Boian, M.G. and O.G.W. Berlin. 1997. Examination of bottled water for non-tuberculous

Mycobacteria. J. Food Protect. 60:185-187.

Hunter, P.R. 1993. The microbiology of bottled natural mineral waters. J. App!.

Bacteriol. 74:345-352.

Hunter, P.R. 1994. Bottled natural mineral water and other bottled waters. Microbiol.

Europe. 2:8-10.

Jeppesen, C. 1995. Media for ileromonas spp., Plesiomonas spp. and Pseudomonas spp.

from and the environment. Int. J. Food Microbiol. 26:25-41.

Levesque, B., Pierre, S., Gauvin, D., Gingras, S., Dewailly, E. and R. Letarte. 1994.

Comparison of the microbiological quality of water coolers and that of municipal water

systems. App!. Environ. Microbiol. 60:1171-1178.

Lynkins, B.U., Clark, R.M. and A.G. James. 1992. Point-of-use/point-of-entry for

drinking water treatment. Lewis, Chelsea, MI.

102

Manaia, C.M., Nunes, 0.C., Morais, P.V. and M.S. Da Costa. 1990. Heterotrophic platecounts and the isolation of bacteria from mineral waters on selective andenrichment media. J. App!. Bacteriol. 69:871-876.

McFeters, G.A. 1990. Drinking water microbiology. G.A. McFeters (ed.), SpringerVerlag. New York, NY.

Naranjo, J.E., Chaidez, C., Quinonez, M. Gerba, C.P., Olson, J. and J. Dekko. 1997.Evaluation of portable water purification system for the removal of enteric pathogens.Wat. Sci. Tech. 35:55-58.

Ogan, M. 1992. Microbiological quality of bottled water sold in retail in Nigeria. I App!.Bacteriol. 73:175-181.

Olson, B.H. and L.G. Nagy. 1984. Microbiology of potable water. Adv. Appl. Microbiol.30:73-132.

Payment, P., Franco, E. and Siemiatycki. 1991. Gastrointestinal health effects associatedwith the consumption of drinking water produced by point-of-use domestic reverseosmosis filtration units. App!. Environ. Microbiol. 57:945-948.

Payment, P., Siemiatycki, J., Richardson, L., Renaud, G., Franco, E. and M. Prevost.1997. A prospective epidemiological health effects due to the consumption of drinkingwater. Int. i Environ. Hit/i. Res. 7:5-31.

Payment, P. 1989. Bacterial colonization of domestic reverse-osmosis filtration units.Can. J. Microbiol. 35:1065-1067.

Reasoner, D.J., Blarmon, J.C. and E.E. Geldreich. 1987. Microbiological characteristicsof third-faucet point-of-use devices. J. Amer. Water Works Assoc. 79:60-66.

Rusin, P.A., Rose, J.B., Haas, C.H. and C. P. Gerba. 1997. Risk assessment ofopportunistic bacterial pathogens in drinking water. Rev. Environ. Con tam. Toxicol.152:57-83.

Rusin, P.A., Rose, J.B. and C.P. Gerba. 1997b. Health significance of pigmented bacteriain drinking water. Wat. Sci. Tech. 32:21-27.

Slade, P.J., Falah, M.A. and A.M. Al-Ghady. 1986. Isolation of Aeromonas hydrophilafrom bottled waters and domestic water supplies in Saudi Arabia. J. Food Protect.49:471-476.

103

Snyder, J.W., Craig, N.M., Anderson, R.E and G.K. Bissonnette. 1995. Effect on point-of-use, activated carbon filters on the bacteriological quality of rural groundwatersupplies. Appl. Environ. Microbiol. 61:4291-4295.

Swanson, S. 1996. International bottled water association: an update. Water Cond. Purif.38:40-41.

Symons, M.J. 1997. Home facts. Texas A&M University Press, Texas.

Taylor, R.H., Allen, M.J. and E.E. Geldreich. 1979. Testing of home use carbon filters.Amer. Water Works Assoc. 71:577-579.

Tobin, R.S., Smith, D.K. and J.A. Lindsay. 1981. Effects of activated carbon andbacteriostatic filters on microbiological quality of drinking water. Appl. Environ.Microbiol. 41:646-651.

Tsai, G.J. and S.C. Yu. 1997. Microbiological evaluation of bottled water uncarbonatedmineral water in Taiwan. Int. I Food Microbiol. 37:137-143.

Warburton, D.W., Dodds, K.L., Burke, R., Johnston, M.A. and P.J. Laffey. 1993. Areview of the microbiological quality of bottled water sold in Canada between 1981 and1989. Can. J. Microbiol. 38:12-19.

Warburton, D.W. 1993. A review of the microbiological quality of bottled water sold inCanada. Part2. The need for more stringent standards and regulations. Can. I Microbiol.39:158-168.

Warburton, D.W., McCormick, J.K. and B. Bowen. 1994. Survival and recovery ofAeromonas hydrophila in water: development and methodology for testing bottled waterin Canada. Can. I Microbiol. 40:145-148.

Wallis, C., Stagg, C.H. and J.L. Melnick. 1974. The hazards of incorporating charcoalfilters into domestic water systems. Wat. Res. 8:111-113.

104

Table 2. Percentage bacterial occurrence in POU-treated water, tap water (with POU), tapwater (without POU), and bottled water

Percentage of bacteria (%)

Organisms POU-treatedwater(57)e

Tap withPOU(57)

Tap withoutPOU(30)

Bottledwater(72)

A. hydrophilaa 7.0 10.5 6.6 6.9P. aeruginosaa 38.5 33.3 16.6 8.3

Acid-fast organisms' 43.8 35 16.6 5.5Total coliformse 82.4 33.3 6.6 15.2

1-1PCd 87.7 61.4 43.3 65.3CFU/500 mLAcid fast stain positivesCFU/100 mL

d Heterotrophic plate count bacteria greater than 500 CFU per rriLe Number of samples

105

Table 3. Average, maximum and minimum values of bacteria in drinking water sources

POU-treated water

Tapwith POU

Tap withoutPOU

Bottledwater`

Av Mi Ma Av Mi Ma Av Mi Ma Av Mi MaA.

hydrophdaa

29.5 0 1600 1.73 0 70 10 0 200 18.2 0 1000

P.aeruginosa

a

102 0 3500 100 0 3000 15 0 320 5 0 100

Totalcoliformsb

139 0 1000 16 0 320 4.66 0 80 31 0 680

Av, average; Mi, minimum; Ma, maximumCFU/500 mL

b CFU/100 mLData without the high value sample

Table 4. Occurrence of the HPC in drinking water sources

Heterotrophic plate count bacteria (%)

HPCCFU/mLa

POU-treatedwater(57)d

Tapwith POU(57)

Tap withoutPOU(30)

Bottledwater(72)

<1 0 0 5.51-10 0 0 3.3 5.5

11- 100 1.7 7.0 26.6 15.3101 - 1,000 24.6 47.4 30.0 12.5

1,001 - 10,000 42.1 26.3 16.6 34.710,001 - 100,000 15.7 10.5 23.3 25

> 100,000 15.7 8.7 0 0Range' 100 - 4x10 7 40- 2x10' 10 - 5x10 4 <1 - 9x104

Median' 4000 700 300 1750Geometric mean' 9692 1841 642 906

Heterotrophic plate count bacteria (7d incubation time at 25°C)All values are percent of samplesNumber of bacteria per mLNumber of samples

106

Table 5. Bacteria isolated and identified in drinking water samples

Organisms POU-treatedwater

Tap water(with POU)

Tap water(without POU)

Bottledwater

Acinetobacter` XEnterobacter

cloacae'X X X X

Enterobacteraglomerans"

X X X

Klebsiella oxytocaa X X XKlebsiella

pnetunoniaeaX X X

Moraxella sp.' XPseudomonas putida b X X XSerratia marcencensb X X X

a Total coliforms"Pigmented bacteria' Heterotrophic bacteria

107

108

Table 6. Average, maximum and minimum values for the physico-chemical parameters inanalyzed water (POU-treated water, tap water with POU, tap water without POU, andbottled water)

POU-treatedWater

Tap with POU Tap withoutPOU

Bottled water

Av Mi Ma Av Mi Ma Av Mi Ma Av Mi MapH 7.6 6.6 8.7 7.8 7.1 8.6 7.8 7.4 8.2 7.3 5.6 8.3

Temperature 23 22 25 24 22 26 23 22 24 n/a nia nia(°C)

Turbidity 0.8 0.2 4.6 0.6 0.2 2.3 0.9 0.2 2.4 0.4 0.1 1.2(NTU)Total

chlorine

(mg/L)

0 0 0 0.3 0 0.8 0.3 0 0.9 0 0 0

Freechlorine

0 0 0 0.2 0 0.7 0.3 0 0.8 0 0 0

(1110-)Av, average; Mi, minimum; Ma, Maximumnia (Not available)

21P 0U-water

Tap welfP OU

Tap witho ut P OU

3 4

weeks

5 6

109

Fig 1. Mean HPC bacteria densities (CFU/mL)

6 —+-- FCUmater

--II-- Tap Wth

—it— Tap withat FCUwaek

110

Fig 2. Mean Pseudomonas aeruginosa densities (CFU/500 mL)

weeksP 0U-water

Tap with P OU

Tap witho ut P OU I

2.5

-0.5

111

Fig 3. Mean Aeromonas hydrophila densities (CFU/500mL)

•

111

. A A. A

1 2 3

weeks4 5 6

—40— POU-water

2.5

0

2

0 5

112

Tap with P OU

Tap without P OU

Fig 4. Mean total conforms densities (CFU/100mL)

113

APPENDIX 4:

AEROMONAS HYDROPHILA AND PSEUDOMONAS AERUGINOSA IN DRINKINGWATER FROM VARIOUS SOURCES: A RISK ASSESSMENT

Cristobal Chaidez Quiroz i and Charles P. Gerba* 2

Department of Nutritional Sciences' and Department of Soil, Water and EnvironmentalSciences'

The University of ArizonaTucson, Arizona 85721

*Corresponding author

114

Abstract

The opportunistic bacterial pathogens, Aeromonas hydrophila and Pseudomonas

aeruginosa have been detected worldwide in drinking water and are associated with a

number of human infections. There have been concern that the presence of these

organisms in drinking water may represent a public health threat. The purpose of this

work was to assess the potential risk of infection (colonization) of these organisms from

exposure to various sources of drinking water (POU-treated water, tap water with POU-

connection, bottled water and water from vending machines and household storage

tanks). Water samples were collected from described sources and tested.

Based on data obtained from human dose-response studies the exponential model

{Pi = 1 - e HiNxENDI was chosen to be used in this risk assessment. The model was used to

develop a microbial risk assessment to estimate daily and yearly risks of infection

(colonization) for the exposure of opportunistic pathogens via drinking water. Analysis

indicate that risks of colonization ranged from 5x10' - 1.6x10 for A. hydrophila, and

7.7x10 -9 - 1.8x10' for P. aeruginosa could result for drinking water from the sources

studied based on the consumption of 2L/day/person. In general, the risk of colonization

may be associated with the most vulnerable (very young, the elderly and

immunocompromised).

Introduction

Risk assessment has been used for setting standards for toxic chemicals in water

and foods (Rose and Gerba, 1991). This process attempts to identify the risks and

quantify the degree of risk to exposed individuals or populations. Risk assessment has

only been used on a limited scale to determine and evaluate the risks involved with

exposure to waterborne pathogenic microorganisms (Haas, 1983; Gerba and Haas, 1988;

Rose et al., 1991; Rose and Gerba, 1991; Regli et al., 1991; Rusin et al., 1997).

Opportunistic bacterial pathogens are members of the heterotrophic bacteria and often are

found in low numbers in drinking water. The risks these may pose to consumers has been

subject to much debate. This debate has centered on the importance of high numbers of

heterorophic (HPC) bacteria in drinking water. It has been suggested that drinking water

should not exceed 500 CFU per mL (EPA, 1989). This recommendation is based on the

findings that this level interferes with the detection of coliform bacteria which are used to

judge the sanitary quality of drinking water. No epidemiological data is available to

indicate that this or higher levels of HPC bacteria in drinking water are harmful. The

premise of this study is that the concentration of specific opportunistic pathogens are

more important than HPC bacteria as a whole. Thus, the occurrence of two common

HPC bacteria (A. hydrophila and P. aeruginosa) in various sources of drinking water

(point-of-use -POU- treated water, tap water with POU-connection. tap water, bottled

115

116

water, water from vending machines, and water from household storage tanks from

Mexico) were studied.

