National Academies Press: OpenBook

Drinking Water Distribution Systems: Assessing and Reducing Risks (2006)

Chapter: 3 Public Health Risk from Distribution System Contamination

« Previous: 2 Regulations, Non-Regulatory Approaches, and Their Limitations
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

3
Public Health Risk from Distribution System Contamination

One of the most challenging facets of reducing the risk of contaminated distribution systems is being able to quantify the existing risk. This is made complicated not only by the plethora of factors that can constitute public health risks, including a diversity of microbial pathogens and chemical compounds, but also by the varying response that a given individual will have when exposed to those factors. This chapter describes three primary mechanisms used to assess the acute public health risk of distribution system contamination, the limitations of these methods, and what conclusions can be derived from currently available data.

INTRODUCTION TO RISK

The process of risk assessment involves determining the likelihood and severity of different adverse impacts given exposure of a population to a hazard. Risk analysis includes the process of risk assessment, as well as risk management activities to decide what an acceptable risk level is and to take actions to reduce risk (NRC, 1983). Risk assessment requires the activities of hazard identification, exposure assessment, and dose-response (or exposure-response) assessment. Hazard identification is the determination of what adverse agents might be present and what adverse impacts they might cause. Exposure assessment is the quantitative determination of the levels of contaminants (in the case of environmental exposures) individuals may consume/inhale/contact over a specific time period. Dose-response assessment is the quantitative determination of the likelihood of an individual having a particular adverse effect from a given exposure. Alternatively, this can be viewed as the proportion of persons in a population who are expected to have the adverse effect were they to have the particular exposure.

Various federal agencies, including the U. S. Environmental Protection Agency (EPA), have developed specific guidelines and procedures for performing risk assessment, particularly for carcinogens and for substances that result in non-carcinogenic toxic effects. In the case of infectious agents (which are frequently the concern in drinking water), methodologies are at a developmental stage.

One of the goals of performing risk assessment within a regulatory framework is to develop regulatory guidance or standards (or decide not to undertake such action) based on the results. This process, which is part of risk analysis,

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

requires additional considerations such as cost and equity. Under the Safe Drinking Water Act, EPA is required to set a maximum contaminant level goal (MCLG) for certain contaminants that is absolutely protective against all adverse health effects, given available risk assessment information. For most contaminants with MCLGs, a regulatory level is then established—a maximum contaminant level (MCL)—or a treatment technique is required, both of which incorporate considerations of feasibility (see Box 3-1).

In determining a regulatory level such as an MCL, implicitly or explicitly the acceptable residual risk (after the implementation of any interventions) must be decided upon. The empirical evidence is that, for human carcinogens, EPA has regarded a window of residual lifetime risk of 1/1,000,000 to 1/10,000 to be acceptable (see Box 3-2 for an explanation of the origins of this value and its extension to infectious agents). In other words, a residual risk resulting in no more than 1 extra cancer in the lifetime of a population of 10,000 to 1,000,000 persons is regarded as being acceptable.

Risks from Drinking Water

Drinking water can serve as a transmission vehicle for a variety of hazardous agents: enteric microbial pathogens from human or animal fecal contamination (e.g., noroviruses, E. coli O157:H7, Cryptosporidium), aquatic microorganisms that can cause harmful infections in humans (e.g., nontuberculous mycobacteria, Legionella), toxins from aquatic microorganisms (such as cyanobacteria), and several classes of chemical contaminants (organic chemicals such as benzene, polychlorinated biphenyls, and various pesticides; inorganic chemicals such as arsenic and nitrates; metals such as lead and copper; disinfection byproducts or DBPs such as trihalomethanes; and radioactive compounds).

Contaminants in drinking water can produce adverse effects in humans due to multiple routes of exposure. In addition to risk from ingestion, exposure can also occur from inhalation and dermal routes. For example, inhalation of droplets containing respiratory pathogens (such as Legionella or Mycobacterium) can result in illness. It is known that DBPs present in drinking water may volatilize resulting in inhalation risk, and these compounds (and likely other organics) may also be transported through the skin (after bathing or showering) into the bloodstream (Jo et al., 1990). Reaction of disinfectants in potable water with other materials in the household may also result in indoor air exposure of contaminants; for example Shepard et al. (1996) reported on release of volatile organics in indoor washing machines. Thus, multiple routes of exposure need to be considered when assessing the risk presented by contaminated distribution systems. It should be noted, however, that the report will not consider such indirect routes of exposure as (1) the loss of pressure and subsequent inadequate fire protection, (2) loss of water for hospitals and dialysis centers, and (3) leaks in household plumbing that lead to toxic mold growth.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

BOX 3-1

U.S. Code, Title 42(6A)(XIIB)§300g-1 (Safe Drinking Water Act as Amended)

  1. Maximum contaminant level goals. Each maximum contaminant level goal established under this subsection shall be set at the level at which no known or anticipated adverse effects on the health of persons occur and which allows an adequate margin of safety.

  2. Maximum contaminant levels. Except as provided in paragraphs (5) and (6)1, each national primary drinking water regulation for a contaminant for which a maximum contaminant level goal is established under this subsection shall specify a maximum contaminant level for such contaminant which is as close to the maximum contaminant level goal as is feasible.

It has been recognized for some years that consumers face risk from multiple hazards, and that action to reduce the risk from one hazard may increase the risk from other hazards given the same exposure. There are prominent examples of this phenomenon in the drinking water arena that have greatly complicated efforts to reduce overall risk from distribution systems. Havelaar et al. (2000) assessed the relative changes in risk from switching to ozone treatment of drinking water in the Netherlands. In this case, there was a projected reduction in risk from waterborne infectious disease (such as Cryptosporidium) while there was a projected increase in risk from DBP formation (the primary one examined was bromate). To compare the net change in overall risk, it is necessary to place the multiple risks (with their different endpoints in terms of disease severity) on the same scale. Havelaar et al. (2000) did this comparison using the methodology of disability adjusted life years (DALY’s). In this approach, the severity of an adverse health effect is quantitatively weighted by an index (disability weight) reflecting the proportional degradation in health (a weight of 0 is reflective of absence of an effect, while a weight of 1 is reflected in total impairment); the integral of the years of diminished functioning multiplied by the disability weight is summed with the reduction in lifespan due to premature mortality to get the aggregate impact to a population. In principle, using such an approach one can optimize for the overall net reduction in risk, considering competing hazards. It is noted that the DALY framework has not been adopted for U.S. regulatory practice and remains controversial for a number of technical and policy reasons (including age equity) (Anand and Hanson, 1997).

When risk is assessed for chemical or microbial exposure, it should be considered that not all segments of the population are at the same degree of risk. This may be due to differences in exposure in terms of either consumption (Gerba et al., 1996) or in concentrations (due to heterogeneity in the environ-

1

Paragraph (5) allows departure upwards from setting the MCL as close to the MCLG as feasible if doing so would result in an increase in risk from other contaminants, or would interfere with the performance of processes used to address other contaminants. Paragraph (6) allows departure upward from the “as close as feasible” criterion in certain circumstances if the benefits would not justify the cost of compliance at that standard.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

BOX 3-2

Origin of the 1/10,000 Acceptable Risk Level for Carcinogens and Infectious Agents

EPA has been at the forefront of the issue of acceptable risk in virtually all of its programmatic areas, primarily as the result of court challenges to its regulations. In response to the 1987 Section 112 Clean Air Act decision (Natural Resources Defense Council vs. U.S. Environmental Protection Agency 824 F. 2nd 1146 [1987]), EPA decided it would base its regulatory decisions on quantitative risk assessments using the general policy that a lifetime added cancer risk for the most exposed person of 1 in 10,000 (1 × 10-4) might constitute acceptable risk and that the margin of safety required by statute and reinforced by the court should reduce the risk for the greatest number of persons to an added lifetime risk of no more than 1 in 1 million (1 × 10-6). However, EPA (along with the courts) has not viewed “safe” as the equivalent of risk-free and has determined that standards should protect against significant public health risks (EPA 49 Fed. Reg. 8386 [1984]; Rodricks et al. 1987; Industrial Union Department, AFL-CIO v. American Petroleum Institute et al. 448 U.S. 607 [1980]). EPA has repeatedly rejected the opinion that it can establish a universal (i.e., brightline) acceptable risk that should never be exceeded under any circumstances, and they maintain that guidance provided under one statute might have little relevance to others because of differing program goals. In practical terms, EPA almost never regulates at a theoretical risk below 1 × 10-6 (de minimis) and almost always regulates at a theoretical risk below 1 × 10-4 (de manifestis)” (NRC, 2004).

Policy with respect to acceptable levels of risk from exposure to infectious agents is less well developed than for chemical carcinogens. However, in framing the Surface Water Treatment Rule (Federal Register, June 29, 1989, page 27486), the rule for reduction of risk from Giardia and viruses was set to achieve a residual estimated risk of infection below 1/10,000 per year. This number derived from the then average waterborne illness rate associated with reported waterborne outbreaks (Regli et al., 1991). However it is now recognized that the waterborne illness rate is substantially greater than this value—due to underreporting of outbreaks, as well as to substantial endemic illness. The use of infection rather than illness as an endpoint was intended to compensate for secondary cases and also for presumed heightened infectivity amongst sensitive subpopulations.

The use by EPA of an acceptable risk window for microorganisms in the 10-6 to 10-4 range as one factor in setting standards continues. As recently as the promulgation of the Long Term 2 Enhanced Surface Water Treatment Rule (Federal Register, January 5, 2006), EPA has stated: “EPA and Advisory Committee deliberations focused on mean source water Cryptosporidium concentrations in the range of 0.01–0.1 oocysts/L as threshold levels for requiring additional treatment…these levels are estimated to result in an annual infection risk in the range of 1.7x10-4 − 6 x 10-3 … for a treatment plant achieving 3-log Cryptosporidium removal (the treatment efficiency estimated for conventional plants under existing regulations).”

ment, e.g., in the distributed water), or to intrinsic differences in susceptibility (Balbus et al., 2000). Unfortunately, our ability to assess quantitative differences in intrinsic susceptibility remains poor, and therefore protection of susceptible subpopulations often relies upon the imposition of safety factors.

Methods for Characterizing Human Health Risk

Characterization of human health risks may be performed using an epide-

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

miological approach or using a risk assessment approach. These methods are complementary and have different strengths and limitations, and each has been used for assessment of drinking water risks in various applications. Epidemiological approaches study the relationship between exposures and disease in actual populations and are descriptive, correlational, or analytic. In the descriptive study, population surveys or systematic disease surveillance (monitoring) describe disease patterns by various factors such as age, seasonality, and geographic location. Correlational (also called “ecologic”) studies collect population level data on disease rates and exposures and look for correlations. Analytical studies (whether experimental or observational) are those in which individual-level data is collected and the investigator tests a formal hypothesis about the association between exposure and disease.

Risk assessment methods, on the other hand, follow the hazard identification, dose-response assessment, exposure assessment, and risk characterization paradigm noted above. Frequently, but not always, the dose-response assessment is based upon extrapolation from results of trials in animals (although results from human exposure may be used where available—for example, in human feeding trials of infectious agents or from studies in populations exposed in occupational or other settings to particular agents of concern).

Epidemiological studies have the advantage of involving human populations, often experiencing the exposure of interest and representing a range of variability in susceptibility and behavior. However to detect a small increase in risk from the baseline, epidemiological studies require very large sample sizes, and thus considerable expense and effort. Epidemiological studies cannot provide direct information on the potential for risk reduction from a proposed change in treatment practice that has not yet been implemented since by definition there is not yet human exposure to conditions expected from the proposed change. However, epidemiological studies can be designed to measure the direct impact of a treatment intervention after it has been implemented. This is very powerful tool and it has provided the evidence base that changes in water treatment have had a positive impact on community health. For example, the recent meta-analysis by Fewtrell and Colford (2004) demonstrates the body of evidence linking improvements in community and household water quality to health.

Risk assessment approaches have the advantage of being flexible in their application to potential (but not yet experienced) situations. A risk assessment can be performed even when the projected risk from a particular exposure or change of exposure is very small. They have the disadvantage of requiring extensive measurement or modeling to ascertain exposure, and also of the need for dose-response studies. Often these dose-response studies are in animals or at higher doses, thereby requiring extrapolation with respect to dose (via a formal mathematical dose-response curve) and/or between species. Generally, whether animal or human data are used to establish the dose-response relationship, the range in variability in susceptibility is small (compared to a full human popula-

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

tion) and therefore some margin of safety may need to be explicitly used to account for more susceptible subpopulations.