A. hydrophila has been isolated from humans since the early 1950s (Carnahan and

Altwegg, 1996). A. hydrophila is commonly isolated from patients with gastroenteritis.

It has also been found in wound and ocular infections (Janda and Abbot, 1996).

Aeromonads are oxidase positive, facultatively anaerobic, glucose fermenting, Gram-

negative bacilli (Hazen et al., 1978). They are a straight rod that measures 0.3 to 1.0 by

1.0 to 3.5 l_tm. Three species have been associated with human gastroenteritis, however

the exact mechanisms for diarrhea is not well defined (A. hydrophila, A. sobria, and A.

caviae) (Holmberg et al., 1986).

Pseudomonas species are Gram-negative, straight or slightly curved rods, motile

with polar flagella (Hoadley, 1977). The genus Pseudomonas contains well-recognized

pathogens (P. malle', P. cepacia, and P. aeruginosa). The source of infection is often

difficult to trace, but colonization has been associated with hospital equipment as well as

water-related reservoirs outside the hospital (i.e. swimming pools, hot tubs and contact

lenses ) (Rusin et al., 1997). P. aeruginosa can be introduced to susceptible tissues of

patients requiring instrumentation (i.e. catheter, aerators and respiratory therapy

equipment) posing a health threat (Fick, 1993).

Risk assessment involves four basic steps: 1) hazard identification; 2) dose-

response assessment; 3) exposure assessment; and, 4) risk characterization (NRC, 1983).

The purpose of the present study was to use the four-tiered approach to evaluate the

117

microbiological hazard from ingestion of drinking water. The data were obtained from

the previously cited studies of the occurrence of A. hydrophila and P. aeruginosa in

drinking water (POU-treated water, tap water with POU-connection, tap water, bottled

water, water vending machines and water storage tanks) (Chaidez et al., 1997; Chaidez et

al., 1998; Chaidez and Gerba, 1998). The risk assessment steps were followed to assess

the risks of acquiring an opportunistic bacterial infection (colonization) from

contaminated drinking water.

Materials and Methods

A. hydrophila and P. aeruginosa were selected for the study due to the importance

as opportunistic pathogens in drinking water and their common occurrence in drinking

water sources. The association of these organisms with water-related illness was also

evaluated.

Quantitative dose-response models were developed using information involving

human volunteer feeding studies (Bergogne-Berezin, 1994; Buck and Cooke, 1969).

Different mathematical models have been developed for evaluating microbial risk

assessment (Haas, 1983; Rose, 1997). However, the exponential model provided the best

fit-value to be used in estimating daily (Pi) and yearly (Pyea) risks of infection for the

study organisms (Rusin et al., 1997). The exponential model is as follow : Pi 1 - e

(1/10x(N)

118

The exposure assessment was conducted by collecting the data obtained in the

three previously cited studies (Chaidez et al., 1997; Chaidez et al., 1998; Chaidez and

Gerba, 1998). This information was then analyzed to perform the risk characterization.

Result

Hazard identification

Aeromonas hydrophila

Bacterial gastroenteritis due to A. hydrophila has been described (Moyer, 1987).

However, the exact mechanisms of diarrhea has not been elucidated. The virulence

factors that may be responsible include cytotoxin related to the cholera toxin, heat-labile

and heat-stable enterotoxins, hemagglutinins and hemolysin (Rusin et al., 1997). A.

hydrophda cause several types of clinical illnesses including diarrhea, wound infections,

endocarditis, meningitis, pneumonia, osteomyelitis, peritonitis, conjuntivitis,

thrombophlebitis, and cholecyititis (Janda et al., 1995; Janda and Duffey, 1988).

Septicemia due to A. hydrophda may be life-threatening. In the U.S. the overall

frequency is low (<1%) (Janda and Abbot, 1996). Common symptoms include fever (94-

95%) and chills (70%) (Harris et al., 1985). Patients who become septic with Aeromonas

often exhibit signs of gastrointestinal involvement, including abdominal pain, nausea,

vomiting and diarrhea (Dryden and Munro, 1989).

119

The health significance of Aeromonas in drinking water is not well understood as

few outbreaks are documented and establishing epidemiological links is difficult. Burke

et al., (1984) suggested a connection between cases of Aeromonas-associated diarrhea

and the numbers of Aeromonas in drinking water. It is accepted that some strains of

Aeromonas are enteropathogens possessing a range of virulence factors (enterotoxins,

cytotoxins, hemolysins and invasive ability) (Kirov, 1993) which can be expressed by

both environmental and clinical strains of Aeromonas (Cahill, 1990). Enteropathogens

strains of A. hydrophila, A. sobria and A. caviae have all been isolated from cases of

gastroenteritis (Deodhar et al., 1991). Ghanem et al., (1993) considered that potable

water was the source of Aeromonas infections after 90% of the samples were positive for

the organism. A. hydrophila isolated from a private unchlorinated well was proposed as

the cause of a long-term diarrhea in an 18-month-old child (Krovaceck and Peterz, 1989).

In a long-term care institution, there were 17 patients involved in acute diarrheal

episode presumably due to contaminated food or drinking water. Four patients were

positive for A. hydrophila, and a fatality was observed in one patient. It was concluded

that previous predisposing factors were associated with such fatality (Bloom and Bottone,

1990). Although Aeromonas are frequently isolated from drinking water systems, and

some strains may exhibit enterotoxic properties, further epidemiological studies are

required to ascertain whether there is any relationship between cases of Aeromonas-

associted diarrhea and the presence of these organisms in drinking water.

120

Pseudomonas aeruginosa

P. aeruginosa is the most important nosocomial pathogen responsible for 9% of

all hospital-acquired infections (Pollack, 1990). The mortality of nosocomial

pneumonias remains as high as 70% (Celis et al., 1988). Cystic fibrosis (CF) is the most

common lethal genetic disease of Caucasians, and chronic Pseudomonas respiratory

infections complicate 90% of CF patients leading to respiratory failure and death

(Gilligan, 1991). Hospitalization leads to greatly increased rates of carnage, especially in

patients with serious burns. Hoadley, (1977) reported that the rate of infection is about

25% in burn patients. Other infections occur due to P. aeruginosa including ear

infections (e.g. otitis externa), especially among swimmers (Hoadley, 1977).

The nutritional requirements of P. aeruginosa are extremely simple, favoring

growth in almost any moist environment (e.g. sinks, weak antiseptic solutions, eye drops,

anesthesia and resuscitation equipment, humidifiers, and stored distilled water)

(Steigbigel et al., 1968). P. aeruginosa may be transmitted by contaminated water, soil,

and vegetables. Kominos et al., (1972) demonstrated that vegetables may be the source

of P. aeruginosa in hospital kitchens. Shooter et al., (1969) implicated foods and

medicines as route of transmission in the hospital environment.

A group of people developed otitis after they were bathed in a wooden hot tub

containing P. aeruginosa (CDC, 1982). In 1947, an epidemic of acute P. aeruginosa

gastroenteritis was reported in a newborn nursery in which milk was implicated.

Symptoms were described as diarrhea, cramps, nausea, and vomiting, being more severe

121

in infants (Hunter and Ensign, 1947). Contaminated drinking water was traced as the

source of P. aeruginosa causing an outbreak of funicular sepsis. Ten newborn in a

hospital of Germany were infected by the bacterium (Weber et al., 1971). The

information of this study was scarce thus, the evidence lacks of credibility. P. aeruginosa

was isolated from a hospital tap water causing urinary tract infections, pneumonia and

sinusitis in nine hospitalized individuals (Bert et al., 1998).

The human body uses a complex cellular interaction to attack P. aeruginosa

infection. P. aeruginosa has a wide variety of virulence factors. These virulence factors

can be divided into two general groups: 1) cell surface components (pili,

lipopolysaccharide-LPS-) and 2) extracellular products (exotoxin A, proteases, and

lecithinase) (Peterson, 1980). No single host factor is of paramount importance in host

defense against P. aeruginosa which depends upon the integrated functions of the

epithelial cells barriers, antibodies, the complement system, phagocytic cells, and

lymphocytes (Peterson, 1980).

Dose-response

The development of a quantitative dose-response relationship is a primary step

in performing a risk analysis (Rose et al., 1991). The dose-response model describes

the quantitative relationship between the dose of a microorganism (the exposure) and

the response of the exposed population (the occurrence of infection). Human and

animal studies have been conducted to elaborate dose-response models (Macler, 1993).

122

In the environment, exposures to microorganisms are usually at doses that are too low to

measure via direct dose-response experiments (Rose et al., 1991). Thus, it is necessary

to use high concentration of pathogenic microorganism in order to observe high

frequencies of infections with a minimum numbers of volunteers. High exposures are

extrapolated down to doses that are representative of environmental scenarios by using

mathematical models (Rose and Gerba, 1991).

The exponential model provides the best fit exposure to A. hydrophila and P.

aeruginosa in drinking water (Rusin et al., 1997). The exponential model may be

expressed as:

Pi = 1 - el 4" )x(Nn

Where Pi is the probability of infection, 1/k is the fraction of ingested

microorganisms that survive to initiate infection, and N is the number of organisms

ingested. The model assumes a random distribution of microorganisms and that one

single microorganism is capable of initiating an infection (Haas, 1983). Yearly risk

were calculated using the following equation:

year = 1 - ( 1 - Pi)365

Exposure assessment

Aeromonas hydrophila

A. hydrophila is an ubiquitous bacterium frequently isolated from drinking water

and aquatic environments (Table 1). Chaidez and Gerba, (1998) found A. hydrophila in

123

6.6% of tap water samples at average concentration of 0.6 CFU/mL, and Levesque et

al., (1994) recovered the bacterium from 12% of the tap water samples. It has also been

recovered in 25% of distribution water samples at an average concentration of 0.001-94

CFU/mL (Millership and Chattopadhyay, 1985; Knocheland and Jeppesen, 1990;

Handfield et al., 1996; LeChevallier et al., 1982; Krovaceck et al., 1992). Taps with a

POU-connection (Faucet-mounted units) have shown to increase bacterial

concentrations when compared with tap water collected without filter device

connection. A. hydrophila was present in 10.5% of tap water samples with a POU-

connection at concentrations ranging from 0.006 to 0.14 CFU/mL (Chaidez and Gerba,

1998). POU-devices connected in households taps had A. hydrophila at concentrations

that ranged from 0.1 to 3.2 CFU/mL (Chaidez and Gerba, 1998). This data suggests

that POU-devices encourage the growth of A. hydrophila to levels greater than that

found in the tap water.

Bottled water has been shown to contain A. hydrophila. Chaidez and Gerba,

(1998) recovered the bacterium in 6.9% of the bottled water tested at concentrations

ranging from 0.1 to 0.3 CFU/mL. This excludes, one bottled water sample obtained

from Costa Rica which had a concentration of 90 CFU/mL. Others studies have

reported the isolation of this bacterium from bottled water, however the actual

concentrations were not provided (Tsai and Yu, 1997; Gonzalez et al., 1987; Manaia et

al., 1990; Slade et al.. 1986). The bacterium was present in 10% of water from vending

124

machines at concentration ranging from 0.03 to 0.2 CFU/mL (Chaidez and Gerba,

1998).

TABLE 1. Isolation of A. hydrophda from drinking water sources (CFU or %)

Reference

Bottled Tap Distribution Tap with POUwater water water POU treated

connection water

Chaidez and Gerba, 0.1-0.3/ 0.4/mL 0.006-1998 mL 0.14/mLLevesque et al., 1994 12%Tsai and Yu, 1997 2.3%Gonzalez et al., 1987 50%Manaia et al., 1990 3.5%Slade et al., 1986 43%Millership and 25%Chattopadhyay, 1985ICnochel and Jeppesen, 0.001-1990 19/mLHandfield et al., 1996 94/mLLeChevallier et al., 0.01-19/mL1982Krovaceck et al., 1992 8.6/mL

0.1-3.2/mL

P.aeruginosa

The results of previous surveys have showed the presence of P. aeruginosa in

drinking water (Table 2). Tap water was found to contain P. aeruginosa at

concentrations from 0.006 to 0.64 CFU/mL (Chaidez and Gerba, 1998). Edberg et al.,

(1996) found the bacterium in 2% of tap waters samples at concentrations of 2-16

CFU/mL. P. aeruginosa was found in 33.3% of tap water with a POU-connection

(Faucet-mounted units) at concentrations from 0.01 to 9 CFU/mL. The bacterium was

125

found in 38.5% of POU-devices mounted on household water taps at concentrations

from 0.008 to 10 CFU/mL (Chaidez and Gerba, 1998).

Chaidez and Gerba, (1998) found P. aeruginosa in 8.3% of bottled water tested

at concentrations from 0.01 to 0.2 CFU/mL. Fewtrell et al., (1997) reported that P.

aeruginosa was present in 12% of bottled water at an average concentration of 0.05

CFU/mL. Manaia et al., (1990) identified P. aeruginosa in 29% of the bottled waters

tested, and Edberg et al., (1996) found that 3% of bottled water contained the bacterium.

P. aeruginosa was found in water from vending machines at concentrations from 0.004

to 0.2 CFU/mL (Chaidez et al., 1997). Drinking water from storage tanks showed the

presence of the bacterium in 28% of the samples at concentrations from 0.002 to 0.6

CFU/mL (Chaidez et al., 1998).

TABLE 2. Isolation of P. aeruginosa from drinking water sources (CFU or %)

Reference Bottled Tap Tap with POU WVMa WSTb

water water POU treatedconnection water

Chaidez and 0.1- 0.006- 0.01-9/mL 0.008-Gerba, 1998 0.2/mL 0.64 10/mL

/mLEdberg et al., 3% 2-16/mL1996Fewtrell et al., 0.05/m1997Manaia et al., 29%1990Chaidez et al.,1997Chaidez et al.,1998

0.004-0.2/mL

0.002-0.6/mL

a Water vending machinesb Water storage tanks

126

During the exposure assessment, the amount of water consumed must also be

considered. An ingestion of 2 Upersoniday for the general population and 4

Li/person/day for the elderly for drinking water exposure was applied (Roseberry and

Burmaster, 1992).

Risk Characterization

Daily risks of colonization based on the consumption of 2L/day/person for A.

hydrophda ranged from 5x10 -1° for tap water with POU-connection to 2.2x10 -7 in

bottled water. On the annual basis the risks ranged from 1.8x10 -7 for tap water with

POU-connection to 7.9x10 -5 for bottled water (Table 3).

Daily risks of colonization based on the consumption of 2L/day/person for P.

aeruginosa ranged from 3x10 -7 for tap water to 2.7x10-6 for POU-treated water. On an

annual basis, the risks ranged from 4.8x10 -5 for water from storage tanks and to 9.9x10-4

for tap water with POU-connection (Table 3).

Also, the risk of colonization based on the consumption of 4L/day/person was

evaluated (Table 4). Daily risks for A. hydrophila ranged from 6.7x10 -9 for tap water to

4.3x10-7 for bottled water, and on the annual basis the risks ranged from 4x10-7 for tap

with POU-connection to 1.6x10 -4 for bottled water. A very low risk of colonization was

observed for bottled water excluding one sample containing high numbers of A.

127

hydrophila (Table 5). For P. aeruginosa. the daily risks ranged from 7.7x10' for water

from vending machines to 5.5x10 -6 for POU-treated water, and on the annual basis from

2.8x10 -6 for water from vending machines to 1.9x10 -3 for POU-treated water. Table 6

shows the bacterial concentration used for the risk of infection analysis.

TABLE 3. Risk of infection of Aeromonas hydrophila and Pseudomonas aeruginosa inselected drinking water sources (2L/day/person)

ORGANISMS RISK (DAY) RISK (YEAR)

POU-treated waterA. hydrophila 9.9x10-9 3.6x10-6P. aeruginosa 2.7x10-6 9.9x10'

Tap water (with POU-connection)A. hydrophila 5x10-1° 1.8x10-7P. aeruginosa 2.6x10-6 9.7x10-4

Tap water (without POU-connection)A. hydrophila 3.3x10' 1.2x10-6P. aeruginosa 3.9x10-7 1.4x10-4

Bottled waterA. hydrophila 2.2x10-7 7.9x10-5P. aeruginosa 1.2x10-7 4.4x10-5

Water from vending machinesA. hydrophila 1.6x10' 5.8x10-7P. aeruginosa 3x1 07 1.1x10-5

Water from storage tanksA. hydrophila Not available Not availableP. aeruginosa 1.3x10-7 4.8x10-5

TABLE 4. Risk of infection of A. hydrophila and P. aeruginosa in selected drinkingwater sources (4L/day/person)

128

ORGANISMS

A. hydrophilaP. aeruginosa

A. hydrophilaP. aeruginosa

A. hydrophilaP. aeruginosa

A. hydrophilaP. aeruginosa

A. hydrophilaP. aeruginosa

A. hydrophilaP. aeruginosa

RISK (DAY) RISK (YEAR)POU-treated water

1.9x10 -8 7.3x10-65.5x10 -6 1.9x10-3Tap water (with POU-connection)

1.1x10 -9 4x10-75.3x10 -6 1.95x10-3Tap water (without POU-connection)6.7x10 -9 2.4x10-67.8x10 -7 2.8x10-4

Bottled water4.3x10-7 1.6x10-42.4x10-7 8.8x10-5

Water from vending machines3.3x10 -9 1.2x10-67.7x10 -9 2.8x10-6

Water from storage tanksNot available Not available

2.6x10-7 9.6x10-5

TABLE 5. Risk of infection of A. hydrophila in bottled water excluding one high value*

RISK (DAY) RISK (YEAR)Bottled water2L/day/person

6.1x10-9 2.2x10-64L/day/person

1.2x10 -8 4.4x10-6*One bottle contained 90 CFU/mL

129

TABLE 6. Daily bacterial concentrations used for the risk of infection based on two andfour liter consumption

ORGANISMS Bacterial concentration

POU-treated waterTwo liters Four liters

A. hydrophila(57)a 6732`(118.1) b 13464(236.2)P. aeruginosa(57) 23340(409.5) 46680(818.2)

Tap water (with POU-connection)A. hydrophila(57) 396(6.9) 792(13.9)P. aeruginosa(57) 22800(400) 45600(800)

Tap water (without POU-connection)A. hydrophila(30) 1200(40) 2400(80)P. aeruginosa(30) 1752(58.4) 3504(116.8)

Bottled waterA. hydrophila(72) 185160(2571.6) 370320 (5143.3)P. aeruginosa(72) 1308(18.2) 2616(36.3)

Water from vending machinesA. hydrophila(30) 588(19.6) 1176(39.2)P. aeruginosa(30) 1380(46) 2760(92)

Water from storage tanksA. hydrophila(25) Not available Not availableP. aeruginosa(25) 496(19.8) 992(39.7)

a Total of number of samples analyzed (N value)Arithmetic averageHighest bacterial concentration

Discussion

Aeromonas hydrophila and Pseudomonas aeruginosa, are well-recognized

opportunistic pathogens and have been responsible for waterborne (Burke et al., 1984;

CDC, 1982) and foodbome (Bloom and Bottone, 1990; Hunter and Ensign, 1947)

outbreaks . Most of the HPC bacteria in drinking water are not human pathogens.

However, some bacteria including A. hydrophila and P. aeruginosa cause human

infections.

130

The microbial risk assessment approach attempts to address the uncertainity of

environmental contamination, generally at low levels of exposure and the potential for

resulting health effects. Although the assessment is made through the use of

assumptions, resulting in quantitations with large range of variation and uncertainity,

the method is useful for ranking the risks, comparing different environmental sources

and different approaches to control (Rose et al., 1991; Rose and Gerba, 1991).

A number of assumptions were used in the application of the selected

mathematical models to drinking water data. First random distribution (poisson)

governs the occurrence of the drinking water microorganisms, the population is equally

susceptible to a single exposure and the exposure is defined as the consumption of

2L/day/person or 4L/day/person (Rose and Gerba, 1991; Haas, 1993).

Risk estimates using existing data would tend to underestimate actual risks as

the culture methods for concentration of opportunistic pathogens are not 100% efficient

and will not detect viable but non-culturable organisms. Risks may be overestimated by

assuming a specific volume of ingestion (2 or 4 L of drinking water), especially for

developing countries where access to drinking water is often difficult (Gerba et al.,

1996; Crabtree, 1996). This assumption may not also be adequate for everyone, as

some persons may consume more or less.

One of the more controversial areas surrounding the modeling is the potential for

a single organism to initiate an infection. However, current scientific data support the

131

"independent-action" (or single-organism) hypothesis (Rubin, 1987). A single

bacterium, or virus, or protozoan can reproduce is a known biological phenomenon

which has been proven in laboratory studies. While it is clear that the host defense

(cellular and humoral immunity) does play a critical role in the determination of which

individuals may develop infection, it has been also suggested that these do not provide

the complete explanation (Rubin, 1987). The evaluation of the dose-response data sets

also support the independent-action hypothesis as in almost every case the exponential

or beta models provided a statistically significant improvement in fit over the log-

normal which could be used to predict a threshold (Haas, 1983). No experimental data

support a threshold phenomenon.

Conclusions

A concern has been generated regarding the health risks associated with

opportunistic bacterial pathogens in drinking waters. Opportunistic bacteria continue to

occur in the United States and other developed countries. It has been recommended that

the microorganisms present a risk no greater than 1:10,000 per year of infection from

drinking water (EPA, 1989). From this assessment, P. aeruginosa collected from POU-

treated water had the greatest risk of colonization (1.9:1,000 per year), and with a lesser

risk from A. hydrophila from any of the drinking water sources. However, the risk of

colonization by opportunistic bacterial pathogens is very low when compared with a

132

waterborne pathogen. Unknown agents, viruses and the protozoa rank as the most serious

threats to water quality. For instance, the annual risk of infection of rotavirus for the

general population is of 8.34x10 -1 (8.34:10 per year)(Crabtree, 1996).

Colonization of the gastrointestinal tract by P. aeruginosa does not confirm that is

a prerequisite for pneumonia or other associated infections. To further narrow the risk to

human health only certain specific hosts are at risks, including patients with profound

neutropenia, cystic fibrosis, severe burns, and those subject to foreign device installation

(Hardalo and Edberg, 1997). A. hydrophila isolated from drinking water rarely infect the

human gastrointestinal tract and the majority of Aeromonas strains occurring in drinking

water differ from those found in diarrhea stools (Havelaar et al., 1992). Apparently, A.

hydrophila which infect humans are from a great variety of environmental sources

including food.

In general, the risk of colonization from opportunistic pathogens is seen as a

transient process. The probability of infection may be very low, and the chances of

illness may be associated to the most vulnerable (very young, the elderly and

immunocompromised). Also, it must be kept in mind that the dose-response models

predict colonization and not disease.

Research needs

Future research should include the comparison of the risk data from various sources of

exposure. For instance , the risk of colonization from POU-treated water was 10 times

133

greater than tap water without POU-connection. Also, more surveys of drinking water

from various sources for specific opportunistic bacterial pathogens should need to be

conducted.

References of Appendix 4:

Altwegg, M., G. Martinetti, J. Luthy-hottenstein, J. and M. Rohrbach. 1991. Aeromonas-associated gastroenteritis after consumption of contaminated shrimp. Ezir. I Clin.Microbiol. Infect. Dis. 10:44-45.

Bergogne-Berezin, E. 1994. Acinetobacter spp., saprophytic organisms of increasingpathogenic importance. Zentralbl. Bakteriol. 281:389-405.

Bert, F., Maubec, E., Bruneau, B. and N. Lambert-Zechovosky. 1998. Multi-resistantPseudomonas aeruginosa outbreak associated with contaminated tap water in aneurosurgery intensive care unit. J. Hosp. Infect. 39:53-62.