This chapter discusses what is known about the human health risks that derive from contamination of the distribution system, relying on three primary approaches: risk assessment methods that utilize data on pathogen occurrence measurements, outbreak surveillance data, and epidemiology studies. A special section is devoted to Legionella, for which all three types of activities have occurred, leading to greater understanding of the risks inherent from growth of this organism in distribution systems. Because the impetus for this study was revision of the Total Coliform Rule, the report focuses primarily on acute risks from microbial contamination of the distribution system. However, there are short-and long-term risks from chemicals that merit mention (particularly DBPs—lead and copper were outside the scope of the study). DBP concentrations in the distribution system can vary significantly depending on water residence time, the types of disinfectants used, and biological and chemical reactions, among many other factors (see Chapter 6). The concentrations of trihalomethanes in finished water tend to increase with increasing water age, while certain haloacetic acids tend to decrease in concentration over time (see Chapter 6; Arbuckle et al., 2002). A number of epidemiologic studies have examined the health significance of DBP exposure and have reported significantly increased risks of bladder, rectal, and/or colon cancers in some populations (King et al., 1996; Koivusalo et al., 1997; Doyle et al., 1997; Cantor et al., 1998; Yang et al., 1998; King et al., 2000) as well as adverse reproductive outcomes (Waller et al., 1998; Dodds et al., 1999; Klotz and Pyrch, 1999; King et al., 2000). However, determining and classifying DBP exposure in these studies has been extremely challenging and has made it difficult to interpret the findings of these studies (Arbuckle et al., 2002, Weinberg et al., 2006). Furthermore, the contribution of distribution systems to the reported risk, as opposed to drinking water treatment or other processes, has not been elucidated. Because epidemiological studies of DBP exposure have been extensively reviewed by others (Boorman et al., 1999; Nieuwenhuijsen et al., 2000; Graves et al., 2001), they are not reviewed here.

EVIDENCE FROM PATHOGEN OCCURRENCE MEASUREMENTS

The risk assessment approach relies on being able to measure or predict (e.g., by modeling) the concentration of an etiologic agent in the water supply. Certain microbial pathogens are indicative of distribution system contamination stemming from both internal and external sources. These include bacteria known to form biofilms—a physiological state in which organisms attach to and grow on a surface (Characklis and Marshall, 1990)—and bacteria that indicate an external contamination event such as intrusion. In distribution systems, the interior pipe walls, storage tanks, sediments, and other surfaces in contact with finished water are colonized by bacteria, which can survive, grow, and detach depending on local conditions. Other types of bacteria (such as coliforms) as

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

well as enteric viruses and protozoa (Quignon et al., 1997; Piriou et al., 2000) are also found in biofilms. However, their presence can also be attributable to an external contamination event or break through of the treatment barrier.

The microbiology of distribution systems can be influenced by a variety of factors (e.g., poor quality source water, inadequate treatment, unsanitary activity, backflow). Given this report’s assumption of adequate treatment, a discussion of all source water microbes and those that would be eliminated during treatment is not warranted. Furthermore, virtually any microorganism in close enough proximity to a vulnerable part of the distribution system (e.g., a cross connection, main break, or leak) could enter during an external contamination event. Control of these events—see Chapters 4 and 5—is important for reducing the risks of not only microbial pathogens but also chemicals that might enter distribution system. Because the complexity of microbes from such diverse sources is beyond the scope of this report, the following section focuses on those organisms most likely to indicate either internal or external contamination of the distribution system.

The Microbiology of Bulk Water

The microbiology of distribution systems essentially consists of two different environments—microorganisms in the bulk water column and those in biofilms attached to the surfaces of pipes, sediments, and other materials. Microorganisms in the bulk water column originate from either the source water, from bacterial growth within the treatment process (e.g., within the treatment filters), from biofilms within the distribution system, or from recontamination of the water from cross connections, intrusion, pipe breaks, or other external sources.

Heterotrophic Bacteria

Heterotrophic bacteria (a broad classification that takes into account all bacteria that utilize organic carbon) are commonly found in the bulk water of distribution systems because they readily form biofilms in such systems. They are measured by using heterotrophic plate counts (HPC). Heterotrophs have traditionally been divided into two primary groups based on their cell wall characteristics—Gram-negative and Gram-positive.

The presence of a disinfectant residual in drinking water has a tremendous selective effect, particularly on Gram-negative bacteria, which are relatively sensitive to inactivation by disinfectants. Identification of bacteria using fatty acid analysis (Norton and LeChevallier, 2000) showed that chlorination resulted in a rapid shift from predominately Gram-negative bacteria (97 percent) in the raw water to mostly Gram-positive organisms (98 percent) in the chlorinated water (see Table 3-1). Bacteria in the raw water were diverse, with Acinetobacter

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

TABLE 3-1 Bacterial Populations Isolated from the Water Column During Treatment

Bacterial Identification

Percentage of Population in Raw Water

Percentage of Population in Ozone Contactor

Percentage of Population in Filter Effluent

Percentage of Population in Distribution System Influent

Gram Negative

 

 

 

 

Acidovorax spp.

2

 

4

7

Acinetobacter spp.

29

6

 

 

Alcaligenes spp.

12

2

1

 

Alteromonas spp.

2

 

 

 

Comamonas spp.

1

 

3

 

Enterobacter spp.

2

 

5

 

Flavobacterium spp.

2

 

5

 

Hydrogenophaga spp.

8

3

1

 

Klebsiella spp.

10

1

3

 

Methylobacterium spp.

1

 

2

 

Pseudomonas spp.

14

53

22

 

Rhodobacter spp.

2

1

 

 

Sphingomonas spp.

2

2

19

 

Stenotrophomonas spp.

2

1

2

 

Xanthobacter spp.

3

 

 

 

Others*

2

1

5

 

Gram Positive

 

 

 

 

Bacillus spp.

 

 

 

7

Nocardia spp.

1

3

7

53

Rhodococcus spp.

 

16

4

 

Staphylococcus spp.

1

1

 

 

Others*

1

1

1

 

Unidentified

3

9

16

33

* Includes organisms isolated from only one site at a frequency of 1%. 100 isolates were identified from each site.

SOURCE: Adapted from Norton and LeChevallier (2000).

spp., Pseudomonas spp., and Klebsiella spp. predominate among the 20 genera identified. Ozonation of the raw water reduced the microbial diversity to 13 genera, dominated by Pseudomonas spp. and Rhodococcus spp. However, following biologically active granular activated carbon filtration, 19 genera were identified in the filter effluent, the majority of which (63 percent) matched isolates observed in the raw water. The predominant genera were Pseudomonas spp. and Sphingomonas spp., which are known to grow attached to the carbon fines of the filter while utilizing natural organic compounds found in the aquatic environment. Final chlorination of the filtered water resulted in a shift to Nocardia spp. as the water entered the pipe system. Nocardia spp. possess characteristic fatty acids that are closely related to Rhodococcus, Mycobacterium, and Corynebacterium. Its partially acid-fast cell wall and possession of the catalase

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

enzyme, which breaks down hydrogen peroxide, are important factors that enable the organism to survive disinfection. Other Gram-positive bacteria found in chlorinated drinking water include Bacillus and Staphylococcus spp. Bacillus spp. form environmentally resistant spores that can withstand prolonged contact with chlorine. Some strains of Bacillus and Staphylococcus aureus can produce toxins when contaminated water is used in food preparation (LeChevallier and Seidler, 1980).

Treated drinking water will include a mixture of Gram-negative and Gram-positive bacteria. In the absence of a disinfectant residual, Gram-negative bacteria will out grow Gram-positive bacteria and dominate the bacterial population. These organisms typically include Pseudomonas, Acinetobacter, Flavobacterium, and Sphingomonas spp. For the most part, these organisms have limited public health significance, except for Pseudomonas aeruginosa, which is a possible opportunistic pathogen in drinking water and in the biofilms of water systems. It is known to colonize point-of-use carbon filters in drinking water systems (de Victoria and Galvan, 2001; Chaidez and Gerba, 2004). Pseudomonas aeruginosa is of concern in bathing waters, especially in swimming pools and spas, where skin infections may result due to exposure. In the case of drinking water, there are a few studies that suggest a relationship between the presence of this organism in the water and disease. In one hospital setting, five of 17 patients with a Pseudomonas infection carried a genotype also detected in the tap water (Trautmann et al., 2001). In another outbreak of pediatric P. aeruginosa urinary tract infections, two isolates had genotypes similar to those in the water. The outbreak was resolved when the taps in the unit were changed (Ferroni et al., 1998).

Despite these specific incidences, a workgroup recently convened by the World Health Organization (WHO) to address this issue concluded that HPC bacteria were not associated with any adverse health effect (Bartram et al., 2003). “Some epidemiological studies have been conducted into the relationship between HPC exposures from drinking water and human health effects. Other studies relevant to this issue include case studies, especially in clinical situations, and compromised animal challenges using heterotrophic bacteria obtained from drinking water distribution systems. The available body of evidence supports the conclusion that, in the absence of fecal contamination, there is no direct relationship between HPC values in ingested water and human health effects in the population at large. This conclusion is also supported indirectly by evidence from exposures to HPC in foodstuffs where there is no evidence for health effects link in the absence of pathogen contamination. There are a small number of studies that have examined possible links between HPC bacteria and non-intestinal outcomes in general populations. The conclusions of these studies do not support a [health] relationship” (WHO, 2002).

One of the difficulties in interpreting the significance of HPC data is that test methods involve a wide variety of conditions that lead to a wide range of quantitative and qualitative results. For this reason, the EPA has not yet issued a health-based standard. However, the Surface Water Treatment Rule requires

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

that distribution system locations without a detectable disinfectant residual maintain HPC levels at or below 500 colony forming units (CFU)/mL in at least 95 percent of the samples each month (EPA, 1989).


Coliform Bacteria. Total coliform bacteria (a subset of Gram-negative bacteria) are used primarily as a measure of water treatment effectiveness and can occasionally be found in distribution systems. The origins of total coliform bacteria include untreated surface water and groundwater, vegetation, soils, insects, and animal and human fecal material. Typical coliform bacteria found in drinking water systems include Klebsiella pneumoniae, Enterobacter aerogenes, Enterobacter cloacae, and Citrobacter freundii. Other typical species and genera are shown in Table 3-2. Although most coliforms are not pathogenic, they can indicate the potential presence of fecal pathogens and thus in the absence of more specific data may be used as a surrogate measure of public health risk. Indeed, the presence of coliforms is the distribution system is usually interpreted to indicate an external contamination event, such as injured organism passage through treatment barriers or introduction via water line breaks, cross connections, or uncovered or poorly maintained finished water storage facilities (Geldreich et al., 1992; Clark et al., 1996). However, biofilms within distribution systems can support the growth and release of coliforms, even when physical integrity (i.e., breaches in the treatment plant or distribution system) and disinfectant residual have been maintained (Characklis, 1988; Haudidier et al., 1988; Smith et al., 1990), such that their presence may not necessarily indicate a recent external contamination event. Coliform regrowth in the distribution system is more likely during the summer months when temperatures are closer to the optimum growth temperatures of these bacteria.

Thermotolerant coliforms (capable of growth at 44.5 °C), also termed “fecal coliforms” have a higher association with fecal pollution than total coliforms. And Escherichia coli is considered to be even more directly related to fecal pollution as it is commonly found in the intestinal track of warm-blooded animals. Although most fecal coliform and E. coli strains are not pathogenic, some strains are invasive for intestinal cells and can produce heat-labile or heat-stable toxins (AWWA, 1999). E. coli and most of the thermotolerant coliforms do not grow in biofilms, although they most likely can be trapped and retained within biofilms.

TABLE 3-2 Coliform Isolates Typically Found in Drinking Water

Citrobacter

Enterobacter

Escherichia

Klebsiella

C. freundii

E. aerogenes

E. coli

K. pneumonia

C. diversus

E. agglomerans

 

K. oxytoca

 

E. cloacae

 

K. rhinoscleromatis

 

 

 

K. ozaena

SOURCE: Adapted from Geldreich and LeChevallier (1999).

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Aeromonas. Aeromonas spp. are Gram-negative bacteria found in fresh and salt water and cause a wide variety of human infections including septicemia, wound infections, meningitis, pneumonia, respiratory infections, hemolytic uremic syndrome, and gastroenteritis (Carnahan and Altwegg, 1996; Alavandi et al., 1999). The ability of these microorganisms to grow at low temperatures and low nutrient conditions are important in their occurrence in drinking water supplies. Through the Unregulated Contaminant Monitoring Rule (see Chapter 2), EPA examined the occurrence of Aeromonas spp. in 308 drinking water systems and found detectable concentrations in 2.6 percent of 5,060 samples and in 13.6 percent of the systems. In a 16-month study conducted on the presence of A. hydrophila in drinking water in Indiana, 7.7 percent of the biofilm samples were positive for A. hydrophila (Chauret et al., 2001). The health significance of detecting aeromonads in drinking water is not well understood. Some countries (such as the Netherlands) have set standards for aeromonads in drinking water leaving the treatment plant (< 20 CFU/200 mL) and in the distribution system (< 200 CFU/100 mL).


Mycobacteria. Organisms of the genus Mycobacteria are also found in drinking water. Of particular concern is the MAC, or Mycobacterium avium complex. Studies have detected M. avium complex organisms in drinking water distribution systems with concentrations ranging between 0.08 and 45,000 CFU/mL (Haas et al., 1983; duMoulin and Stottmeir, 1986; Carson et al., 1988; duMoulin et al., 1988; Fischeder et al., 1991; von Reyn et al., 1993; Glover et al., 1994; von Reyn et al., 1994; Covert et al., 1999). M. avium are resistant to disinfectants, especially free chlorine (Taylor et al., 2000). Indeed, it is postulated that they may in fact be selected for in distribution systems as a result of their resistance to chlorine (Collins et al., 1984; Schulze-Robbecke and Fischeder, 1989; Briganti and Wacker, 1995). However, there is also evidence that MAC are susceptible to chlorine dioxide and chloramine (Vaerewijck et al., 2005).

Falkinham et al. (2001) examined eight, well characterized drinking water systems and reported that 20 percent of the water isolates and 64 percent of the biofilm isolates were identified as M. avium or M. intracellulare. Additionally, 8 percent of the water isolates were identified as M. kansasii. Most of these isolates were detected in raw water samples, with M. avium complex organisms detected in five of six surface water sites ranging from 6 to 35 percent of the organisms isolated. M. avium complex organisms were not detected in any plant or well effluent sample, but were occasionally detected at low levels (< 1 CFU/mL) in drinking water systems. However, M. avium and M. intracellulare were recovered frequently from drinking water biofilm samples, indicating that M. avium levels were increasing in the distribution system. Increases in M. avium levels in drinking water were correlated to levels of AOC (r2 = 0.65, p = 0.029) and BDOC (r2 = 0.64, p = 0.031) (Falkinham et al., 2001; LeChevallier, 2004).

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

The greatest increase in M. avium complex infections have been with acquired immunodeficiency syndrome (AIDS) patients; approximately 25 to 50 percent of these patients suffer debilitating and life-threatening MAC infections (Horsburgh, 1991; Nightingale et al., 1992), although the availability of highly active antiretroviral therapy has reduced the incidence of MAC in AIDS patients in recent years. Members of the MAC are known opportunistic pathogens, with symptoms of pulmonary infection mimicking that of M. tuberculosis (Wolinsky, 1979). The organism infects the gastrointestinal or pulmonary tract, suggesting that food or water may be important routes of transmission for AIDS patients (Singh and Lu, 1994). It should be pointed out that epidemiology studies have not yet identified drinking water as a risk factor for MAC, except perhaps in hospital water systems.

Free-Living Protozoa

Of the genera of protozoa present in distribution systems, Acanthamoeba, Hartmanella and Naegleria are known to feed on bacteria and biofilms by grazing. Previous research has shown that all coliforms as well as bacterial pathogens and opportunistic pathogens may be ingested by protozoa. Ingested bacteria, if not digested, may survive within the protozoa and be protected from residual disinfectant. The survival of Legionella has been the subject of numerous reports in the literature with regards to its increased resistance to disinfectants while in the intracellular state (Levy, 1990).

Of the eucaryotes mentioned above, two are known to be pathogenic—Naegleria spp. and Acanthamoeba. These are usually associated with recreational rather than drinking waters, although Acanthamoeba was included as part of the first Contaminant Candidate List (EPA, 1998) as an opportunistic pathogen affecting contact lens wearers. Previous studies have shown that these organisms are usually found at the source. However, cysts have also been isolated from drinking water distribution systems in France (Jacquemin et al., 1981; Geldreich, 1996).

Routine monitoring for free-living protozoa is rarely done. Isolation and identification of these organisms are accomplished only when there is evidence for disease outbreak or when research studies are being conducted. As interest in the ability for protozoa to harbor bacterial pathogens increases, it is probable that more effort will be expended in determining their presence in distribution systems, including premise plumbing.

Fungi

Although many fungi have been found in drinking water systems, their levels are typically low and the organisms have not been directly associated with disease (Kelley et al., 2003). The origin of fungi in drinking water systems has

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

not been well characterized, but it is assumed that they come from environmental sources including surface water and groundwater, soils, and vegetation. The four most frequently occurring genera of filamentous fungi isolated from chlorinated and unchlorinated distribution systems in southern California were Penicillium, Sporocybe, Acremonium, and Paecilomyces (Nagy and Olson, 1982). Aspergillus fumigatus was the predominant species detected in the distribution system water supplies in Finland (Niemi et al., 1982). A variety of fungi (Cephalosporium sp., Verticillium sp., Trichodorma sporulosum, Nectria veridescens, Phoma sp., and Phialophora sp.) were identified from water service mains in England (Bays et al., 1970; Dott and Waschko-Dransmann, 1981). Outside of specialized research studies, potable water supplies are not routinely tested for fungi.

The Microbiology of Distribution System Biofilms

Biofilms in drinking water pipe networks contain all of the organisms mentioned above that are found in bulk distribution system water, as well as others. The microbial composition of any given pipe segment can be highly variable, and in most cases is poorly, if ever, characterized. The pipe surface itself can influence the composition and activity of biofilm populations. Studies have shown that biofilms developed more quickly on iron pipe surfaces than on plastic PVC pipes, despite the fact that adequate corrosion control was applied, that the water was biologically treated to reduce AOC levels, and that chlorine residuals were consistently maintained (Haas et al., 1983; Camper, 1996).

In addition to influencing the development of biofilms, the pipe surface has also been shown to affect the composition of the microbial communities present within the biofilm (Figure 3-1). Iron pipes supported a more diverse microbial population than did PVC pipes (Norton and LeChevallier, 2000). Undoubtedly part of the reason that certain bacteria associate with certain pipe types is because materials may leach compounds that support bacterial growth. For example, pipe gaskets and elastic sealants (containing polyamide and silicone) can be a source of nutrients for bacterial proliferation. Colbourne et al. (1984) reported that Legionella were associated with certain rubber gaskets. Organisms associated with joint-packing materials include populations of Pseudomonas aeruginosa, Chromobacter spp., Enterobacter aerogenes, and Klebsiella pneumoniae (Schoenen, 1986; Geldreich and LeChevallier, 1999). Coating compounds for storage reservoirs and standpipes can contribute organic polymers and solvents that may support regrowth of heterotrophic bacteria (Schoenen, 1986; Thofern et al., 1987). Liner materials may contain bitumen, chlorinated rubber, epoxy resin, or tar-epoxy resin combinations that can support bacterial regrowth (Schoenen, 1986). PVC pipes and coating materials may leach stabilizers that can result in bacterial growth. Studies performed in the United Kingdom reported that coliform isolations were four times higher when samples were collected from plastic taps than from metallic faucets (cited in Geldreich and

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

LeChevallier, 1999). The purpose of these studies was not to indicate that certain pipe materials are preferred over another, but to demonstrate the importance of considering the type of materials that come into contact with potable water. Although procedures are available to evaluate the growth stimulation potential of different materials (Bellen et al., 1993), these tests are not applied in the United States by ANSI/NSF.

FIGURE 3-1 Microbial populations isolated from iron pipe (A) or PVC (B) surfaces. SOURCE: Adapted from Norton and LeChevallier (2000).

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

***

For both bulk drinking water and biofilms, the identification of microorganisms typically relies on culturing bacteria from potable supplies, which has important limitations. Culture methods do not detect all microbes that may exist in water, such that only a fraction of viable organisms is recovered (Amann et al., 1995). In addition, most culture methods only detect relatively rapidly growing heterotrophic bacteria, and slowly growing organisms, fastidious or autotrophic organisms, and anaerobes are generally not examined. Diagnostic kits are unreliable for many heterotrophic bacteria because the methodology often requires the analyst to perform a Gram stain, which is difficult because of the slow growth and acid-fast or partially acid-fast nature of bacteria surviving in disinfected drinking water.

An alternative method includes fatty acid profiling. As shown above, this approach can be used to identify organisms from drinking water (Norton and LeChevallier, 2000) but in this study the organisms were cultured prior to identification and therefore the limitations associated with culturing are still present. Additionally, for identification, the lipid profile must match an established profile in a database; these databases are predominated by medical (and not environmental) organisms. The use of fatty acid profiles was further developed by Smith et al. (2000) who used biofilm samples without prior culturing to demonstrate that predominantly Gram-negative bacteria were present, but no further identification was accomplished. A similar approach was taken by Keinanen et al. (2004) who compared profiles from two drinking water systems and showed that they differed, but again, no identifications were obtained. Although fatty acid profiling has been used in these studies to provide some insight on microbial ecology, the limitations associated with the method preclude it from extensive use in characterizing mixed microbial communities.

Molecular methods offer the promise of a more complete determination of the microbiology of drinking water (see Chapter 6 for details). DNA extraction coupled with polymerase chain reaction (PCR) amplification can be used to identify waterborne microbes (Amann et al., 1990, 1995). These procedures can be combined with quantitative real-time PCR, fluorescence in-situ hybridization, or flow cytometry to provide quantitative assessments of bacterial populations. However, careful quality assurance is necessary to ensure complete extraction and recovery of environmental DNA. Martiny et al. (2003) utilized terminal restriction fragment length polymorphisms to identify members of a biofilm consortium over a three-year time period. In this study, several organisms were identified (Pseudomonas, Sphingomonas, Aquabacterium, Nitrospira, Planctomyces, Acidobacterium) but for the majority of the peaks no sequence match could be made.

It is telling that there is very little published information about the microbial ecology of distribution systems. At this point in time, the detection methods are expensive, are time consuming, require optimization for specific conditions, and are appropriate only for the research laboratory. As a consequence, there is

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

a lack of information about the types, numbers, and activities of microorganisms in drinking water. It is also unknown how the ecology of the main distribution system is related to that in premise plumbing, how the populations vary between distribution systems in different locations, and how the populations respond to water quality changes within a distribution system. This translates into a lack of understanding about whether organisms of potential public health concern may be present in water systems and further complicates the ability to assess risk due to their presence.

As mandated by the Safe Drinking Water Act, the EPA has issued a second Contaminant Candidate List that includes 10 microbes (or microbial products) for potential future regulation (EPA, 2004) (see Table 3-3). For most of these microbes, methods do not exist for routine testing of drinking water supplies, and basic research is needed on their occurrence, survival, and importance in potable water. Where the current list includes organisms that are not discussed above, they are considered to be of primary concern in untreated or inadequately treated source waters and not in distribution systems, such that a more detailed discussion is beyond the scope of the report.

It can be hard to determine whether the detection of frank or opportunistic pathogens in drinking water poses an unacceptable risk. In addition to the monitoring techniques being difficult, time-consuming, expensive, and of poor sensitivity, the methods do not detect specific virulence determinants, such that many environmental isolates (e.g., E. coli, Aeromonas, Legionella, etc.) are indistinguishable from their clinical strains. Therefore even when monitoring for potentially pathogenic organisms is done, the public health significance of the results is often in question. Furthermore, there is insufficient supporting information (in terms of occurrence data for exposure assessment, dose-response data, health effects, and models that can predict pathogen occurrence for different distribution system contamination scenarios such as contamination via cross connections, main breaks, or intrusion) to conduct a risk assessment for many waterborne microbes. For all these reasons, measurement of the microbe itself is

TABLE 3-3 Contaminant Candidate List Microbes

Bacteria

Mycobacterium avium

 

Helicobacter

 

Aeromonas

Viruses

Caliciviruses

 

Echovirus

 

Coxsackieviruses

 

Adenovirus

Protozoa

Microsporidium

Toxins

Cyanobacterial toxins

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

typically insufficient to make a public health determination. Until better monitoring methods, pathogen occurrence models, dose-response data, and risk assessment data are available, pathogen occurrence measurements are best used in conjunction with other supporting data on health outcomes. Such supporting data could include enhanced or syndromic surveillance in communities, as well as the use of microbial or chemical indicators of potential contamination.

EVIDENCE FROM OUTBREAK DATA

Most information on the risks of waterborne disease in the United States comes from surveillance and investigation of waterborne disease outbreaks. A passive voluntary surveillance system for waterborne disease outbreaks started in 1971 and is a collaboration between the Centers for Disease Control and Prevention (CDC), the EPA, and state and regional epidemiologists. This surveillance system includes outbreaks associated with both drinking and recreational water, and outbreaks due to both microbial and chemical agents. The objectives of the surveillance system are to (1) characterize the epidemiology of waterborne disease outbreaks, (2) identify the etiologic agents that cause the outbreaks, (3) determine the risk factors that contributed to the outbreak, (4) inform and train public health personnel to detect and investigate waterborne disease outbreaks, and (5) collaborate with local, regional, national and international agencies on strategies to prevent waterborne diseases (Stanwell-Smith et al., 2003).

From 1971 through 2002, 764 drinking water outbreaks have been reported through this surveillance system. Although this is believed to be an underestimate of the true number of outbreaks that occurred during this period, the information collected in this surveillance system has been extremely valuable for improving our understanding of the agents that cause waterborne disease and the risk factors involved in waterborne disease outbreaks. The data collected in this surveillance system includes:

  • Type of exposure (drinking water or recreational water)

  • Location and date of outbreak

  • Actual or estimated number of persons exposed, ill, hospitalized, dead

  • Symptoms, incubation period, duration of illness

  • Etiologic agent

  • Epidemiological data (attack rate, relative risk or odds ratio)

  • Clinical laboratory data (results of fecal and serology tests)

  • Type of water system

    • Community, non-community, or individual homeowner drinking water supply

    • Swimming pool, hot tub, water park, or lake for recreational water

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
  • Environmental data (results of water analyses, sanitary survey, water plant inspection)

  • Factors contributing to contamination of water

The surveillance data are summarized in biannual reports (Morbidity and Mortality Weekly Report Surveillance Summaries) that are published by the CDC and distributed to public health authorities and practitioners throughout the country. The information is also available on the Internet at http://www.cdc.gov/mmwr. These reports (Herwaldt et al., 1991; Moore et al., 1993; Kramer et al., 1996; Levy et al., 1998; Barwick et al., 2000; Lee et al., 2002a; Blackburn et al., 2004) indicate three main trends:

  1. The overall number of reported waterborne disease outbreaks associated with drinking water is declining from a peak of over 50 reported outbreaks in 1980 to eight reported outbreaks in 2002.