Bloom, H.G. and E.J. Bottone. 1990. Aeromonas hydrophila diarrhoea in a long-termcare setting. J. Amer. Geriatr.Soc. 38:804-806.

Buck, A.C. and E.M. Cooke. 1969. The fate of ingested Pseudomonas aeruginosa innormal persons. J. Med. Microbiol. 2:521-525.

Burke, V., Robinson, J., Gracey, M., Peterson, D. Meyer, N. and V. Haley. 1984.Isolation of Aeromonas spp. from unchlorinated domestic water supply. Appl. Environ.Microbiol. 48:367-370.

Carnahan, A.N. and M. Altwegg. 1996. Taxonomy. p.1. In Wiley (ed.) The genusAeromonas. Chichester, England.

Cahill, M.M. 1990. Virulence factors in motile Aeromonas species. J. Appl. Bacteriol.69:1-16.

Center for Disease Control. 1982. Otitis due to Pseudomonas aeruginosa 0:01 associatedwith a mobile redwood tub system-North Carolina. Morb. Mortal. Weekly Report.31:541-542.

134

Celis, R., Tones, A., Gate11, J.M. and M. Alnela. 1988. Nosocomial pneumonia amultivariate analyses of risk and prognosis. Chest. 93:318-324.

Chaidez, C., Candil, A., Uribe, M. and C.P. Gerba. 1998. Microbiological survey ofprivate roof water tanks in Culiacan, Mexico. J Environ. Sci. Hlth. In press.

Chaidez, C. and C.P. Gerba. 1998. Microbiological comparison of the quality of POU-treated water, tap water, and bottled water. University of Arizona, Tucson, AZ.(unpublished report).

Chaidez, C., Rusin, P.A., Naranjo, J. and C.P. Gerba. 1997. Microbiological quality ofwater vending machines. Int. J Food Microbial. In press.

Crabtree, K.D. 1996. Risk assessment of viruses in water. Ph.D. dissertation. Universityof Arizona. Tucson, AZ.

Deodhar, L.P., Saraswathi, K. and A. Vardukar. 1991. Aeromonas spp. and theirassociation with human diarrheal diseases. J. Clin. Microbial. 29:853-856.

Dryden, M. and R. Munro. 1989. Aeromonas septicemia: relationship of species andclinical features. Pathol. 21:111-114.

Edberg, S.C., Gallo, P. and C. Kontnick. 1996. Analysis of the virulence characteristicsof bacteria isolated from bottled water, water cooler, and tap water. Microbial EcoL Hlth.Dis. 9:67-77.

Environmental Protection Agency (EPA). 1989. Drinking water health effects task force.Health effects of drinking water treatment technologies. Lewis Pubs. Chelsea, MI.

Fewtrell, L., Kay, D., Wyer, M., Godfree, A. and G. O'neall. 1997. Microbiologicalquality of bottled water. Wat. Sci. Tech. 35:47-53.

Fick, R.B. 1993. Pseudomonas aeruginosa: The opportunist. CRC press.

Flynn, T.J. and I.G. Knipp. 1995. Seafood shucking as an etiology for Aeromonashydrophila infection. Arch. Intern. Med. 147:1816-1817.

Gerba, C.P. and C. N. Haas. 1988. Assessment of risks associated with enteric viruses incontaminated drinking water. American society for testing and materials. Philadelphia.

Gerba, C.P., J.B. Rose, C.N. Haas and K.D. Crabtree. 1996. Waterborne rotavirus: a riskassessment. Wat. Res. 30:2929-2940.

135

Ghanem, E.H., Mussa, M.E. and H.M. Eraki. 1993. Aeromonas-associated gastroenteritisin Egypt. Zentralbl. Mikrobiol. 148:441-447.

Gilardi, G.L. 1980. Medical Microbiology. p. 25-30. In Hans Huber (ed.) Pseudomonasaeruginosa, the organism. Switzerland.

Gilligan, P.H. 1995. Pseudomonas and Burkholderia. p. 509-519. In ASM press (ed.)Manual of clinical microbiology. Washington, DC.

Gilligan, P.H. 1991. Microbiology of airway disease in patients with cystic fibrosis. Clin.Microbiol. Rev. 4:35-51.

Glover, D., Ross, A. and J. Lugsdin. 1992. Gastroenteritis outbreak at an industrial camp-British Columbia. Can. Commun. Dis. Rep. 18:66-68.

Gonzalez, C. Gutierrez, C. and T. Grande. 1987. Bacterial flora in bottled uncarbonatedmineral drinking water. Can. I Mierobiol. 33:1120-1125.

Haas, C.N. 1983. Estimation of risk due to low doses of microorganisms: a comparison ofalternatives methodologies. Amer. I EpidemioL 118:537-582.

Handfield, M., Simard, P., Coullard, M. and R. Letarte.1996. Aeromonas hydropilaisolated from food and drinking water: Hemaglutination, hemolysis, and cytotoxity for ahuman intestinal cell line (HT-29). App!. Environ. Microbiol. 62:3459-3461.

Hardalo, C. and S.C. Edberg. 1997. Pseudomonas aeruginosa: Assessment of risk fromdrinking water. Crit. Rev. Microbiol. 23:47-75.

Harris, R.L., Fainstein, V., Elting, L., Hopfer, R.L. and G.P. Bodey. 1985. Bacteremiacaused by Aeromonas species in hospitalized cancer patients. Rev. Infect. Dis. 7:314-320.

Hazen, T.C., Flierman, G.B., Hirsch, R.P. and G.W. Esch. 1978. Prevalence anddistribution of Aeromonas hydrophila in the United States. App!. Environ. Microbiol.36:731-738.

Havelaar, A.H., Schets, F.M., Van Silfhout, A., Jasen, W.H., Wieten, G. and D. Van DerKooij. 1992. Typing Aeromonas strains from patients with diarrhoea and from drinkingwater. I Appl. Bacteriol. 72:435-444.

136

Hoadley, A. W. 1977. Pseudomonas aeruginosa in surface waters. p. 31-57. In young,V.M. Raven press. (ed.) Pseudomonas aentginosa: ecological aspects and patientscolonization. New York, NY.

Holmberg, S.D., Shell, W.L., Fanning, G.R., Wachsmoth, I.K., Hiickman, Brenner, F.W.and P.A. Blake. 1986. Aeromonas intestinal infections in the United States. Ann. Intern.Med. 105:109-114.

Hunter, C.A. and P.R. Ensign. 1947. An epidemic of diarrhea in a newborn nurserycaused by Pseudomonas aeruginosa. Amer. J. Public Hlth. 37:1166-1169.

Janda, J.M. and S.L. Abbott. 1996. Human pathogens. p. 151-173. In Wiley (ed.) Thegenus Aeromonas. Chichester, England.

Janda, J.M., Abbott, S.L. and A.M. Carnahan. 1995. Aeromonas and Plesiomonas. p.477-482. In ASM press (ed.) Manual of clinical microbiology. Washington, DC.

Janda, J.M. and P.S. Duffey. 1988. Mesophilic Aeromonads in human disease: Currenttaxonomy, laboratory and infections disease spectrum. Rev. Infect. Dis. 10:980-997.

Kirov, S.M. 1993. The public health significance of Aeromonas spp. in foods. Mt. J.Food Microbiol. 20:179-198.

Knochel, S. and C. Jeppesen. 1990. Distribution and characteristics of Aeromonas in foodand drinking water in Denmark. Mt. J Food Microbiol. 10:317-322.

Kominos, S.D., Copeland, C.E., Grosiak, B. and B. Postic. 1972. Introduction ofPseudomonas aeruginosa into a hospital via vegetables. Appl. Microbiol. 24:567-570.

Krovaceck, K. Dumontet, S., Ericksson, E. and S.B. Balada. 1995. Isolation and virulenceprofiles of Aeromonas hydrophila implicated in an outbreak of food poisoning inSweden. Microbiol. Immunol. 39:655-661.

Krovaceck, K., Faris, A., Baloda, S.B., Lindberg, T., Peterz, M. and I. Mansson. 1992.Isolation and virulence profiles of Aeromonas spp. from different municipal drinkingwater supplies in Sweden. Food Microbiol. 9:215-222.

Krovaceck, K. and M. Peterz. 1989. Enterotoxigenicity and drug sensitivity ofAeromonas hydrophila isolated from well water in Sweden: a case study. Mt. J. FoodMicrobiol. 8:149-154.

137

LeChevallier, M.W., Evans, T.M., Seidler, R.J., Daily, 0.P., Merrell, B.R., Rollins, D.M.and S.W. Joseph. 1982. Aeromonas sobria in chlorinated drinking water supplies.Microb. Ecol. 8:325-333.

Levesque, B., Pierre, S., Gauvin, D., Gingas, S., Dewailly, E. and R. Letarte. 1994.Comparison of the microbiological quality of water coolers and that of municipal watersystems. Appl. Environ. Microbiol. 60:1171-1178.

Macler, B.A. and S. Regli. 1993. Use of microbial risk assessment in setting UnitedStates drinking water standards. Int. J. Food Microbiol. 18:245-256.

Manaia, C.M., Nunes, 0.C., Morais, P.V. and M.S. Da Costa. 1990. Heterotrophic platecounts and the isolation of bacteria from mineral waters on selective andenrichment media. J. Appl. Bacteriol. 69:871-876.

Millership, S.E. and B. Chattopadhyay. 1985. Aeromonas hydrophila in chlorinated watersupplies. J. Hosp, Infect. 6:75-80.

Moyer, N.D. 1987. Clinical significance of Aeromonas species isolated from patientswith diarrhea. J. Clin. Microbiol. 25:2044-2048.

National Research Council. 1983. Risk assessment in the federal government: managingthe process. National Academy Press, Washington, DC.

Peterson, P.K. 1980. Host defense against Pseudomonas aeruginosa. p. 103-118. In HansHuber (ed.) Pseudmonas aeruginosa, the organism. Switzerland.

Pollack, M. 1990. Pseudomonas aeruginosa. p. 1673-1687. In Mandell, G.H., DouglasR.G. Jr., Benett, J.E. (ed.) Principles and practice of infectious diseases. Livingstone, NY.

Regli, S., Rose, J.B., Haas, C.N. and C.P. Gerba. 1991. Modeling the risk from Giardiaand viruses in drinking water. J. Amer. Water Works. Assoc. 83:76-84.

Rose, J.B. and C.P. Gerba. 1991. Use of risk assessment for the development of microbialstandards. Wat. Sci. Tech. 24:29-34.

Rose, J.B., Haas, C.N. and S. Regli. 1991. Risk assessment and control of waterbornegiardiasis. Amer. J. Pub. Hlth. 1:709-713.

Rose, J.B. 1997. Environmental ecology of Cryptosporodium and public healthimplications. Annu. Rev. Public Hlth. 18:135-161.

138

Roseberry, A.M. and D.E. Burmaster. 1992. Lognormal distribution for water intake bychildren and adults. Risk Anal. 12:99-104.

Rubin, L.G. 1987. Bacterial colonization and infection resulting from multiplication of asingle organism. Rev. Infect. Dis. 9:488-493.

Rusin, P.A., Rose, J.B., Haas, C.H. and C. P. Gerba. 1997. Risk assessment ofopportunistic bacterial pathogens in drinking water. Rev. Environ. Contam. Toxicol.152:57-83.

Sims, R.V., Hauser, R.J., Adewale, A.O., Maislin, G., Skeie, S., Lavizzo-Mourey, R.J.and H. Rubin. 1995. Acute gastroenteritis in three community-based nursing homes. J.Gerontol. A. Biol. Sci. Med. Sci. 50:M252-256.

Shooter, R.A., Cooke, E.M., Gaya, H., Kumar, P., Patel, N., Parker, T., Tham, T. andD.R. France. 1969. Food and medicaments as possible sources of hospital strains ofPseudomonas aeruginosa. LANCET. 1:1227-1229.

Slade, P.J., Falah, M.A. and A.M. Al-Ghady. 1986. Isolation of Aeromonas hydrophilafrom bottled waters and domestic water supplies in Saudi Arabia. J. Food Protect.49:471-476.