  2. For a substantial portion of drinking water outbreaks, the pathogen is not identified and the outbreaks are classified as “acute gastrointestinal illness of unknown etiology” (AGI). From 1986 through 2002, approximately 41 percent of the over 250 outbreaks reported during this period were classified as AGI, and this proportion varies by reporting period from a peak of 68 percent in 1991–1992 to 17 percent in 1993–1994. Overall, Giardia and Cryptosporidium are the most commonly reported etiologic agents of waterborne disease when a pathogen is identified and are associated with about 20 percent of reported outbreaks associated with drinking water since the mid-1980s. However, with the recent addition of Legionella outbreaks to the surveillance system, Legionella is now the single most common cause of outbreaks involving drinking water (as discussed below).

  3. Most drinking water outbreaks involve groundwater systems, especially untreated groundwater systems. Forty (40) percent of the 25 drinking water outbreaks reported between 2001 and 2002 involved untreated groundwater systems (Blackburn et al., 2004).

Declining Number of Drinking Water Outbreaks

Since the mid-1980s, the number of waterborne outbreaks has declined (Figure 3-2). The reason for the decrease is largely attributed to the promulgation of more stringent drinking water regulations, including the Surface Water Treatment Rule, the Total Coliform Rule, and others. In addition, many water utilities have made voluntary improvements, such as the Partnership for Safe Water program to reduce the risk of waterborne cryptosporidiosis. The Partnership program entails a comprehensive evaluation of treatment practices with a focus on achieving filtered drinking water turbidities less than 0.1 nephelometric turbidity units (NTU). The number of reported outbreaks began to decrease sharply beginning with the 1985–1986 reporting period; this was attributable

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

FIGURE 3-2 Number of drinking water disease outbreaks in the United States, 1971–2002. Individual—private or individual water systems (9 percent of U.S. population or 24 million users); Community—systems that serve > 25 users year round (91 percent of U.S. population or 243 million users); Noncommunity—systems that serve < 25 users and transient water systems such as restaurants, highway rest areas, parks (millions of users yearly). SOURCE: Blackburn et al. (2004).

primarily to fewer community and noncommunity outbreaks. With the institution and enforcement of better regulations that chiefly affect these types of water systems (particularly community systems), a marked drop in the number of outbreaks was seen. In contrast, the increase in outbreaks reported during 1999– 2000 was attributable primarily to individual homeowner systems, which affect fewer persons, are less regulated, or are more subject to changes in surveillance and reporting. In 2001–2002, individual homeowner systems comprised 40 percent of the waterborne outbreaks (Figure 3-3).

Etiologic Agents Associated With Drinking Water Outbreaks

The agents responsible for waterborne disease outbreaks were predominantly undefined, microbial (parasitic, bacterial, or viral), or chemical. Indeed, surveillance data on waterborne disease outbreaks associated with drinking water in the United States from 2001 to 2002 indicate that almost 30 percent of reported outbreaks were due to bacterial agents, 16 percent were due to protozoa, 16 percent were due to viral agents, 16 percent were due to chemical contaminants, and 23 percent had an unidentified etiology. Figure 3-4 shows the

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

FIGURE 3-3 Waterborne outbreaks by etiological agent, water system, water source, and deficiency—United States, 20012002. SOURCE: Blackburn et al. (2004).

etiology of waterborne disease outbreaks over time. The large number of waterborne disease outbreaks associated with protozoa in the early 1980s was mostly caused by Giardia and was greatly reduced by the implementation of the Surface Water Treatment Rule in 1989 (Barwick et al., 2000). Relatively few outbreaks due to viruses have been reported, in part because of the difficulty of the detection methodologies for these organisms. However, the number of reported viral outbreaks has increased significantly since 1999 with the development of better diagnostic techniques for noroviruses. Nine of the 15 drinking water outbreaks associated with noroviruses that have been reported since 1986 occurred between 1999 and 2002 (Herwaldt et al., 1991; Moore et al., 1993; Kramer et al., 1996; Levy et al., 1998; Barwick et al., 2000; Lee et al., 2002a; Blackburn et al., 2004).

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

FIGURE 3-4 Agents responsible for waterborne outbreaks. SOURCE: Blackburn et al. (2004).

Over the past 30 years, there has been a wide range of chemical agents associated with drinking water outbreaks, including arsenic, benzene, chlordane, chlorine, chromate, copper, cutting oil, developer fluid, ethyl acrylate, ethylene glycol, fluoride, fuel oil, furadan, lead, leaded gasoline, lubricating oil, kerosene, nitrate, nitrite, phenol, polychlorinated biphenyls, selenium, sodium hydroxide, toluene, xylene, and unidentified herbicides. From 1993 through 2002, most drinking water outbreaks associated with chemical agents have been due to copper (eight outbreaks, usually related to premise plumbing) followed by nitrates/nitrites (six outbreaks, usually related to contamination of groundwater) (Kramer et al., 1996; Levy et al., 1998; Barwick et al., 2000; Lee et al., 2002a; Blackburn et al., 2004).

Outbreaks Associated With Groundwater Systems

In recent years, as treatment of surface water supplies has improved, waterborne outbreaks have increasingly involved groundwater supplies (Figure 3-3). There is increasing recognition that many groundwater supplies have microbial contamination, yet the use of untreated groundwater continues in many small communities and by individual homeowners. A survey of 448 wells in 35 states reported that 31 percent of the sites were positive for at least one virus, and enterovirus RNA was detected in approximately 15 percent, rotavirus RNA in 14 percent, and hepatitis A virus RNA in 7 percent of the wells by reverse-tran-

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

scription-polymerase chain reaction (RT-PCR) (Abbaszadegan et al., 2003). Fout et al. (2003) examined 321 samples from 29 groundwater sites by RT-PCR and reported that 72 percent of the sites were virus positive. Borchardt et al. (2004) collected monthly samples from four municipal wells in one city in Wisconsin for a 12-month period and detected enteric viruses by RT-PCR in 50 percent of the samples. Two studies in Ontario, Canada examined the relationship between E. coli in well water and acute gastrointestinal illness in households using the water for drinking (Raina et al., 1999, Strauss et al., 2001). In the first study of 181 households with untreated well water, water samples were collected five times during the one-year study, and E. coli was detected in 20 percent of the household wells. The second study included 235 households in four rural communities (Strauss et al., 2001) and reported that 20 percent of the households had at least one water sample that exceeded the national standards for total coliforms or E. coli.

Outbreaks Associated With Distribution Systems

Among the seven outbreaks associated with community water systems in 2001–2002, four (57.1 percent) were related to problems in the water distribution system. Preliminary results from the 2003–2004 surveillance report indicate that distribution systems were associated with 38 percent of the outbreaks associated with drinking water systems during this period (Liang et al., 2006). Other epidemiological and outbreak investigations conducted in the last five years suggest that a substantial proportion of waterborne disease outbreaks, both microbial and chemical, is attributable to problems within distribution systems (Craun and Calderon, 2001; Blackburn et al., 2004) (see Figure 1-1). Craun and Calderon (2001) examined causes of reported waterborne outbreaks from 1971 to 1998 and noted that, in community water systems, 30 percent of 294 outbreaks were associated with distribution system deficiencies, causing an average of 194 illnesses per outbreak. Distribution system contamination was observed to be the single most important cause of outbreaks in community water systems over that time period.

The reason for the apparent increase in the proportion of outbreaks associated with water distribution systems is not entirely clear. Outbreaks associated with distribution system deficiencies have been reported since the surveillance system was started. However, there may be more attention focused on the distribution system now that there are fewer outbreaks associated with inadequate treatment of surface water. Also, better outbreak investigations and reporting systems in some states may result in increased recognition and reporting of all the risk factors contributing to the outbreak, including problems with the distribution system that may have been overlooked in the past. Although waterborne disease outbreaks in general are still under-reported, the surveillance system has become more mature, and outbreak investigations and analyses are becoming more sophisticated.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

The CDC surveillance system for waterborne disease outbreaks attempts to collect information on outbreaks and their contributing causes. For example, from 1981 to 1998, the CDC documented 57 waterborne outbreaks related to cross-connections, resulting in 9,734 detected and reported illnesses (Craun and Calderon, 2001). Contamination from cross-connections and backsiphonage were found to cause the majority of the outbreaks associated with distribution systems (51 percent), compared with contamination of water mains following breaks (39 percent) and contamination of storage facilities (the remaining 10 percent). A separate compilation by the EPA of backflow events revealed many more incidents of backflow and resulting outbreaks—a total of 459 incidents resulting in 12,093 illnesses from backflow events from 1970 to 2001 (EPA, 2002). The situation may be of even greater concern because incidents involving premise plumbing are even less recognized.

Most reported outbreaks associated with distribution systems occur in community water systems because of their greater size and complexity. For example, from 1999 to 2002 there were 18 reported outbreaks in community water systems, and nine (50 percent) of these were related to problems in the water distribution system (Lee et al., 2002b; Blackburn et al., 2004). However, there have been a number of reported outbreaks associated with noncommunity water systems that have been attributed to deficiencies in the distribution system. Finally, the magnitude and severity of reported outbreaks associated with distribution systems vary, with an average about almost 200 illnesses per outbreak (Craun and Calderon, 2001) and a total of 13 deaths.

The Extent of Underestimation

The number of identified waterborne disease outbreaks is considered an underestimate because not all outbreaks are recognized, investigated, or reported to health authorities (Blackburn et al., 2004). For example, outbreaks occurring in national parks, tribal lands, or military bases might not be reported to state or local authorities. Factors influencing whether a waterborne outbreak is recognized include awareness of the outbreak, availability of laboratory testing, and resources available for surveillance and investigation of outbreaks. The detection and investigation of waterborne outbreaks is primarily the responsibility of the local, state, and territorial public health departments with varying resources and capacities. Differences in the capacity of local and state public health agencies and laboratories to detect an outbreak might result in reporting and surveillance bias, such that the states with the majority of outbreaks might not be the states with the majority of waterborne disease. Outbreaks are more likely to be recognized when they involve acute illnesses with symptoms requiring medical treatment, or when sensitive laboratory diagnostic methods are readily available. These and other limitations are discussed below.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Underreporting of Outbreaks Involving Individual Homeowner Systems

Although the surveillance system has always included outbreaks associated with individual homeowner water systems, it is likely that most sporadic cases and small clusters of waterborne disease associated with individual homeowner water systems are not recognized or reported because small numbers of people are involved. Furthermore, a cluster of cases of gastroenteritis within a single household may easily be attributed to food contamination or person-to-person transmission, such that the possibility of waterborne transmission may not be considered or investigated. From 1971 to 1980, 37 (11.6 percent) of the 320 reported drinking water outbreaks were associated with individual homeowner systems, and most of these outbreaks involved chemical agents when an etiologic agent was identified (Craun, 1986). From 1993 to 2002, 41 (28.7 percent) of the 143 reported drinking water outbreaks were associated with individual homeowner water systems, suggesting that there may be increased recognition and reporting of these smaller outbreaks in the past ten years of surveillance.

Underreporting of Outbreaks Involving Premise Plumbing

Outbreaks associated with premise plumbing are not specifically identified in the CDC surveillance reports. Adverse health effects associated with premise plumbing problems are less likely to be recognized and reported in this surveillance system, especially if they occur within a single household. However, a number of outbreaks associated with drinking water have been reported from public building settings such as schools, restaurants, churches, factories, and apartment buildings. Some of these outbreaks were due to contamination of a private well that serves the building. Other outbreaks in public buildings were classified as due to distribution system deficiencies and appeared to involve cross-connections and/or backsiphonage problems. Examples of the latter type of outbreak include:

  • an outbreak of copper poisoning in the early 1980s that occurred when “backsiphonage of corrosive water containing carbon dioxide from a sodamixing dispenser caused copper to be leached from piping in a building (Craun, 1986);

  • a norovirus outbreak in 1995 at a high school in Wisconsin that affected 148 persons. The school was connected to the community water supply. However, water in the school became contaminated from backsiphonage of water from hoses submerged in a flooded football field (Levy et al., 1998);

  • a chemical outbreak in 1995, in which 13 persons in a healthcare facility in Iowa became ill after drinking water that was contaminated with concentrated liquid soap. A valve on the water supply hose to the soap dispenser had been left open and allowed the soap to enter the water supply in the building.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Although the building had vacuum breakers to prevent backsiphonage, these were installed incorrectly at the soap dispensers (Levy et al., 1998);

  • a chemical outbreak in 1999, in which four residents of an apartment building in Florida had acute gastroenteritis that was attributed to unidentified chemical poisoning. A cross-connection was discovered between their drinking water and an improper toilet flush-valve. Residents of the apartment had noticed on several occasions that their tap water was blue before the onset of illness (Lee et al., 2002a).