Steigbigel, N.H., Kislak, J.W., Tilles, J.G. and M. Finland. 1968. Clinical evaluation ofcephaloridine. Arch. Mt. Med. 121:24-38.

Tsai, G.J. and S.C. Yu. 1997. Microbiological evaluation of bottled water uncarbonatedmineral water in Taiwan. Mt. J. Food Microbiol. 37:137-143.

Weber, G., Werner, H.P. and H. Matsching. 1971. Death cases in newborns caused byPseudomonas aeruginosa contaminated drinking water. Zentralbl. Bakteriol. 216:210-214.

APPENDIX 5:

CONCLUSION

Cristobal Chaidez Quiroz

Department of Nutritional SciencesThe University of Arizona

Tucson, AZ. 85721

139

140

Microbial risk assessment is relatively new and has many potential applications in

risk management decisions. It can be used to evaluate treatment options and objectives,

guidelines and standards, as well as cost benefit analysis, and to competing risks. The

presented study in this dissertation shows how microbial risk assessment can be applied

to microbial water quality. Data from various drinking water sources were obtained for

microbial risk assessment. Quantitative microbial risk assessment was applied to offer

risk managers another means of interpreting water quality data. Attending the needs for

specific occurrence data, opportunistic bacterial pathogens were identified from point-of-

use (POU)-treated water, tap water with POU-connection, tap water, bottled water, and

water from vending machines and households storage tanks. The data were designed to

be used in the microbial risk assessment. The opportunistic bacterial pathogens, as

potential waterborne pathogens, were placed in a framework for use in risk

characterization.

Microbial risk assessment has only been used on a limited scale for assessing the

risk of infection from waterborne pathogens. Dose-response models for two specific

opportunistic bacterial pathogens (Aeromonas hydrophda and Pseudomonas aeruginosa)

have previously modeled (Rusin et al., 1997), and were applied to perform the risk

characterization in this study. The selected bacterial pathogens belong to the

141

heterotrophic plate count (HPC) bacteria and are commonly found in water. However,

there is no clear-cut evidence that EIPC bacteria as a whole pose a significant public

health risk. Therefore, it is necessary to evaluate specific opportunistic bacterial

pathogens when assessing drinking water quality. Opportunistic bacterial pathogens

continue to occur in drinking water and may pose a threat to segments of the population.

Opportunistic bacterial pathogens requires very specific immune defects in order for

people to become infected (colonized). Infection (colonization) outside these well-

defined groups are very rare. Therefore, under normal circumstances, the risk to the

general population from opportunistic bacteria is not significant. However in the specific

compromised patient, prevention, or early detection, and eradication of colonization are

likely to be more effective than attempting to control the number of opportunistic bacteria

due to the ubiquitous nature of the species, it is not practical to eliminate the species from

the drinking water.

More data is needed on the occurrence of specific opportunistic bacterial

pathogens in selected drinking water sources to fully assess the risks of HPC bacteria in

drinking water. Conventional methods as well as Molecular approaches should be

applied to identify clinical and environmental isolates. Although the assessment is made

based on assumptions. resulting in a large range of variations and uncertainity, the use of

microbial risk assessment approach can provide a better understanding of the public

health significance of the occurrence of bacteria in drinking water.

APPENDIX 6:

BACTERIAL DATA

Cristobal Chaidez Quiroz

Department of Nutritional SciencesThe University of Arizona

Tucson, AZ. 85721

142

Occurrence of bacteria in water vending machines (WVM)

WVM Pseudomonasaeruginosa

Aeromonashydrophila

Plesiomonasshigelloides

Acid-fastorganisms

Totalcoliforms

Heterotrophicplate count

CFUa/ 500 mL CFU/ mL1 0 0 0 0 6002 0 0 0 0 503 0 0 0 + 0 20,0004 0 0 0 0 12,0005 0 0 0 0 9,2006 0 0 0 400 1,7207 0 0 0 0 10,0008 0 0 0 0 9009 0 0 0 + 600 2,00010 0 0 0 0 6,08011 0 0 0 + 37 1,00012 0 0 0 + 0 24013 0 0 0 561 2,30014 0 0 0 + 0 4,60015 0 100 0 + 0 30016 10 0 0 + 100 38017 250 0 0 + 21 1,80018 0 0 0 + 150 6019 0 0 0 14 19020 0 17 0 + 1 60021 0 0 0 + 0 20022 0 0 0 0 86023 5 0 0 + 0 6024 50 0 0 1,000 98025 0 0 0 0 1,90026 30 30 0 0 1,30027 0 0 0 + 0 60028 0 0 0 + 0 18,00029 0 0 0 + 0 60030 0 0 0 3 300

CFU, colony-forming unitsPositives samples based on acid-fast stain

143

144

Occurrence of bacteria in water storage tanks (WST)

Vv'STP. aeruginosaCFUa/100mL

Total coliformsCFU/100rnL

Fecal coliformsCFU/100mL

HPC bacteriaCFU/mL

1 0 1 0 625001 0 2 0 195003 0 0 0 725004 100 1 1 224000:_ 3 0 0 170006 2 0 0 5000007 0 20 20 2000008 10 10 10 2180009 3 0 0 40000010 5 120 0 63001 1 0 0 0 2300012 0 0 0 2000013 0 0 0 125014 0 0 0 14200015 0 0 0 155016 0 0 0 300017 1 0 0 15200018 0 15 15 10000019 0 0 0 30200020 0 15 3 600021 0 0 0 271022 0 0 0 56023 0 0 0 880024 0 0 0 800025 0 0 0 2500

' CFU, colony-forming units

Occurrence of bacteria in Bottled water (BW)(Mexico, Costa Rica, Panama, Japan, Korea, and Greece)

BW P.aeruginosa

A.hydrophila

P. Acid fastshieelloides organism'

Totalcoliforms

HPC6

OE -Pi 500 mL CFU/100 mL CFU/ mL1 0 0 0 0 4302 0 0 0 0 2,8003 0 0 0 0 3,0004 0 0 0 19 16,0005 0 0 0 0 206 0 0 0 0 10,0007 0 0 0 0 1308 0 0 0 - 0 80,0009 0 0 0 - 0 60,00010 0 0 0 0 011 0 0 0 0 30,00012 0 0 0 - 0 10,00013 0 0 0 0 014 0 0 0 0 1,00015 100 0 0 0 1,00016 0 45,000 0 - 18 72,00017 0 150 0 - 0 12,00018 0 0 0 0 1,52219 0 90 0 0 25,00020 0 0 0 0 3,00021 0 0 0 0 14,00022 0 0 0 - 0 1,60023 0 0 0 - 0 6,40024 0 0 0 1 14025 7 0 0 600 10,00026 0 0 0 500 70,00027 60 0 0 - 680 19,00028 50 0 0 23 2,00029 0 0 0 0 2030 0 0 0 300 90,00031 100 1,000 0 0 16,00032 0 0 0 - 0 90,00033 0 0 0 - 0 20034 0 0 0 - 0 24,00035 0 0 0 - 0 3036 0 0 0 0 9,60037 0 0 0 - 35 100

a CFU, colony-forming unitsb HPC, Heterotrophic plate count

Positives samples based on acid-fast stain

145

Occurrence of bacteria in bottled water (BW)(United States)

(BW) P.aeruginosa

A.hydrophila

P.shigelloides

Acid-fastorganismsb

Totalcoliforms

Heterotrophicplate count

CFliaf 500 mL CFUI 100mL CFU/mL1 0 0 0 0 4,2002 0 0 0 0 1,1003 0 0 0 2 2,0004 0 0 0 0 15 0 0 0 0 106 0 0 0 0 207 0 0 0 -4- 0 1,9008 0 0 0 0 1,2009 0 0 0 + 0 0lo 0 0 0 + 0 2011 0 0 0 0 1,30012 0 0 0 0 4,00013 0 0 0 0 1014 0 0 0 + 0 3015 0 0 0 0 15016 10 0 0 0 28,00017 0 0 0 0 6018 0 0 0 0 4019 0 0 0 0 1020 0 0 0 0 2021 0 0 0 0 1,00022 0 0 0 0 1,20023 0 0 0 0 10,08024 0 0 0 0 30025 0 50 0 0 9,20026 0 0 0 0 90027 0 0 0 0 2,80028 0 0 0 0 20,00029 0 0 0 0 3,00030 0 0 0 0 26,80031 0 0 0 0 3032 0 0 0 0 033 0 0 0 0 3,60034 0 0 0 0 6,20035 0 0 0 - 60 8,000

CFU, colony-forming unitsb Positives samples based on acid-fast stain

146

Occurrence of bacteria in tap water (TW)(TUCSON, AZ)

TW P.aeruginosa

A.hydrophila

P. Acid fastshigelloides organisms .'

Totalcoliforms

Heterotrophicplate count

CFUa/500

Week #1

CFU/100mL

CFU/mL

1 0 0 0 0 202 0 0 0 0 3003 0 0 0 0 104 0 0 0 + 0 305 0 0 0 0 60

Week #26 0 0 0 + 0 38,0007 0 0 0 0 40,0008 0 0 0 0 50,0009 0 0 0 60 4,00010 0 100 0 80 50,000

Week #311 0 0 0 - 0 50,00012 0 0 0 0 20,00013 100 0 0 0 12,00014 0 0 0 0 15015 0 0 0 0 100

Week #416 10 0 0 0 13017 0 0 0 - 0 20018 0 0 0 - 0 30019 320 0 0 0 15020 0 0 0 0 20

Week #521 0 0 0 0 4,00022 0 0 0 0 15023 5 0 0 + 0 5,20024 0 0 0 + 0 10025 0 0 0 0 1,600

Week #626 0 0 0 0 35027 0 200 0 0 2028 0 0 0 0 6029 0 0 0 - 0 1,00030 3 0 0 0 1,200

a CFU, colony-forming unitsh Positives samples based on acid-fast stain

147

WEEK#1Occurrence of bacteria in tap water with POU-connection

sample P.aeruzinosa

.4.hydrophda

P.shigelloides

Acid-fastorganisms'

Totalcoliforms

Heterotrophicplate count

CFUa/500 mL CFU/100 mL

CFU/rnL

unit-1 0 0 0 - 80 350

unit-2 0 70 0 - 30 70

unit-3 0 0 0 + 10 30,000

unit-4 0 0 0 0 250

unit-5 0 0 0 30 1,200

unit-6 0 0 0 - 15 400

unit-7 0 0 0 0 1,000

unit-8 20 0 0 - 3 200

unit-9 760 0 0 - 0 1,090

unit-10 0 0 0 - 0 40

WEEK#2Occurrence of bacteria in tap water with POU-connection

sample P.aeruginosa

A.hydrophila

P.shigelloides

Acid-fastorganisms'

Totalcoliforms

Heterotrophicplate count

CHP/500 mL CFU/100niL

CFU/mL

unit-1 0 0 - 0 - 0 600

unit-2 0 3 0 - 0 150

unit-3 14 0 0 .,- 0 1,150

unit-4 0 3 0 0 150

unit-5 0 0 0 - 100 530

unit-6 0 0 0 160 2,200

unit-7 0 0 ' 0 + 0 1,300

unit-8 80 0 0 - 30 200

unit-9 100 0 0_ - 0 10,000,000

unit-10 0 0 0 - 0 250

CFU, colony-forming unitsPositives samples based on acid-fast stain

148

VvIEK#3Occurrence of bacteria in tap water with POU-connection

sample P.aeruginosa

A.hydrophda

P.shigeiloides

Acid-fastorganisms'

Totalcoliforms

Heterotrophicplate count

CFUa/500 mL CFU/100raL

CFU/ML

unit-1 0 0 0 A- 0 450

unit-2 50 0 0 - 0 450

unit-3 0 0 0 5 400

unit-4 0 0 0 8 500

unit-5 0 0 0 5 10,800

unit-6 50 0 0 -,- 0 10,000,000

unit-7 0 0 0 0 1.000,000

unit-8 240 0 0 - 0 900

unit-9 3000 0 0 - 0 10,000,000

unit-10 0 0 0 - 0 300

WEEK#4Occurrence of bacteria in tap water with POU-connection

sample P.aeruginosa

A.hydrophda

P.shigelloides

Acid-fastorganisms'