  • a small waterborne disease outbreak at a middle school in Florida in 2001 due to a cross-connection between the air conditioning unit and the potable water supply. A maintenance worker used the potable water system to dilute the ethylene glycol solution in the chiller unit. The higher water pressure in the chiller unit forced the diluted ethylene glycol into the school’s water supply and pink-colored water was observed in the school bathrooms. Three students became ill with gastrointestinal symptoms (Blackburn et al., 2004).

Underreporting of Outbreaks Involving Chemical Agents

From 1971 to 1980, 38 (11.9 percent) of the 320 reported drinking water outbreaks were attributed to chemical agents (Craun, 1986), and from 1993 to 2002, 25 (17.5 percent) of the 143 reported drinking water outbreaks were attributed to chemical agents (Kramer et al., 1996; Levy et al., 1998; Barwick et al., 2000; Lee et al., 2002a; Blackburn et al., 2004). The CDC believes that waterborne chemical poisonings are underreported for many reasons. First, most of these are probably due to copper and lead leaching from plumbing in private residences and affect relatively few people and are consequently unlikely to be recognized by public health authorities. Furthermore, exposure to chemicals in drinking water can often cause non-specific symptoms that may not be recognized as chemical poisoning or may not be linked to a specific chemical. The detection, investigation, and reporting of waterborne disease outbreaks linked to chemical exposures are not as well established as the methods for dealing with outbreaks associated with infectious agents. Finally, many physicians may have difficulty recognizing and diagnosing chemical poisonings unless they have had additional training in this area (Barwick et al., 2000).

Revisions of the CDC Waterborne Disease Outbreak Surveillance System

The CDC is making several changes to its waterborne disease outbreak surveillance system that are relevant to better understanding the role of distribution systems, including premise plumbing. Previously, the risk factors or deficiencies that contributed to a waterborne disease outbreak were classified as: (1) use of untreated surface water, (2) use of untreated groundwater, (3) treatment deficiency, (4) distribution system problem, or (5) miscellaneous. The 2003–2004

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

MMWR Surveillance Summary will use a new and detailed classification system for risk factors that contributed to the outbreak, and it will distinguish between deficiencies before or after entry into a building or home. This distinction is important because drinking water before it enters a building is usually managed by the water utility and subject to EPA drinking water regulations. However, drinking water problems that occur after entry into a building, such as those due to Legionella colonization in premise plumbing, cross-connections, point-of-use devices, or drink mix machines, may not be the responsibility of the water utility or regulated by EPA (lead and copper are an exception—see Chapter 2). Preliminary results from the surveillance system for 2003–2004 indicate that 48 percent of the outbreaks associated with drinking water were associated with deficiencies in source water, water treatment, and the distribution system and 52 percent of the outbreaks were due to deficiencies after the point of entry. In this latter group of outbreaks, approximately 47 percent involved Legionella and 35 percent involved chemical agents (including copper) (Liang et al., 2006). In addition, the surveillance system will now report all the identified deficiencies that contributed to the waterborne disease outbreak rather than reporting only the primary deficiency. Finally, CDC is moving toward a web-based system for reporting outbreaks and developing a public access database on waterborne disease outbreaks that will allow investigators to examine and analyze these data.

EPIDEMIOLOGY STUDIES

Three basic epidemiological study designs can be used to assess the public health risk of contaminated water supplies (Steenland and Moe, 2005): descriptive, correlational or ecological, and analytic. In the descriptive study, population surveys or systematic disease surveillance describe disease patterns by various factors such as age, seasonality, and geographic location. These studies do not test a formal hypothesis about the relation between a specific exposure (or risk factor) and disease, but they can help identify specific populations or geographic regions for further study. This category includes the systematic surveillance of outbreaks discussed in the previous section as well as endemic cases. Surveillance systems are useful for showing trends in the causes and risk factors of waterborne disease, but they are not very sensitive and cannot serve as a rapid warning system of a water-related health problem in a specific community because of reporting delays. In addition to the waterborne disease outbreak surveillance system, there is also a national system of notifiable diseases in the United States that mandates that health care providers report specific infections, including a number of potentially waterborne infections such as cholera, cryptosporidiosis, E. coli O157:H7, giardiasis, hepatitis A virus, legionellosis, poliomyelitis, salmonellosis, shigellosis, tularemia, and typhoid fever. Like the outbreak surveillance system, the surveillance for notifiable diseases is a voluntary passive surveillance system with low sensitivity and reporting delays. Finally,

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

the descriptive framework has been used in the Foodnet surveillance program to assess occurrence of common gastroenteric illnesses in the population and gather information on the prevalence of various risk factors for diarrheal disease (such as food consumption habits, water consumption habits, and recreational water contact). Although the notifiable disease surveillance system and the Foodnet program provide valuable data on disease occurrence, they provide no information on what proportion of these diseases are related to drinking water.

Correlational or ecologic studies collect population level data on disease rates and exposures and look for correlations. For example, bladder cancer rates in cities with chlorinated surface water can be compared to cities with chlorinated groundwater to see if there may be a correlation between chlorination of surface water, formation of DBPs, and bladder cancer. However, these studies do not collect information on individual risk factors or confounders that may be related to risk of disease, such as smoking. Correlational studies do not test a formal hypothesis and are considered weaker than studies that collect individual-level data. But they can provide valuable information for generating hypotheses. Time-series studies are another example of correlational studies and have been used to examine the relationship between changes in water quality indicators (such as turbidity) and disease rates in the population served by the water supply (such as emergency department visits for gastroenteritis) (Schwartz et al., 1997). These studies have the advantage of comparing the same population at different points in time (thus controlling for confounding) so that only the variables that change are those that are being studied—i.e., water quality and disease rates.

Analytical studies are those in which individual-level data are collected, and the investigator tests a formal hypothesis about the association between exposure and disease. Analytical studies can be experimental, such as a clinical trial where some households are given bottled water to drink and other households are asked to drink tap water, and then disease rates between the two study groups are compared to determine the risk of disease attributable to drinking water. In these clinical trials, study participants are randomly assigned to a study group in order to ensure that other potential risk factors for disease are equally distributed among the study groups. An example of this design is the study of Colford et al. (2002) in which home water purification devices were installed in the homes of a test group of study participants and the control group consisted of homes in which “sham” devices were installed. Both groups kept health diaries to record symptoms of gastroenteritis and other health effects. At the end of the observation period, incident rates of disease were compared as a ratio, e.g., diarrhea episodes per person-year in the “exposed group” (those with the sham device) divided by diarrhea episodes per person-year in the “unexposed group” (those with additional purification).

Other analytical studies can be observational or natural experiments, where the investigator examines disease rates over time in study groups that have different exposures. Observational studies can use a cohort design, case-control design, or cross-sectional design. In the cohort design, all study participants are

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

disease-free at the beginning of the study and disease rates over time are compared between study participants who are exposed to various risk factors vs. those who are not exposed. This design allows the consideration of multiple health outcomes and can be either prospective or retrospective. The cohort design is useful for rare exposures because the study deliberately recruits a cohort of individuals who are more likely to become exposed because of their occupation or geographic location. An example of this is the study of Frost et al. (2005), who assessed the illness rate of cryptosporidiosis and the presence of antibodies to Cryptosporidium in two populations (one exposed to surface water and one to groundwater). They concluded that populations receiving surface-derived water had higher antibody prevalence (but not higher illness rate) than individuals receiving groundwater. Cohort studies are not well suited for rare diseases because the purpose of this study design is to compare how frequently the disease occurs in the exposed group vs. the unexposed group. If the disease is rare, then a very large cohort must be recruited in order to make a meaningful comparison.

Case-control studies are often used to study rare diseases and start with recruiting a group of individuals with the disease of interest (cases) and another group of individuals without the disease (controls). The study individuals are then queried about their past exposure to the specific risk factors of interest. In a case-control study, the measure of association is the “risk odds ratio” which compares the odds of exposure to a specific risk factor among the cases to the odds of exposure among the controls. In contrast to the cohort study, a casecontrol study can look at only one health outcome but can examine multiple risk factors. An example of the case-control design is the study of Steinmaus et al. (2003) who examined associations of risk factors with bladder cancer in the western U.S. This study found no association of bladder cancer with daily arsenic ingestion in drinking water below 80 µg/day and found some association in smokers at ingestions of greater than 200 µg/d of arsenic.

Cross-sectional studies are similar to ecologic studies in that exposure rates and disease rates are measured at the same time. However, cross-sectional studies collect individual-level data whereas ecologic studies collect population-level data. Seroprevalence surveys are a form of cross-sectional study where, for example, prevalence of antibodies to Cryptosporidium can be measured in populations served by different types of water supplies. The use of epidemiological methods to study health risks associated with drinking water has been reviewed by Savitz and Moe (1997).

Descriptive Studies of Endemic Waterborne Disease

The risk of endemic waterborne disease (sporadic cases) is difficult to estimate, although various authors have made educated guesses. Bennett et al. (1987) estimated that the incidence of waterborne disease in the United States was 940,000 cases per year and resulted in 900 deaths. Although the purpose of

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

the study was to rank the importance of various disease categories (water ranked next to the last, above zoonotic diseases) and to define the opportunities for prevention, the study has been criticized as little more than an exercise in guess work. Morris and Levin (1995) used incidence rates for enteric diseases and prevalence rates for specific groups of pathogens detected in water to give waterborne infectious disease estimates of 7.1 million mild infections, 560,000 cases of moderate or severe illness, and 1,200 deaths annually in the United States. The authors concluded, however, that available data were inadequate to refine the estimates.

Recent data on the incidence of diarrheal disease in the U.S. is available from the FoodNet population-based surveillance system (managed by the CDC). The disease estimates from the FoodNet system are based on telephone surveys that used random-digit-dialing and interviewed one individual per household to recall their occurrence of diarrhea in the four weeks prior to the interview. As shown in Table 3-4, the overall diarrhea prevalence rates from these surveys range from 5 to 11 percent, resulting in an estimated incidence of around 0.7 to 1.4 episodes/person/year. Diarrhea prevalence rates were consistently higher in children under five years of age.

Other CDC estimates based on the FoodNet data and other sources suggest that there are 211 million episodes of acute gastroenteritis in the United States each year that result in over 900,000 hospitalizations and 6,000 deaths (Mead et al., 1999). Mead et al. (1999) estimated the incidence of gastrointestinal illness to be 0.79 episodes/person/year. These FoodNet data are valuable for providing a measure of baseline diarrhea incidence in the U.S. population and the public health and economic burden associated with diarrheal diseases in an industrialized country. However, it is important to point out that these data offer no information on the proportion of diarrheal disease attributable to drinking water. Furthermore, these data probably underestimate the total burden of acute gastroenteritis in the population because cases with only vomiting were not included in the estimate (Imhoff, 2004), and vomiting is a common symptom for most gastroenteritis due to noroviruses and other viral agents.

TABLE 3-4 Burden of Diarrheal Disease in the U.S. based on FoodNet Telephone Survey Data

Year

No. of States

Total # respondents in analysis

Overall prevalence of acute diarrheal illness in past four weeks

Estimated incidence of episodes/person/year

Diarrhea prevalence in children < five years old

1996–1997

5

8,624

11%

1.4

10%

1998–1999

7

12,075

6%

0.72

9%

2000–2001

8

14,046

5%

NA

9%

2002–2003

9

15,578

5%

NA

9%

NA = the authors did not report an estimate of the incidence rate.

SOURCES: Herikstatd et al. (2002); Imhoff et al. (2004) ; Hawkins et al. (2002) ; McMillan et al. (2004).

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Analytical Epidemiological Studies

Determining the proportion of diarrheal disease that is attributable to water contamination is best done through analytical, experimental epidemiological studies. There have been four analytical epidemiological studies of acute gastroenteritis and drinking water systems relevant to distribution systems, all of which focused on risks from microbiological agents.

Laval Studies

Payment et al. conducted two epidemiology studies (Payment et al, 1991; Payment et al, 1997) in a suburb of Montréal known as Laval that examined the health of people who drank tap water and compared the group to people receiving water treated by reverse osmosis to determine which group had higher levels of gastrointestinal illness. In the 1991 study, reverse osmosis units were installed in 299 households (1,206 persons), and another 307 households (1,202 persons) were followed as controls with no device installed. Both groups were monitored for a 15-month period. Highly credible gastrointestinal illness (HCGI) was defined as (1) vomiting or liquid diarrhea with or without confinement to bed, consultation with a doctor, or hospitalization, or (2) nausea or soft diarrhea combined with abdominal cramps with or without absence from school or work, confinement to bed, consultation with a doctor, or hospitalization. The water source for the study area was a river that was contaminated by human sewage discharges, including combined sewer overflows. The community had a single water treatment plant with pre-disinfection, alum flocculation, rapid sand filtration, ozonation, and final disinfection with chlorine or chlorine dioxide. The quality of the finished water leaving the plant included an average of 0.6 mg/L total chlorine and approximately 0.4 mg/L free chlorine, an average turbidity of 0.26 NTU, and no detection of indicator bacteria or human enteric viruses in weekly samples (Payment et al., 1991). The overall incidence of highly credible gastroenteritis was 0.66 episodes/person/year and was highest in children five years of age and younger. The authors concluded that approximately 35 percent of the self-reported gastrointestinal illnesses was attributed to tap water consumption.