Totalcoliforms

Heterotrophicplate count

CFUa/500 mL CFU/100mL

CFU/mL

unit-1 620 0 0 0 700

unit-2 0 0 0 - 0 20,000

unit-3 0 0 0 + 0 400unit-4unit-5 0 0 0 + . 10 3,600

unit-6 0 0. 0- + 0 520

unit-7 0 0 0 + 8 400

unit-8 6 0 0 0 600

unit-9 200 0 0 - 0 1,550

unit-10 0 0 0 - 0 1,900

CFU, colony-forming unitsb Positives samples based on acid-fast stain

149

WEEK#5Occurrence of bacteria in tap water with POU-connection

sample P.aeruginosa

.4.hydrophda

P.shigelloides

Acid-fastorganisms b

Totalco1iforms

Heterotrophicplate count

CFL;a/500 mL CPU/100mL

CFU/mL

unit-1 0 0 0 0 1,500

unit-2 0 5 0 - 0 200

unit-3 0 0 0 - 0 4,000

unit-4unit-5 30 0 0 - 10 2,000

unit-6 200 0 0 - 0 3,500

unit-7 0 0 0 - 0 380

unit-8 10 0 0 - 0 500

unit-9 300 0 0 - 57 2,500

unit-10 0 0 0 - 0 100

WEEK#6Occurrence of bacteria in tap water with POU-connection

sample P.aeruginosa

A._ hydrophila

P.shigelloides

Acid-fastorganisms b

Totalconforms

Heterotrophicplate count

CFtia/500 mL CFU/100rnL

CFU/mL

unit-1 0 0 0 - 30 450unit-2 0 10 0 - 0 20,000,000

unit-3 0 0 0 + 0, 100

unit-4 .unit-5 60 8 0 - 320 14,000

unit-6 80 — 0 0 -_

0 16,000unit-7 0 0 0 - - 0 1,600

unit-8 0 0 0 - 0 4,000

unit-9 120 - 0 0 - 15 16,000unit-10 0 0 - 0 - 0 200

CFU, colony-forming unitsb Positives samples based on acid-fast stain

150

WEEK#1Occurrence of bacteria in Point-Of-Use (POU)-treated water

sample P.aeruginosa

A.hydrophila

P.shigelloides

Acid-fastorganisms'

Totalcoliforms

Heterotrophicplate count

CFUN500 mL CFU/100mL

CFU/mL

unit-1 0 0 0 100 700

unit-2 0 0 0 - 14 400

unit-3 0 0 0 - 16 10,000

unit-4 0 0 0 + 5 150unit-5 0 0 0 - 260 30,000,000

unit-6 30 0 ' 0 340 2,000,000

unit-7 0 0 0 - 0 2,600unit-8 0 0 0 - 240 1,200unit-9 240 0 0 - 0 4,000

unit-10 400 0 0 + 100 120

WEEK#2Occurrence of bacteria in Point-Of-Use (POU)-treated water

sample P.aeruginosa

A.hydrophila

P.shigelloides

Acid-fastorganisms'

Totalcoliforms

Heterotrophicplate count

CFUa/500 rnL CFU/100mL

CFU/mL

unit-1 0 () 0 - 200 80,000unit-2 4 2 8 0 - 70 30,000unit-3 25 0 0 + 35 30,000,000

unit-4 0 0 0 + 62 40,000unit-5 150 0 0 — 140 100

unit-6 74 0 0 + 260 7,400unit-7 0 0 0 - 0 30,000,000unit-8 0 0 0 - 145 1,400unit-9 100 0 0. -. 0 510

unit-10 0 0 0 - 150 4,000CFU, colony-forming units

b Positives samples based on acid-fast stain

151

"vVEEK#3Occurrence of bacteria in Point-Of-Use (POU)-treated water

sample P.aeruginosa

A.hydrophda

P.shigelloides

Acid-fastorganisms'

Totalcoliforms

Heterotrophicplate count

CFUa/500 mL CFU/100mL

CFU/mL

unit-1 0 0 0 — 800 8,000

unit-2 15 0 0 — 35 350

unit-3 100 0 0 180 4,000unit-4 0 0 0 62 900

unit-5 0 0 0 - 10 4,700

unit-6 100 0 0 161 28,000

unit-7 0 0 0 - 507 900

unit-8 0 0 0 - 3 750

unit-9 3500 0 0 - 0 4,000

unit-10 0 0 0 -1- 480 3,200

WEEK#4Occurrence of bacteria in Point-Of-Use (POU)-treated water

sample P.aeruginosa

A.hydrophila

P.shigelloides

Acid-fastorganisms'

Totalcoliforms

Heterotrophicplate count

CFUa/500 mL CFU/100mL

CFU/mL

unit-1 0 0 0 0 4,560

unit-2 6 0 . 0 - 10 25,000,000

unit-3 0 0 - 0 - 10 1,800unit-4

unit-5 0 0 0, 80 6,400

unit-6 10 0 0, + 20 40,000,000

unit-7 0 0 0 - 130 1,000unit-8 0 0 0 - 62 2,000

unit-9 190 0 0 - 0 8,800unit-10 0 0 0 - 60 7,600CFU, colony-forming units

" Positives samples based on acid-fast stain

152

WEEK#5Occurrence of bacteria in Point-Of-Use (POU)-treated water

sample P.aerueinosa

A.hvdrophila

P.shigelloides

Acid-fastorganisms'

Totalcoliforms

Heterotrophicplate count

CR. /500 m_L CFU/100mL

CFU/mL

unit-1 10 0 0 ± 17 500unit-2 100 25 0 500 2,500unit-3 0 0 0 - 25 5,600unit-4

unit-5 0 0 0 194 18,000unit-6 100 0 0 + 32 15,000unit-7 0 0 0 100 1.200unit-8 0 0 0 + 0 10,000,000unit-9 266 0 0 - 0 400

unit-10 0 0 0 -t- 300 6,000

WEEK#6Occurrence of bacteria in Point-Of-Use (POU)-treated water

sample P.aeruginosa

A.hydrophila

P.shigelloides

Acid-fastorganisms'

Totalcoliforms

Heterotrophicplate count

CFUa/500 mL CFU/100rnL

CFU/mL

unit-1 0 0 0 + 130 12,000unit-2 100 1,600 0 - 50 20,000,000unit-3 0 0 0 - 0 1,000unit-4

unit-5 0 30 0 - 1,000 16,800unit-6 120 0 0 610 16,400unit-7 0 0 0 - 84 2,400unit-8 0 0 0 - 5 4,400unit-9 180 0 0 - 30 2,000,000

unit-10 0 0 0 - 80 400CFU, colony-forming unitsPositives samples based on acid-fast stain

153

154

REFERENCES

Alonso, J.L., Garay, E. and E. Hernandez. 1989. Membrane filter procedure for theenumeration of Pseudomonas aeruginosa in water. Wat. Res. 23:1499-1502.

Altwegg, M., Reeves, R., Altwegg-Bissig, R. and D.J. Brenner. 1991. Multilocus enzymeanalysis of the genus Aeromonas and its use for specific identification. Zentralbl.Bakteriol. 257:28-45.

American Public Health Association (APHA). 1995. Standard Methods for theExamination of Water and Wastewater, 18 ed. Washington, DC.

Arai, T., Ikejima, N., Itoh, T., Sakai, S., Shimada, T. and R. Sakazaki. 1980. A survey ofPlesiomonas shigelloides from aquatic environments, domestic animals, pets andhumans. 1 Hyg. Camb. 84:203-211.

Bermudez, L.E., Petrotsky, M., Koloski, P. and L.S. Young. 1992. An animal model ofMycobacterium avium complex disseminated infection after colonization of the intestinaltract. J. Infect. Dis. 165:75-79.

Bergogne-Berezin, E. 1994. Acinetobacter spp., saprophytic organisms of increasingpathogenic importance. Zentralbl. Bakteriol. 281:389-405.

Bert, F., Maubec, E., Bruneau, B. and N. Lambert-Zechovosky. 1998. Multi-resistantPseudomonas aeruginosa outbreak associated with contaminated tap water in aneurosurgery intensive care unit. J. Hosp. Infect. 39:53-62.

Bitton, G. 1994. Wastewater Microbiology. Wiley-Liss. NY.

Bodey, G.P., Bolivar, R., Faisten, V. and L. Jadeja. 1983. Infections caused byPseudomonas aeruginosa. Rev. Infect. Dis. 5:279-313.

Brown, L.M. and J. Gauthier. 1993. Cell density and growth phase as factor in theresistant of a biofilm of Pseudomonas aeruginosa (ATCC 27853) to iodine. Appl.Environ. Microbiol. 59:2320-2322.

Buck, A.C. and E.M. Cooke. 1969. The fate of ingested Pseudomonas aeruginosa innormal persons. J. Med. Microbiol. 2:521-525.

Cabelli, V.J. 1978. Swimming-associated disease outbreaks. Water Poll. Control Fed. J50:1374-1377.

155

Caroli, G., Levre, E., Armani. G., Biffi-gentili, S. and G. Molonari. 1985. Search for acid-fast in bottled water. J. Appl. Bacteriol. 58:461-464.

Celis, R., Torres, A., Gatell, J.M. and M. Alnela. 1988. Nosocomial pneumonia amultivariate analyses of risk and prognosis. Chest. 93:318-324.

Center for Disease Control. 1982. Otitis due to Pseudomonas aeruginosa 0:10 associatedwith a mobile redwood tub system-North Carolina. Morbid. Mortal. Weekly Rep. 31 :541-542.

Chaidez, C. and C.P. Gerba. 1998. Microbiological comparison of the quality of POU-treated water, tap water, and bottled water. University of Arizona, Tucson, AZ.(unpublished report).

Chaidez, C., Rusin, P.A., Naranjo, J. and C.P. Gerba. 1997. Microbiological quality ofwater vending machines. Int. I Environ. Hlth. In press.

Chin, D.P. Hopewell, P.C., Yajko, D.M., Vittinghoff, E., Horsburgh, C.R. jr., Hadley,W.K., Stone, E.N., Nassos, P.S., Ostroff, S.M., Jacobsono, M.A., Matkin, C.C. and A.L.Reingold. 1994. Mycobacterium avium complex in the respiratory or gastrointestinal tractand the risk of M avium bacteremia in patients with human immunodeficiency virusinfection. J. Infect. Dis. 169:289-295.

Clark, J.A., Burger, C.A. and L.E. Sabatinos. 1982. Characterization of indicator bacteriain municipal raw water, drinking water and new main water samples. Can. J. Microbiol.28:1002-1013.

Collins, C.H., Grange, J.M. and M.D. Yates. 1984. Mycobacteria in water. J. App!.Bacteriol. 57:193-211.

Cooper, R.G. and G.W. Brown. 1968. Plesiomonas shigelloides in south Australia. J.Clin. Pathol. 21:715-718.

Dennis, P.J.L., Lee, J.V., Walker, J.T., West, A.A. and C.W. Keevil. 1989. Biofilms:Detection, implications, and solutions. Watershed 89: The future for water quality inEurope. Vol. II. Pergamon press. New York, NY.

Difco laboratories. 1984. Difco manual: dehydrated culture media and reagentsfor microbiology, 10' ed. Difco laboratories, Detroit, MI.

Dooley, J.R., Nims, L.J., Lipp, V.H., Beard, A. and L.T. Delancy. 1980. Meningitis ofinfants caused by Flavobacterium meningosepticum. I Trop. Ped. 26:24-30.

156

Du Moulin, G.C., Sherman, I.H., Hoaglin, D.C. and K.D. Stottmeier. 1985.Mycobacterium avium complex, an emerging pathogen in Massachusetts. J. Clin.Microbiol. 22:9-12.

Du Moulin, G.C., Stottmeier, K.D., Pelletier, P.A., Tsang, A.Y. and J. Hedley-White.1988. Concentration of Mycobacterium avium by hospital hot water systems. JAMA.260:1599-1601.

Du Moulin, G.C. and K.D. Stottmeier. 1978. Use of cetylpiridinium chloride in thedecontamination of water for culture of Mycobacteria. App!. Environ. Microbiol. 36:771-773.