The 1997 study included groups receiving (1) regular tap water, (2) tap water from a continuously purged tap, (3) bottled plant effluent water, or (4) bottled plant effluent water purified by reverse osmosis. Differences in gastroenteritis rates between groups 1 and 2 versus group 3 was assumed to be due to changes in water quality that occurred between the time the water left the treatment plant and the time the water reached the household. The water ingested by group 1 represented tap water that had gone through the distribution system and also had residence time in the household plumbing. The water ingested by group 2 represented tap water quality in the distribution system without any significant residence time in the household plumbing. It should be noted that be-

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

tween the time of the first and second study, the water treatment plant was significantly upgraded with higher disinfection doses and better filtration. Estimated Giardia removal/inactivation exceeded 7.4 logs, and estimated virus inactivation by chlorine exceeded 10 logs. The average turbidity of the finished water was 0.1 NTU and never exceeded 0.5 NTU. However, periods of “microfailures” in individual filters were reported (Susan Shaw, EPA, personal communication, 2006).

This second study attributed 14 percent to 40 percent of the gastrointestinal illness to the consumption of tap water (which met Canadian guidelines). Payment et al. (1997) concluded that the distribution system played a role in waterborne disease because the rates of HCGI were similar for group 3 (ingested purified bottled water) and group 4 (ingested bottled water from the treatment plant), but groups 1 and 2 (ingested water from the distribution system) had higher HCGI rates than group 4. Interestingly, there appeared to be no correlation between the relatively short residence time of the water in the distribution system (which varied from 0.3 to 34 hours) and the incidence of HCGI in a family. Furthermore, microbiological testing of the water in the distribution system did not indicate any bacterial indicators of contamination, but these water samples were not tested for viruses or protozoa. Contrary to their expectation, the investigators observed higher HCGI rates in families that ingested water from the continuously purged taps compared to families with regular tap water that may be subject to bacterial regrowth in household pipes. The investigators suggested that the shorter residence time for water from the continuously purged taps may have transported pathogens in the distribution system to the household sooner than regular tapwater and that there may have been inadequate contact time with residual chlorine in the distribution system to inactivate any introduced pathogens.

Transient pressure modeling (Kirmeyer et al., 2001) found that the distribution system studied by Payment et al. was extremely prone to negative pressures, with more than 90 percent of the nodes within the system drawing negative pressures under certain modeling scenarios (e.g., power outages). The system reported some pipe breaks, particularly during the fall and winter when temperature changes placed added stresses on the distribution system. Although the system employed state-of-the-art treatment, the distribution network suffered from low disinfectant residuals, particularly at the ends of the system. Low disinfectant residuals and a vulnerability of the distribution system to pressure transients (suggesting intrusion as a possible mechanism of contamination) could account for the observed illnesses.

Melbourne Study

A double-blinded, randomized trial was recently completed in Melbourne, Australia, to determine the contribution of drinking water to gastroenteritis (Hellard et al., 2001). Melbourne, with a population of about 3 million, draws its

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

drinking water from protected forest catchments of the Upper Yarra and Thomson rivers. The catchments, which are approximately 1,550 square kilometers (600 square miles) in area, are closed to public access and have no permanent human habitation or activity except for logging in limited areas. Water from these catchments is stored in two major reservoirs (Silvan and Cardinia) with detention times of approximately two and 33 months, respectively. Water from both reservoirs is treated by chlorination, fluoridation (slurry or acid), and pH adjustment with lime.

Routine water quality monitoring at sampling points in the distribution system included total and fecal coliforms, HPC bacteria, and total and free chlorine. Free chlorine levels in the distribution system ranged from 0 to 0.94 mg/L, with a median of 0.05 mg/L, and 90 percent of samples had < 0.20 mg/L. Total coliform bacteria were detected in 18.9 percent of 1,167 routine 100-mL water samples, but fecal coliform bacteria were not detected. Median HPC concentrations were 37 CFU/mL with 13 percent of samples greater than 500 CFU/mL. During the study, water quality monitoring included testing a weekly composite sample from four water mains for selected pathogens: Campylobacter sp., Aeromonas sp., Clostridium perfringens, Cryptosporidium sp. and Giardia sp. These distribution system samples were positive for Aeromonas spp. (50 percent of 68 weekly samples), Campylobacter (one occasion), and Giardia (two positive samples by reverse transcriptase-PCR). No samples had detectable C. perfringens spores or Cryptosporidium parvum oocysts.

The study area in Melbourne is a growing area with relatively new houses and many families with young children. Six hundred (600) families (with at least two children one to 15 years of age) were recruited into the study. Approximately one third of the study households lived in areas of the distribution system with average water residence times of one to 1.5 days. Approximately two thirds of the study households lived in areas of the distribution system with average water residence times of three to four days (maximum six days).

Study households were randomly assigned to receive either a real or placebo water treatment unit installed under the kitchen sink. Functional units were designed to remove viruses, bacteria, and protozoa using microfiltration and ultraviolet light treatment. The study participants completed a weekly health diary reporting gastrointestinal symptoms during the 68-week observation period. The rates of HCGI ranged from 0.79/person/year for those with functional treatment units and 0.82/person/year with the sham devices. The study concluded that the water was not a source of measurable gastrointestinal disease (the ratio of illness rates between the group drinking treated water compared to the normal tap water was 0.99, with a 95 percent confidence interval of 0.85– 1.15; p = 0.85). Analysis of 795 fecal specimens from participants with gastroenteritis did not reveal any difference in pathogen detection rates between the two groups.

This study was not designed to examine the risks from the distribution system separately from the risks associated with the entire water system. However, since there appeared to be no measurable contribution to illness due to drinking

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

water, one may assume that the risks from degraded water quality in the distribution system were also below the detection limit of the study.

Davenport Study

The 1996 amendment to the Safe Drinking Water Act included a mandate to the CDC and the EPA to conduct studies to determine the occurrence of waterborne disease. To address this mandate, EPA scientists conducted several epidemiological studies of waterborne disease, and EPA funded several studies by external investigators, including the pilot study and full-scale study in Davenport, Iowa.

As a preliminary trial to the subsequent epidemiology study, a randomized, triple-blinded, home drinking water intervention trial of 77 households was conducted for four months in Contra Costa County, California (Colford et al., 2002). The drinking water was treated using an under-the-kitchen-sink device that incorporated ultraviolet light and microfiltration. Although the purpose of the trial was to evaluate the “blinding” of the study (e.g., could the participating households detect the active and identical-looking placebo devices), analysis of the data showed that the incidence rate ratio of highly credible gastrointestinal illness (HCGI) (incidence rate of the placebo group divided by the active device group, adjusted for clustering) was 1.32, with a 95 percent confidence interval of 0.75 to 2.33. Given the small study size, the higher rate of HCGI among the placebo group was not statistically significant. The authors concluded, however, that the relative rates of HCGI were consistent with those observed by Payment et al. (1991, 1997). This pilot study is interesting because it provides another estimate of self-reported HCGI rates in a cohort of households followed over time, and it confirmed that study subjects could successfully be blinded to the type of water treatment device they had during the intervention trial.

The full-scale Water Evaluation Trial was conducted in Davenport, Iowa to determine the incidence of gastrointestinal illness associated with consumption of drinking water meeting all federal and state treatment guidelines (LeChevallier et al., 2004; Colford et al., 2005). The municipal water system used a single source (the Mississippi River) and was treated at a single plant with conventional treatment consisting of coagulation, flocculation, sedimentation, prechlorination, filtration (dual filters with granular activated carbon and sand), and post-filtration chloramination. The average turbidity of the finished water was 0.05 NTU.

A total of 456 households with 1,296 participants were randomized into two groups. One group received a household water treatment device with a 1-micron absolute ceramic filter and UV light with 35,000–38,000 uW-second/cm2 output. The other group received a sham device that was identical to the active device but had an empty filter chamber and a UV light that was shielded to block the transmission of radiation but still generated the same light and heat as the active unit. Each study household had an active device for six

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

months and a sham device for six months and was blinded to the status of their device during the study. Study participants recorded the occurrence of any symptoms in daily health diaries. HCGI was defined as in the previous studies as (1) vomiting, (2) watery diarrhea, (3) soft diarrhea and abdominal cramps, or (4) nausea and abdominal cramps.

Incidence of HCGI varied by season and ranged during the study period from 1.64 to 2.80/person/years at risk (Wade et al., 2004). The overall HCGI rate for households with the sham device was 2.12 episodes/person/year and 2.20 episodes/person/year for households with the active device. The overall HCGI rate for the entire study population was 2.16 episodes/person/year. Multivariate analyses showed no effect of the household water treatment device on illness rates during the 12-month study period. As in the studies by Payment et al., the highest illness rates were in children five years of age and younger. The overall conclusion was that less than 11 percent of the gastrointestinal illness observed in this community was due to drinking water. Unlike the studies by Payment et al., this study included households without children, and it is possible that the number of young children in the study was too small to be able to detect an effect in this more vulnerable group.

United Kingdom Study

A study conducted in Wales and northwest England from 2001 to 2002 found a very strong association (p < 0.001) between self-reported diarrhea and reported low water pressure at the home tap based on a postal survey of 423 subjects (Hunter et al., 2005). This study was part of a larger case-control study of risk factors associated with sporadic cryptosporidiosis and was not specifically designed to study waterborne disease. Cryptosporidiosis cases and controls were identified from family physician practices in Wales and northwest England, and a postal survey asking a number of questions about potential risk factors for diarrhea was mailed to 662 cases of cryptosporidiosis and 820 controls. The survey included questions on travel outside the U.K., eating habits, food preparation habits, contact with animals, contact with young children, consumption of unboiled water, contact with other persons with diarrhea, and age. Questionnaires were returned by 427 controls, and 423 were included in the analyses. Of these, 28 (6.6 percent) reported having diarrhea in the two weeks before receiving the survey.

Four risk factors for diarrhea in the control group remained significant in the logistic regression model using a stepwise comparison strategy: feeding a child under five years old, contact with another person who had diarrhea, loss of water pressure at home, and how often the subject ate yogurt. The first three risk factors had a positive association with diarrhea (Odds Ratios of 2.5, 7.0, and 12.5, respectively, after adjusting for the effects of the other variables in the model). Yogurt consumption had a protective effect against diarrhea and showed a dose-response relationship (more frequent consumption was associ-

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

ated with lower risk). The investigators suggested that the strength of the association between loss of water pressure and risk of diarrhea indicates that this was not a spurious association and was not likely to be affected by recall bias because it was just one of many potential risk factors that was investigated.

The study populations were drawn from two large regions that include both heavily industrialized areas and rural areas and about 240 water treatment plants. The overall microbiological water quality for the utilities in these regions was described to be excellent with less than 0.05 percent of water samples positive for E. coli during this study period. The investigators hypothesized that most of the reported episodes of pressure loss were due to main breaks in which contamination entered the distribution system. However, no attempt was made to collect information on recorded main breaks in the systems where the controls lived. The investigators concluded that up to 15 percent of gastrointestinal illness may be associated with consumption of drinking water that was contaminated from main breaks or other pressure loss events, and that the associated costs of this illness should be taken into account when weighing the costs of replacing aging water supply distribution systems. Although there had previously been concern about possible health risks from pressure loss and pathogen intrusion in water distribution systems (LeChevallier et al., 2003), this was the first study to provide solid evidence of that risk, with policy implications for how to manage low pressure events in public water supplies.

***

The body of evidence from these epidemiological studies does not eliminate consumption of tap water that has been in the distribution system from causing increased risk of gastrointestinal illness. The conflicting results between the Laval and U.K. studies, which indicated risk associated with distribution system water, versus the Melbourne and Davenport studies, which showed no increased risk of gastrointestinal illness associated with tap water, may be due to a number of differences between the study designs and the individual water systems.

With respect to the latter, all four cohort studies were in cities that used surface water supplies. In Laval and Davenport, the rivers received upstream sewage discharges and were known to be contaminated. With the Davenport study in particular, it is possible that the reason they found no contribution to disease from the water supply was because the investigators chose a well-operated and maintained system. In Melbourne, the source water came from a highly protected watershed. In Laval and Davenport, the water treatment plants used conventional filtration and disinfection—indeed, Laval had both ozonation and chlorination although the average turbidity of the finished water during the first study was quite high (0.26 NTU). The water treatment plant in Melbourne did not practice filtration. There is no information on the water supplies in the U.K. study. Little to no information on the distribution systems was provided in the descriptions of the Laval or Melbourne studies except that the residence time in

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

the Laval system was relatively short (0.3 to 34 hours), while the residence time for most of the study families in the Melbourne study was 72 to 96 hours.

Differences in study design such as population size and composition and follow-up period also played a role. As shown in Table 3-5, the size of the study population in the Davenport study is approximately half of the study population in the Laval and Melbourne studies (although the Davenport study used a crossover design to try to compensate for the smaller sample size). The Davenport study also had the shortest follow-up period of the four studies. Unlike the Laval and Melbourne studies that only recruited households with children, households enrolled in the Davenport study were not required to have children, and the average household size was smaller in the Davenport study (2.84 persons) compared to the Laval and Melbourne studies (Laval 1988–1989: 3.97 persons; Laval 1993–1994: 3.84 persons; Melbourne: 4.69 persons). The smaller sample size, shorter follow-up period, and possibly lower proportion of children (a vulnerable sub-population), may be reasons why the Davenport study did not detect a significant risk of waterborne illness.