Dupont, H.L., C. Chappell, C. Sterling, P. Okhuysen, J.B. Rose and W. Jakubowski.1995. Infectivity of Cryptosporidium parvum in healthy individuals volunteers. N. Eng. J.Med. 332, 885, 1.

Edberg, S.C., Gallo, P. and C. Kontnick. 1996. Analysis of the virulence characteristicsof bacteria isolated from bottled water, water cooler, and tap water. Microbial Eco!. Hlth.Dis. 9:67-77.

Environmental Protection Agency (EPA). 1989. Drinking water health effects task force.Health effects of drinking water treatment technologies. Lewis Pubs. Chelsea, MI.

Favero, M.S., Carson, L.A. Bond, W.W. and N.J. Petersen. 1971. Pseudomonasaeruginosa: Growth on distilled water from hospitals. Science. 173:836-838.

Fewtrell, L., Kay, D., Wyer, M., Godfree, A. and G. O'neall. 1997. Microbiologicalquality of bottled water. Wat. Sci. Tech. 35:47-53.

Fick, R.B. 1993. Pseudomonas aeruginosa: The opportunist. CRC press.

Freund, S.M., Koburger, J.A. and C.I. Wei. 1988. Enhanced recovery of Plesiomonasshigelloides following an enrichment technique. J. Food Protect. 51:110-112.

Geldreich, E.E. 1996. Microbial quality of water supply in distribution systems. CRC,Boca Raton, FL.

Geldreich, E.E. 1986. Potable water: New directions in microbial regulations. ASM news.52:530-534.

157

Geldreich, E.E., Nash, H.D., Reasoner, D.J. and R.H. Taylor. 1975. The necessity ofcontrolling bacterial population in potable water-bottled water, and emergency watersupplies. J. Amer. Water Works Assoc. 67:117-124.

Gerba, C.P. and C. N. Haas. 1988. Assessment of risks associated with enteric viruses incontaminated drinking water. American society for testing and materials. Philadelphia, p.489-494.

Gerba, C.P., J.B. Rose, C.N. Haas and K.D. Crabtree. 1996. Waterborne rotavirus: a riskassessment. Wat. Res. 30:2929-2940.

George, S.E., Kohan, M.J., Walsh, D.B., Stead, A.G. and L.D. Claxton. 1989.Polychlorinated phenyl-degrading Pseudomonads: Survival in mouse intestines andcompetition with normal flora. J. Toxicol. Environ. Hlth. 26:19-37.

George, W.L., Nakata, M.M., Thompson, J. and M.L. White. 1985. Aeromonas-relateddiarrhea in adults. Arch. Intern. Med. 145:2207-2211.

George, K.L., Parker, B.C., Grufth, H. and J.O. Falkinham III. 1980. Epidemiology ofinfection by non-tuberculous Mycobacteria. Am. Rev. Respir. Dis:122:89-94.

Gilligan, P.H. 1995. Pseudomonas and Burkholderia. p. 509-519. In ASM press (ed.)Manual of clinical microbiology. Washington, DC.

Gilligan, P.H. 1991. Microbiology of airway disease in patients with cystic fibrosis. Clin.Microbiol. Review. 4:35-51.

Gleeson, C. and N. Gray. 1997. The Coliform Index and Waterborne Disease. E and FNSpon (ed.). U.K.

Glover, N., Holtzman, A., Aronson, T., Fromm, S., Brolin, 0.G.W., Dominguez, P.,Kunkel, K.A., Overturf, G., Stelma, G., Smith, C. and M. Yakrus. 1994. The isolation andidentification of Mycobacterium avium complex (MAC) recovered from Los Angelespotable water, a possible source of infection in AIDS patients. Mt. J. Environ. Hlth. Res.4:63-72.

Gonzalez, C. Gutierrez, C. and T. Grande. 1987. Bacterial flora in bottled uncarbonatedmineral drinking water. Can. J. Microbiol. 33:1120-1125.

Goslee, S. and E. Wolinsky. 1976. Water as a source of potentially pathogenicMycobacteria. Am. Rev. Respir. Dis. 113:287-292.

158

Grabow, W.O.K., Burger, J.S. and E.M. Nupen. 1980. Evaluation of acid-fast bacteria,Candida albicans, enteric viruses and conventional indicators for monitoring wastewaterreclamation system. Prog. Water Tech. 12 :803-817.

Graham, L., Warren, N.G., Tsang, A.Y. and H.P. Dalton. 1988. Mycobacterium aviumcomplex pseudobacteriuria from a hospital water supply. J. Clin. Microbiol. 26:1034-1036.

Haas, C.N. 1983. Estimation of risk due to low doses of microorganisms: a comparison ofalternatives methodologies. Amer. J Epidemiol 118:537-582.

Haas, C.N., J.B. Rose, C.P. Gerba and S. Regli. 1993. Risk assessment of virus indrinking water. Risk Anal. 13:545-552.

Haas, C.N., C.S. Crockett, J.B. Rose, C.P. Gerba and A.M. Fazil. 1996. Assessing therisk posed by oocysts in drinking water. J Amer. Water Assoc. 9:131-136.

Handfield, M., Simard, P., Coullard, M. and R. Letarte.1996. Aeromonas hydropilaisolated from food and drinking water: Hemaglutination, hemolysis, and cytotoxity for ahuman intestinal cell line (HT-29). App!. Environ. Microbiol. 62:3459-3461.

Hardalo, C. and S.C. Edberg. 1997. Pseudomonas aeruginosa: Assessment of risk fromdrinking water. Critical Rev. Microbiol. 23:47-75.

Havelaar, A.H. and M. Vonk. 1988. The preparation of ampicillin dextrin agar for theenumeration of Aeromonas in water. Lett. App!. Microbiol. 7:169-171.

Hernandez-Duquino, H. and I.A. Rosenberg. 1987. Antibiotic resistant Pseudomonas inbottled drinking water. Can. J Microbiol. 33:286-289.

Hernandez, P. and R. Rodriguez de Garcia 1997. Prevalence of Plesiomonas shigelloidesin surface waters. Arch. Lat. Nutr. 47:47-49.

Herrington, D.A., Tzipori, S., Robins-Browne, R.M., Tall, B.D. and M.M. Levine. 1987.In vitro and in vivo pathogenicity of Plesiomonas shigelloides. Infect. Immun. 55:979-985.

Hoadley, A. W. 1977. Pseudomonas aeruginosa in surface waters. p. 31-57. In young,V.M. Raven press (ed.) Pseudomonas aeruginosa: ecological aspects and patientscolonization. New York, NY.

159

Holmberg, S.D., Shell. W.L., Fanning, G.R., Wachsmoth, I.K., Hiickman, Brenner, F.W.and P.A. Blake. 1986. Aeromonas intestinal infections in the United States. Ann. Intern.Med. 105:109-114.

Holtanan, A.E., Aronson, T.W., Glover, N., Froman, S., Stelma, G.N., Sebata, S.N.,Boian, M.G. and 0.G.W. Berlin. 1997. Examination of bottled water for non-tuberculousMycobacteria. J Food Protect. 60:185-187.

Horsburgh, C.R. 1991. Mycobacterium avium complex infection in the acquiredimmunodeficiency syndrome. N. Engl. J. Med. 19:1332-1338.

Hunter, P.R. 1994. Bottled natural mineral water and other bottled waters. Micro biol.Europe. 2(4):8-10.

Hunter, C.A. and P.R. Ensign. 1947. An epidemic of diarrhea in a newborn nurserycaused by Pseudomonas aeruginosa. Amer. J. Public Hlth. 37:1166-1169.

Janda, J.M. 1991. Recent advances in the study of the taxonomy, pathogenicity, andinfectious syndromes associated with the genus Aeromonas. Clin. MicrobioL Rev.4:397-410.

Janda, J.M., Abbott, S.L. and A.M. Carnahan. 1995. Aeromonas and Plesiomonas. p.477-482. In ASM press (ed.) Manual of clinical microbiology. Washington, DC.

Janda, J.M. and P.S. Duffey. 1988. Mesophilic Aeromonads in human disease: Currenttaxonomy, laboratory and infections disease spectrum. Rev. Infect. Dis. 10:980-997.

Jeppesen, C. 1995. Media for Aeromonas spp., Plesiomonas spp.and Pseudomonas spp.from and the environment. Int. J. Food Microbiol. 26:25-41.

Jones, A.R., Bartlett, J. and J.G. McCormack. 1995. Mycobacterium avium complex(MAC) osteomyelitis and septic arthritis in immunocompetent host. J. Infect. 30:59-62.

Kirschner, R.A. jr., Parker, B.0 and J.O. III Falkinham. 1992. Mycobacterium avium,Mycobacterium intracellulare, and M scrofulaceum in acid brown-water swamps of thesoutheastern United States and their association with environmental variables. Am. Rev.Respir. Dis. 145:271-275.

Kominos, S.D., Copeland, C.E., Grosiak, B. and B. Postic. 1972. Introduction ofPseudomonas aerugmosa into a hospital via vegetables. App!. Microbiol. 24:567-570.

160

Koprowski, H. 1956. Immunization against poliomyelitis with living attenuated virus.JAMA 160:954-966.

Korvick, J. 1996. Mycobacterium avium complex infection. Marcel Dekker (ed.), NewYork, NY.

Kubica. G.P. and L.G. Wayne. 1984. The Mycobacteria. Marcel Dekker (ed.), New York,NY.

Kuijper, E.J., Van Alphen, L., Leenders, E. and H.C. Zanen. 1989. In vitro susceptibiltiesof tropical strains of Aeromonas species from Queensland, Australia. to 22 antimicrobialagents. J Clin. Microbiol. 27:1280-1285.

Kwaga, J.K., Adesiyun, A.A., Bello, C.S. and S.U. Abdullahi. 1988. Occurrence ofPlesiomonas shigelloides in humans and water in Zaria, Nigeria. Microbiologica. 11:165-167.

LeChevallier, M.W., Evans, T.M., Seidler, R.J., Daily, 0.P., Merrell, B.R., Rollins, D.M.and S.W. Joseph.1982. Aeromonas sobria in chlorinated drinking water supplies. Microb.Ecol. 8:325-333.

LeChevallier, M.W., Seidler, R.J. and T. M. Evans. 1980. Enumeration andcharacterization of standard plate count bacteria in chlorinated and raw water supplies.App!. Environ. Microbiol. 40:922-930.

Levesque, B., Pierre, S., Gauvin, D., Gingras, S., Dewailly, E. and R. Letarte. 1994.Comparison of the microbiological quality of water coolers and that of municipal watersystems. Appl. Environ. Microbiol. 60:1171-1178.

Liesenfeld, O., Weinke, T. and H. Hahn. 1993. Three-year prevalence ofenteropathogenic bacteria in an urban population in Germany. Infect. 21:101-105.

Lombardo, L.R., West, P.R. and J.L. Holbrook. 1985. A comparison of various mediaand incubation temperatures used in the heterotrophic plate count analysis. Water Quai.Tech. Conf, Amer. Water Works Assoc. Denver, CO. p.251-270.

Lynkins, B.U., Clark, R.M. and A.G. James. 1992. Point-of-use/point-of-entry fordrinking water treatment. Lewis, Chelsea, MI.

Macler, B.A. and S. Regli. 1993. Use of microbial risk assessment in setting UnitedStates drinking water standards. Int. J Food MicrobioL 18:245-256.

161

Manaia, C.M., Nunes, O.C. , Morais, P.V. and M.S. Da Costa. 1990. Heterotrophic platecounts and the isolation of bacteria from mineral waters on selective andenrichment media. J. Appl. Bacteriol. 69:871-876.

Mansfield, K.G. and A.A. Lackner. 1997. Simian immunodeficiency virus-inoculatedmacaques acquire Mycobacterium avium from potable water during AIDS. J Infect.175:184-187.

Martinetti, L. and M. Altwegg. 1992. rRNA gene restriction pattern as taxonomic toolsfor the genus Aeromonas. Int. J. Syst. Bacteriol. 24:384-389.

McFeters, G.A. 1990. Drinking water microbiology. G.A. McFeters (ed.), SpringerVerlag. New York, NY.