TABLE 3-5 Comparison of Population Parameters from the Epidemiology Studies

Study

Laval

1988-1989

Laval

1993-1994

Melbourne

1997-1999

Davenport

2000-2002

# households in tapwater group

307

346 (tap water)

330 (tap w/valve)

300

229

# of persons in tapwater group

1,202

1,296 (tap water)

1,300 (tap w/valve)

1,399

650

# households in purified water group

299

339 (purified)

354 (bottled plant)

300

227

# of people in purified water group

1,206

1,360 (purified)

1,297 (bottled plant)

1,412

646

% children in tapwater group

6.2 <6 yrs

12.8 <6 yrs (tap)

16.5 <6 yrs (tap valve)

40.2 < 10 yrs

NA

% children in purified water group

9.6 <6 yrs

15.1 <6 yrs (purified)

15.4 <6 yrs (bottled plant)

40.9 < 10 yrs

NA

Weeks of observation time

Approx 60

Approx 69

68

54

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Statistical power in a cohort study is determined by the size of the study population, the follow-up time, and the frequency of the health outcome of interest (incidence of HCGI), with the number of outcomes being more relevant that the size of the study population (Hulley and Cummings, 1988). The Davenport study was designed to have the statistical power to detect an 11 percent or greater risk of HCGI due to water (Colford et al., 2005). The Melbourne study, with the larger sample size and longer follow-up period, was designed to detect a 15–20 percent reduction in the overall rate of HCGI in the group with the active point-of-use treatment devices. However, the total number of HCGI episodes measured in both study populations was very similar (tap water: Melbourne = 1,500 episodes, Davenport = 1,431 episodes; purified water: Melbourne = 1,459 episodes, Davenport = 1,476 episodes). Thus, the higher HCGI rates detected in the Davenport study and the cross-over design appear to have mitigated the effects of the smaller sample size and shorter follow-up period on the statistical power of the study. As shown in Table 3-5, all of these studies had relatively large study populations and measured thousands of illness episodes, and thus had similar statistical power.

There was limited assessment of exposure among the studies. All of the studies monitored water quality at the treatment plant, but there was a wide range in the amount of sampling and analyses of water in the distribution system. For example, monitoring in the Davenport study was extensive, with tap water samples and treatment device samples collected from about one-fourth of the study households at three times during the study. They documented higher coliform and HPC levels in water from the treatment devices compared to tap water (LeChevallier et al., 2002). None of the studies reported pathogen detection in the tap water, except for three occasions in the Melbourne study. It should be noted that the microbiological analyses of water differed for each study. Finally, all four studies attempted to measure the volume of tap water ingested via surveys, and these surveys indicated that subjects in the purified water groups also consumed regular tap water (reported range 14.5 to 40 percent).

All four cohort studies used similar approaches for recording symptoms of gastrointestinal illness and similar definitions of HCGI. Different rates of HCGI were observed in the four cohort studies. It is striking that the rates reported by the Davenport study and the Contra Costa County pilot study are more than twice as high as the rates reported by the Laval and Melbourne studies and about three times higher than the FoodNet rates of diarrheal disease (see Table 3-6). The reason for these higher rates is unknown because the investigators state that they used similar case definitions as the Laval and Melbourne studies. If there were several significant transmission routes of enteric pathogens in these communities that were responsible for these higher reported illness rates, then an intervention study targeted only to waterborne disease transmission may not show any effect (see Briscoe, 1984). However, the use of the cross-over design in Davenport should have been valuable in this regard because the effect of other transmission routes is better controlled for using this design.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

TABLE 3-6 Rates of Highly Credible Gastrointestinal Illness from the Epidemiology Studies

Study

Estimated rate* of HCGI in tap water groups

Estimated rate* of HCGI in purified water group

Estimated rate* of HCGI in all study participants

Laval

1988-1989

0.76

0.50

0.66

Laval

1993-1994

0.66 (tap)

0.70 (tap valve)

0.58

0.60

Melbourne

1997-1999

0.82

0.79

0.80

Contra Costa County, CA

1999

3.48

2.63

3.05

Davenport

2000-2002

2.12

2.20

2.16

FoodNet

ND

ND

Approx 0.72

* rate expressed as episodes/person/year

The conflicting results of these epidemiological studies raise a number of questions. The fact that these were carefully conducted studies by research teams with considerable experience implies that there are detectable elevated risks of waterborne disease associated with some water systems and not others. However, not enough information was gathered to know what characteristics of the water systems posed increased risk, whether it be the source water, the treatment plant, or the distribution system.

For the studies that showed no detectable association between gastrointestinal symptoms and consumption of tap water (Melbourne and Davenport), it is not clear if they suffered from an inadequate design and sample size in order to detect an association, or if there simply was no association. The randomized clinical trial design used in Laval, Melbourne, and Davenport is one of the most rigorous analytical study designs and is less likely to be affected by error and confounding. However, it is possible that selection bias in the recruitment of the study population, misclassification of drinking water exposure, or inaccurate reporting of health outcomes may have affected the results of these studies. It must be kept in mind that epidemiological studies are not able to prove that there is zero risk associated with a specific exposure; they can only report that the risk is below the level that the study had the power to detect, which was 15 to 20 percent (Melbourne) or 11 percent (Davenport).

For the studies that did show an association between gastrointestinal symptoms and consumption of tap water (Laval study), or an association between gastrointestinal symptoms and a water pressure drop (UK study), it is not clear what portion of the observed risk was due to water contamination in the distribution system as opposed to water contamination at the source and/or inadequate

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

water treatment. The second Laval study examined the risks associated with the distribution system by including a study group that received bottled plant effluent as well as groups that ingested tap water and continuous-flow tap water (“tap valve” group). Tap water drinkers had elevated risk of HCGI compared to those who ingested bottled water from the treatment plant or purified bottled water, suggesting that water in the distribution system posed an increased health risk (although routine water quality monitoring of the distribution system did not provide evidence of compromised quality). However, there was also an indication of some increased risk of illness from water with reduced residence time in the distribution system (tap valve group) compared to water with average residence times (from 0.3 to 34 hours in this system). This suggests that additional contact time with disfinectants in the distribution system may be helpful in reducing risks. The UK study suggests that pressure drops in the distribution system was associated with increased gastrointestinal illness, but this association needs to be tested more systematically and rigorously in further studies.

One of the major challenges for designing an epidemiology study of health risks associated with water quality in the distribution system is separating the effect of source water quality and treatment from the effect of distribution system water quality. Knowledge of how water distribution systems become contaminated from anecdotal evidence and outbreak data (main breaks, sudden changes in pressure and intrusion, backpressure or backsiphonage, etc.) suggests that the exposure to contamination in the distribution system is likely to be intermittent and may be very difficult to capture in an epidemiological study. Nonetheless, new approaches to deal with this challenge were tested in a pilot study in the southeastern U.S. and a third approach is being tested in a study in the Midwestern U.S. These studies were designed by multidisciplinary teams of university and research foundation scientists with input from outside experts including EPA and CDC staff. Support for these studies came from the EPA STAR Grant Program, and they are part of a series of studies funded by or conducted by the EPA to develop a national estimate of waterborne disease risks. These three approaches are described in Box 3-3 as examples. Other study designs may also be useful for addressing the question of endemic disease risks associated with water quality in the distribution system.

RISKS FROM LEGIONELLA

The role of biofilms and microbial risk can best be illustrated by the example of the bacterium Legionella pneumophila in water systems, for which occurrence data, outbreak data, and epidemiological data are available. Legionella are widely distributed in the aqueous environment and have been found in drinking water (Stout et al., 1985; Rogers et al., 1994) and biofilms (Rogers et al., 1994; Pryor et al., 2004; Thomas et al., 2006). Although the bacteria have been isolated from biofilms in water distribution systems, there is evidence that the

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

BOX 3-3

Three Approaches to Designing an Ideal Epidemiology Study that would Determine the Distribution System Component to Waterborne Disease

Method 1

This method relies on conducting a vulnerability assessment of the water distribution system and identifying areas in the distribution system that are more vulnerable and less vulnerable to contamination—based on pipe age and composition, history of main breaks, history of coliform detections, estimates of residence time, and chlorine residual. The study population (families with one or more children < six years old) should be recruited in the most vulnerable and the least vulnerable geographic areas of the distribution system. It is important to randomize the study population in each geographic area into two groups. The researchers would provide purified bottled water to half of the study households, and ask the other half of the study population to drink tap water. All study households would be asked to record health symptoms in a health diary. The difference in the rates of reported gastrointestinal symptoms (GI) for families drinking tap water to the rates for families drinking purified bottled water would then be compared. This difference (GItap-GIbottle) represents the risk of GI symptoms due to source water and distribution system water. Part of the analysis would be to compare this difference (GItap-GIbottle) for the study populations in the most vulnerable areas (where the degradation of distribution system water quality would be the greatest) to the difference (GItap-GIbottle) for the study populations in the least vulnerable areas (where there should be little or no impact from degradation of water quality in the distribution system). This difference between the study groups in different parts of the distribution system should represent the impact of the distribution system on risk of GI illness (see Figure 3-5). Although the study is not blinded, the technique of “comparing the difference of the difference” controls for lack of blinding. This “double-difference methodology” is commonly used in economics studies and program evaluation to assess the impact of a specific intervention by comparing the differences between intervention and control groups at baseline and at a follow-up time point (Maluccio and Flores, 2005).

FIGURE 3-5 Study Design to Examine Risks from Water Quality in the Distribution System: Method 1.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Method 2

This approach is identical to the first, except that the study population in each geographic area is randomized into three groups. The researchers would provide purified bottled water to one-third of the study households, bottled finished water directly from the treatment plant to the other third of the study households, and bottled water from the most vulnerable part of the distribution system to the final third of the study population. As before, study households would be asked to record health symptoms in a health diary. This study, which has a cross-over design, is shown in Figure 3-6. The advantage of this approach over the first approach is that the study is blinded because everyone receives bottled water. Furthermore, one can recruit study subjects in any geographic location because drinking water is delivered to their home. This design is similar to a human challenge study because the investigators control exposure to the study water. The disadvantages are that bottled distribution system water will not capture temporal changes in water quality. Also, possible changes in water quality during bottling and storage may not reflect quality of distribution system water. However, these disadvantages could be mitigated by detailed microbiological studies of distribution system water quality in the study site prior to starting the epidemiologic study, bottling the distribution system water more frequently, bottling composite samples of the distribution system water over time and geographic area, and characterizing changes in distribution system water quality during bottling and storage.

FIGURE 3-6 Study Design to Examine Risks from Water Quality in the Distribution System: Method 2 Cross-over Study.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Method 3

A third approach is being attempted in the Wisconsin groundwater study (WAHTER) in several communities that use untreated groundwater. This study uses a community level intervention where UV disinfection is added at the wellhead, and community gastrointestinal symptom rates are compared before and after the UV intervention. The risk from the distribution system will be estimated using a risk assessment approach. Enteric virus concentrations are being measured in water samples from well heads (representing contamination in the groundwater) and compared to virus concentration measurements in water samples from study households (representing contamination from both the groundwater and the distribution system). The difference in virus concentration will be attributed to the distribution system. In those study communities with UV disinfection installed at the wellheads, viruses measured at the households could only have originated from intrusions into the distribution system. Note that the feasibility of this approach depends on studying a water supply where pathogens are detected with some frequency. For a water supply where a high proportion of water samples do not have detectable pathogens, the application of this study design is uncertain.

The study also measures the incidence of gastrointestinal symptoms in a cohort of children in the study communities using a health diary. The researchers intend to model the illness rate in the study population as a function of household pathogen concentration using dose-response models where incidence of acute gastrointestinal illness in the study population is a function of the pathogen dose in the household water (calculated as concentration of virus in the volume of water ingested over a defined period of time). The investigators will then use quantitative risk assessment to estimate the community illness rates if the population drank water directly from the wellhead. The difference between the measured illness rates in the study population and the estimated illness rates associated with source water will represent the risk from pathogens in distribution system.

One of the challenges of this approach is that there are different dose-response relationships for different waterborne viruses. Thus, information on the etiology of the predominant viral infections in the community will be used to guide the modeling analyses.

SOURCE: Available online at http://cfpub.epa.gov/ncer_abstracts/index.cfm/fuseaction/display.abstractDetail/abstract/7430/report/0. Accessed August 10, 2006.

organism must be taken up by protozoa to proliferate (Nahapetian et al., 1991; Barbaree et al., 1986; Barbaree, 1991; Murga et al., 2001). Some studies have reported that the presence of amoebae is a predictor of Legionella colonization in plumbing systems (Moore et al., 2006).