Medema, G. and C. Schets. 1993. Occurrence of Plesiomonas shigelloides in surfacewater: relationship with fecal pollution and trophic state. Zentralbl. Hyg. Umweltmed.194:398-404.

Meitert, E.A., Vlad, A. and E. Sima. 1984. Food Poisoning associated with Pseudomonasaeruginosa. Arch. Roum. Pathol. Exp. Microbiol. 43:115-119.

Miller, L.M. and J.A. Koburger.1985. Plesiomonas shigelloides: An opportunistic foodand waterborne pathogen. J. Food Protect. 48:449-457.

Morbidity and Mortality Weekly Report (MMWR). 1998. Plesiomonas shigelloides andSalmonella serotype Hartford infections associated with contaminated water supply-Livingstone County, NY. MMWR. 47:349-396.

Morgan, D.R., Johnson, P.C., DuPont, H.L., Satterwhite, T.K. and L.V. Wood. 1985.Lack of correlation between known virulence properties of Aeromonas hydrophila andenteropathogenicity for humans. Infect. Immun. 50:62-65.

Moyer, N.D. 1987. Clinical significance of Aeromonas species isolated from patientswith diarrhea. J Clin. Microbiol. 25:2044-2048.

Murphey, S.A, Jungkind, D.L. and E. Adams. 1983. Mycobacterium avium-intracellularein a hospital hot water system: epidemiological investigation. In: Proceedings andabstracts of the 2nd Interscience Conference on Antimicrobial Agents and Chemotherapy,Las Vegas. Washington, DC: American Society for Microbiology, p. 277.

National Research Council. 1983. Risk assessment in the federal government: managingthe process. National Academy Press, Washington, DC.

162

National Research Council. 1994. Science and judgment in risk assessment. NationalAcademy Press, Washington, DC.

Nolte, F. and B. Metchock. 1995. Mycobacterium. p. 400-437. In Murray R.R. (ed.).Manual of clinical microbiology. American society for microbiology. Washington DC.

Ogan, M. 1992. Microbiological quality of bottled water sold in retail in Nigeria. J App!.Bacteriol. 73:175-181.

Palleroni, N.J. 1984. Genus I Pseudomonas migula 1894, 237AL , p. 141-199. In N.R.Kreig and J.G. Holt (ed.) Bergey's Manual of Systematic Bacteriology. Baltimore.

Payment, P., Franco, E. and Siemiatycki. 1991. Gastrointestinal health effects associatedwith the consumption of drinking water produced by point-of-use domestic reverseosmosis filtration unites. App!. Environ. Microbiol. 57:945-948.

Payment, P., Gamache, F. and G. Paquette. 1988. Microbiological and virologicalanalysis of water from two filtration plants and their distribution system. Can. J.Microbiol. 34:1304-1309.

Payment, P., Siemiatycki, J., Richardson, L., Renaud, G., Franco, E. and M. Prevost.1997. A prospective epidemiological health effects due to the consumption of drinkingwater. Int. J. Environ. Hlth. Res. 7:5-31.

Peters, M., Muller, C., Rusch-Gerdes, S., Seidel, C., Gobe!, U., Pohle, H.D. and B. Ruf.1995. Isolation of atypical Mycobacteria from tap water in hospitals and homes: Is this apossible source of disseminated MAC infection in AIDS patients. J. Infect. 31:39-44.

Pollack, M. 1990. Pseudomonas aeruginosa. p. 1673-1687. In Mandell, G.H., DouglasR.G. Jr., Benett, J.E. (ed.) Principles and practice of infectious diseases. Livingstone, NY.

Prince, D.S., Peterson, D.D., Steiner, R.M., Gottlieb, J.E., Scott, R., Israel, H.L.,Figueroa, W.G. and J.E. Fish. 1989. Infections with Mycobacterium avium complex inpatients without predisposing factors conditions. N. Engl. J. Med. 321:863-868.

Rautelin, H., Sivanen, A., Kuikka, A., Renkonen, U.V., Valtonen, V. and T.V. Kosunen.1995. Enteric Plesiomonas shigelloides infections in Finnish patients. Scand. J. Infect.Dis. 27:495-498.

Reasoner, D.J. 1990. Monitoring heterotrophic bacteria in potable water. p. 452-477 InDrinking water microbiology. Gordon A. McFeters (ed.) Spring-Verlag. NY.

163

Reasoner, D.J., Blannon, J.C., Geldreich, E.E. and J. Barnick. 1989. Non-photosyntheticpigmented bacteria in a potable water treatment and distribution system. App!. Environ.Microbiol. 55:912-921.

Reasoner, D.J. and E.E. Geldreich. 1985. A new medium for the enumeration andsubculture of bacteria from potable water. App!. Environ. Microbiol. 49:1-7.

Regli, S., Rose, J.B.. Haas, C.N. and C.P. Gerba. 1991. Modeling the risk from Giardiaand viruses in drinking water. J. Amer. Water Works. Assoc. 83:76-84.

Ridgway, H.F. and B.H. Olson. 1982. Chlorine resistant patterns of bacteria from twodrinking water distribution systems. App!. Environ. Microbiol. 44:972-987.

Ridgway, H.F., Rigby, M.G. and D.G. Argo. 1985. Bacterial adhesion and fouling ofreverse osmosis membranes. J Amer. Water Works Assoc. 97:106-109.

Rippey, S.R. and V.J. Cabelli. 1979. Membrane filter procedure for enumeration ofAeromonas hydrophila in fresh water. App!. Environ. Ann. Intern. Med. 38:180-113.

Rose, J.B. and C.P. Gerba. 1991. Use of risk assessment for the development of microbialstandards. Wat. Sci. Tech. 24:29-34.

Rose, J.B., Haas, C.N. and S. Regli. 1991. Risk assessment and control of waterbornegiardiasis. Amer. J. Pub. Hlth. 1:709-713.

Roseberry, A.M. and D.E. Burmaster. 1992. Lognormal distribution for water intake bychildren and adults. Risk Anal. 12:99-104.

Rosenzweig, D.Y. and D.P. Schlueter. 1981. Spectrum of clinical disease in pulmonaryinfections with Mycobacterium avium-intracellulare. Rev. Infect. Dis. 3:1046-1052.

Rusin, P.A., Rose, J.B., Haas, C.H. and C. P. Gerba. 1997. Risk assessment ofopportunistic bacterial pathogens in drinking water. Rev. Environ. Contam. Toxicol.152:57-83.

Rutala, W.A., Sarubbi, F.A. jr., Finch, C.S., MacCormack, J.N. and G.E. Steinkraus.1982. Oyster-associated outbreak of diarrhoeal disease possibly caused by Plesiomonasshigelloides. LANCET. 27:739.

Schulze-Robbecke, R., Janning, B. and R. Fischerder. 1992. Occurrence of Mycobacteriain biofilms samples. Tub. Lung Dis.73:141-144.

164

Seidler, R.J., Allen, D.A, Lockman, H., Colwell, R.R., Joseph, S.W. and O.P. Daily.1980. Isolation, enumeration, and characterization of Aeromonas from polluted watersencountered in diving operations. Appl. Environ. Microbiol. 39:1010-1018.

Severin, B.F. 1976. Inactivation of new indicators organisms of disinfection efficiency.Part II. Combined chlorine as chloramines. Presented at the 96th annual meeting. Amer.Water Works Assoc. New Orleans, LO.

Shooter, R.A., Cooke, E.M., Gaya, H., Kumar, P., Patel, N., Parker, T., Tham, T. andD.R. France. 1969. Food and medicaments as possible sources of hospital strains ofPseudomonas aeruginosa. LANCET 1:1227-1229.

Sloodley, B.J. and B.T. Thom.1970. Observations on the intestinal carriage ofPseudomonas aeruginosa. J Med. Microbiol. 3:367-375.

Stiles, M.E. 1989. Less recognized or presumptive foodborne pathogenic bacteria. p.689-692. In Foodbome bacterial pathogens. Marcel Dekker (ed.). NY.

Surucu, F. and C.N. Haas. 1976. Inactivation of new indicator organisms of disinfectionefficiency. Part I. Free available chlorine species kinetics. Presented at the 96th annualmeeting. Amer. Water Works Assoc. New Orleans, LO.

Tamagini, L.M. and R.D. Gonzalez. 1997. Bacteriological stability and growth kinetics ofPseudomonas aeruginosa in bottled water. J Appl. Microbiol. 83:91-94.

Tsai, G.J. and S.C. Yu. 1997. Microbiological evaluation of bottled water uncarbonatedmineral water in Taiwan. Int. J. Food Microbiol. 37:137-143.

Tsukamoto, T., Kinoshita, Y., Shimada, T. and R. Sakazaki. 1978. Two epidemics ofdiarrhoeal possibly caused by Plesiomonas shigelloides. J Hyg. Camb. 80:275-280.

Van Der Kooij, D. 1988. Properties of Aeromonas and their occurrences and hygienicsignificance in drinking water. Zentralbl. BakterioL MikrobioL Hyg. Ser. B. 187:1-17.

Van Der Kooij, D. 1991. Nutritional requirements of Aeromonads and theirmultiplication in drinking water. Experientia. 47:444-446.

Von Gravenitz, A. and A.H. Mensh. 1968. The genus Aerornonas in human bacteriology:Report of 30 cases and review of the literature. N. Eng. J. Med. 278:245-249.

165

Von Reyn, C.F., Barber, T.W., Arbeit, R.D., Sox, C.H., Oconnor, G.T., Brindle, R.J.,Gilks, C.F., Hakkarainen, K., Ranki, A., Bartholomew, C., Edwards, J., Tostesson,A.N.A. and M. Magnusson. 1993. Evidence of previous infection with Mycobacteriumavium-M intracellulare complex among healthy subjects: An international study ofdominant mycobacterial skin test reactions. J. Infect. Dis. 168:1553-1558.

Von Reyn, C.F., Manslow, J.N., Barber, T.W., Falkinham III, J.O. and R.D. Arbeit. 1994.Persistent colonization of potable water as a source of Mycobacterium avium infection inAIDS. LANCET. 343:1137-1141.

Warburton, D.W., Bowen, B. and A. Konkle. 1994a. The survival and recovery ofPseudomonas aeruginosa and its effects upon Salmonellae in water: Methodology to testbottled water in Canada. Can. J. Microbiol. 40:987-992.

Warburton, D.W., McCormick, J.K. and B. Bowen. 1994b. Survival and recovery ofAeromonas hydrophila in water: Development of methodology for testing bottled waterin Canada. Can. J Microbiol. 40:145-148.

Ward, R.L., S. Krugman, J. Giles, M. Jacobs and O. Bodansky. 1958. InfectiousHepatatis-studies of its natural history and prevention. N. Eng. J Med. 258:402-416.

Ward, R.L., D.I. Bernstein and E.C. Young. 1986. Human Rotavirus studies in volunteersof infectious dose and serological response to infection. J. Infect. Dis. 154:871-877.

Weber, G., Werner, H.P. and H. Matsching. 1971. Death cases in newborns caused byPseudomonas aeruginosa contaminated drinking water. Zentralbl. Bakteriol. 216:210-214.

Werthamer, S. and M. Weiner. 1972. Subacute bacterial endocarditis due toFlavobacterium meningosepticum. Amer. J. Clin. Pathol. 57:410-412.

World Health Organization (WHO). 1993. Guidelines for the drinking water quality. 2'd

edition. Geneva, Switzerland.

Yakrus, M.A. and R.C. Good. 1990. Geographic distribution, frequency, and specimensource of Mycobacterium avium complex serotypes isolates from patients with acquiredimmunodefieciency syndrome. J. Clin. Microbiol. 28:926-929.

Yamamoto, T., Shibagaki, T., Yamori, S., Saito, H., Yamada, H., Ichiyama, S. and K.Shimokata. 1993. Polymerase chain reaction for the differentiation of Mycobacteriumintracellulare and M. avium. Tub. Lung dis. 74:342-345.

166

Zwadyk, P., Down, J.A., Myers, N. and M.S. Dey. 1994. Rendering of Mycobacteria safefor molecular diagnostic studies and developmental of lysis method for stranddisplacement amplification and PCR. J. Clin. Microbiol. 32:2140-2146.