Levels of legionellae in potable water systems are typically low, but amplification can occur in cooling towers, recirculating hot water systems, and hot tubs (EPA, 1999). Legionella species have been shown to proliferate in biofilms in institutional and premise plumbing (Pryor et al., 2004; Thomas et al., 2006) and can be found in water heaters, shower heads, and cooling towers (Wadowsky and Yee, 1983, 1985; Stout et al., 1985; Rogers et al., 1994). Indeed, in a study of legionellosis in the United Kingdom, 528 of the examined 604 cases were attributed to contaminated cooling towers, 70 (or 12 percent) were caused by contaminated drinking water, and six were caused by contaminated whirlpools (VROM, 2005).

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Legionella is an example of an organism that is an efficient pulmonary pathogen when inhaled as large aggregates or biofilm fragments. Inhalation of large numbers of the bacteria overwhelms the pulmonary defenses, and Pontiac fever results. Aspiration of smaller numbers of organisms as biofilm fragments may cause Legionnaire’s disease. Epidemiological studies have linked water contaminated with both Legionella and protozoa to outbreaks of legionellosis (Fields et al., 1989; Breiman et al., 1990). A review paper by Lin et al. (1998) suggests that hospitals take routine samples for the organism in their distribution systems and determine the efficacy of any disinfection processes by measuring a reduction in Legionella counts.

Legionella are specifically mentioned in the EPA’s Surface Water Treatment Rule, with the MCLG set at zero. For this reason, the bacterium was not included on the Contaminant Candidate List for methods development and potential future regulation. However, there is little evidence that filtration and disinfection of surface water prevents the growth of Legionella species in distribution system plumbing. In fact, since Legionella was incorporated into the waterborne disease outbreak surveillance system starting in 2001, several outbreaks have been attributed to the microorganism. During 2001–2002, the six drinking water outbreaks attributed to Legionella species (19.4 percent of the total) caused illness in 80 persons and resulted in 41 hospitalizations and four deaths. All of these outbreaks occurred in large buildings or institutional settings and were related to multiplication of Legionella species in the respective distribution systems. As mentioned previously, Legionella is now the single most common cause of outbreaks involving drinking water (Liang et al., 2006). These outbreaks underscore the importance of remaining vigilant about the possibility of growth of Legionella species in building complexes and the need to take measures to reduce this threat (see Chapter 8).

In an epidemiological study, Kool and colleagues (1999) examined 32 nosocomial outbreaks of Legionnaires’ disease from 1979 to 1997 where drinking-water was implicated and tabulated the characteristics of the hospital (size, transplant program) and the primary disinfectant treatment, disinfectant residual, water source, community size, and pH of the water. The researchers found that the odds of a nosocomial Legionella outbreak was 10.2 (95 percent confidence interval of 1.4–460) times higher in systems that maintained free chlorine than in those using a chloramine residual. They estimated that 90 percent of waterborne Legionella outbreaks could be prevented if chloramine was universally used. Heffelfinger et al. (2003) reported that 25 percent (38) of 152 hospitals surveyed had definite reported cases or outbreaks of hospital-acquired Legionnaires’ disease during the period 1989 to 1998. However, hospitals supplied with drinking water disinfected with monochloramine were less likely (odds ratio 0.20; 95 percent confidence interval, 0.07 to 0.56) to have hospital-acquired Legionnaires’ disease than other hospitals. Cunliffe (1990) reported that suspensions of Legionella pneumophila were more sensitive to monochloramine disinfection, with a 99 percent level of inactivation when exposed to 1.0 mg monochloramine/L for 15 minutes, compared with the 37-

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

minute contact time required for Escherichia coli inactivation under similar conditions. Donlan et al. (2002) reported that monochloramine was significantly more effective than free chlorine at eradicating laboratory-grown biofilms of L. pneumophila.

Legionella has also been the subject of pathogen occurrence measurements. Researchers at the CDC conducted a study of Legionella occurrence in 53 public buildings before and after the conversion of the San Francisco water supply from free chlorine to chloramine (Fields, 2005; Flannery et al., 2006). They showed that the concentration of legionellae was reduced more than 20-fold by the conversion from free chlorine to chloramine. Interestingly, the incidence rate of Legionella infections was low (only one laboratory-confirmed case in the two years prior to the switch to chloramine) despite the fact that the major serotype detected included the clinically significant Legionella pneumophila serogroup 1. The results illustrate the difficulty in relating the detection of microbes in drinking water to a documented risk of waterborne disease.

Another recent study examined the impact of switching from chlorine to monochloramine disinfection on Legionella occurrence in Pinellas County, Florida (Moore et al., 2006). In this study, water samples were collected from 96 buildings (public buildings and individual homes) for a four-month period when chlorine was the primary disinfectant and from the same sampling sites for a four-month period after monochloramine was introduced into the municipal water system. In the first period, 20 percent of the buildings were colonized with Legionella in at least one sampling site. Legionella colonization was reduced by 69 percent within a month after monochloramine introduction. Monochloramine appeared to be more effective in reducing Legionella in hotels and single-family homes than in county government buildings, perhaps because of more consistent water usage. As in the San Francisco study, the reported incidence of legionellosis in the study area during this time was too low (nine cases) to determine if the change to monochloramine had an impact on human disease.

Given that 20 percent of reported outbreaks involving drinking water are attributed to Legionella, additional attention should be given to the control of this potential pathogen, especially in institutional and premise plumbing (see Chapter 8).

CONCLUSIONS AND RECOMMENDATIONS

Accurate estimates are not yet available for the prevalence of adverse health effects attributable to deficiencies in distribution systems from pathogen occurrence measurements, waterborne disease outbreak surveillance, or epidemiological studies. Pathogen occurrence measurements are rare due to limitations in detection methods and cost issues. Models to quantitatively predict pathogen occurrence in distribution systems (e.g., by cross-connections, main breaks, or intrusion) have not yet been developed. Despite under-reporting and limited data on risk factors, the voluntary waterborne disease outbreak surveillance sys-

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

tem provides the best available evidence of public health risks associated with distribution systems in the United States. These data suggest that about one-third to one-half of reported waterborne disease outbreaks are associated with distribution system problems. To date, only one epidemiological study (the second Laval study) has been specifically designed to examine the contribution of the distribution system to endemic disease occurrence. Until better data are available from these three approaches, it will not be possible to accurately assess the magnitude of the health impacts resulting from distribution system deficiencies. The following conclusions and recommendations are made.


The distribution system is the remaining component of public water supplies yet to be adequately addressed in national efforts to eradicate waterborne disease. This is evident from data indicating that although the number of waterborne disease outbreaks including those attributable to distribution systems is decreasing, the proportion of outbreaks attributable to distribution systems is increasing. Most of the reported outbreaks associated with distribution systems have involved contamination from cross-connections and backsiphonage. Furthermore, Legionella appears to be a continuing risk and is the single most common etiologic agent associated with outbreaks involving drinking water. Initial studies suggest that the use of chloramine as a residual disinfectant may reduce the occurrence of Legionella, but additional research is necessary to determine the relationship between disinfectant usage and the risks of Legionella and other pathogenic microorganisms.


Distribution system ecology is poorly understood. There is very little information available about the types, activities, and distribution of microorganisms in distribution systems. Limited HPC data are available for some systems, but these data are not routinely collected, they underestimate the numbers of organisms present, and they include many organisms that do not necessarily present a health risk. To more adequately assess risk, more information on the microbial ecology of distribution systems, including premise plumbing, is needed.


There is inadequate investigation of waterborne disease outbreaks associated with distribution systems, especially in premise plumbing. Legionella has only recently been added to the outbreak surveillance system. Existing data on outbreaks due to other etiologic agents would rarely implicate premise plumbing because backflow and regrowth events likely would not be recognized and reported unless an institutional building with large numbers of people was affected. The Centers for Disease Control and Prevention are commended for revising the format used to report waterborne disease outbreaks to the surveillance system such that outbreaks arising from events in premise plumbing are now more clearly identified.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Epidemiology studies that specifically target the distribution system component of waterborne disease are needed. Recently completed epidemiological studies have either not focused on the specific contribution of distribution system contamination to gastrointestinal illness, or they have been unable to detect any link between illness and drinking water. Epidemiological studies of the risk of endemic disease associated with drinking water distribution systems need to be performed and must be designed with sufficient power and resources to adequately address the deficiencies of previous studies.


This chapter highlights the lack of information available to assess the public health risk of contaminated distribution systems. One of the consequences of this fact is that the committee was forced to rely heavily on its best professional judgment to prioritize contamination events into high, medium, and low priority (see Appendix A). Better public health data, including data on waterborne outbreaks, from epidemiological studies, and on distribution system water quality, could help refine distribution system risks and provide additional justification for the rankings.

The following three chapters consider the roles of physical, hydraulic, and water quality integrity. Protection of public health requires that water professionals incorporate approaches that combine all three into a comprehensive program of best practices to maintain the highest level of water quality.

REFERENCES

Abbaszadegan, M., M. LeChevallier, and C. Gerba. 2003. Occurrence of viruses in U.S. groundwaters. J Amer. Water Works Assoc. 95(9):107–120.

Alavandi, S. V., M. S. Subashini, and S. Ananthan. 1999. Occurrence of haemolytic and cytotoxic Aeromonas species in domestic water supplies in Chennai. Indian J. Med. Res. 110:50–55.

Amann, R. I., B. J. Binder, R. J. Olson, S. W. Chisholm, R. Devereux, and D. A. Stahl. 1990. Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Appl. Environ. Microbiol. 56:1919–1925.

Amann, R. I., W. Ludwig, and K.-H. Schleifer. 1995. Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol. Rev. 59:143–169.

American Water Works Association (AWWA). 1999. Waterborne pathogens: manual of water supply practices—M48, First Edition. Denver, CO: AWWA.

Anand, S., and K. Hanson. 1997. Disability-adjusted life years: a critical review. J. Health Econ. 16(6):685–702.

Arbuckle, T. E., S. E. Hrudey, S. W. Krasner, J. R. Nuckols, S. D. Richardson, P. Singer, P. Mendola, L. Dodds, C. Weisel, D. L. Ashley, K. L. Froese, R. A. Pegram, I. R. Schultz, J. Reif, A. M. Bachand, F. M. Benoit, M. Lynberg, C. Poole, and K. Waller. 2002. Assessing exposure in epidemiologic studies to disinfection by-products in drinking water: Report from an international workshop. Environ. Health Perspect. 110 (Suppl 1):53–60.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Balbus, J., R. Parkin, and M. Embrey. 2000. Susceptibility in microbial risk assessment: definitions and research needs. Environmental Health Perspectives 108(9):901–905.

Barbaree, J. M. 1991. Controlling Legionella in cooling towers. ASHRAE Journal 33:38–42.

Barbaree, J. M., B. S. Fields, J. C. Feeley, G. W. Forman, and W. T. Martin. 1986. Isolation of protozoa from water associated with a legionellosis outbreak and demonstration of intracellular multiplication of Legionella pneumophila. Appl. Environ. Microbiol. 51:422–424.

Bartram, J., J. Cotruvo, M. Exner, C. Fricker, and A. Glasmacher (eds). 2003. Heterotrophic Plate Counts and Drinking-water Safety. London: IWA Publishing.

Barwick, R. S., D. A. Levy, G. F. Craun, M. J. Beach, and R. L. Calderon. 2000. Surveillance for waterborne-disease outbreaks—United States, 1997–1998. MMWR 49(SS-4):1–34.

Bays, L. R., N. P. Burman, and W. M. Lavis. 1970. Taste and odour in water supplies in Great Britain: a survey of the present position and problems for the future. Journal Society Water Treatment Examination 19:136–160.

Bellen, G. E., S. H. Abrishami, P. M. Colucci, and C. Tremel. 1993. Methods for assessing the biological growth support potential of water contact materials. Denver, CO: AwwaRF.

Bennett, J. V., S. D. Holmberg, M. F. Rogers, and S. L. Soloman. 1987. Infectious and parasitic diseases. In: Closing the Gap: The Burden of Unnecessary Illness. R. W. Amler and H. B. Dull (eds.). New York: Oxford University Press.

Blackburn, B. G., G. F. Craun, J. S. Yoder, V. Hill, R. L. Calderon, N. Chen, S. H. Lee, D. A. Levy, and M. J. Beach. 2004. Surveillance for waterborne-disease outbreaks associated with drinking water—United States, 2001–2002. MMWR 53(SS-8):23– 45.

Borchardt, M. A., N. L. Haas, and R. J. Hunt. 2004. Vulnerability of drinking-water wells in La Crosse, Wisconsin, to enteric-virus contamination from surface water contributions. Appl Environ Microbiol. 70(10):5937–5946.

Boorman, G. A., V. Dellarco, J. K. Dunnick, R. E. Chapin, S. Hunter, F. Hauchman, H. Gardner, M. Cox, and R.E. Sills. 1999. Drinking water disinfection by-products: review and approach to toxicity evaluation. Environ. Health Perspect. 107:207–17.

Breiman, R. F., B. S. Fields, G. N. Sauden, L. Volmer, A. Meier, and J. S. Spilka. 1990. Association of shower use with Legionnaires' disease. J. Amer. Med. Assoc. 263:2924–2926.

Briganti, L. A., and S. C. Wacker. 1995. Fatty acid profiling and the identification of environmental bacteria for drinking water utilities. Denver, CO:AWWA.

Briscoe, J. 1984. Intervention studies and the definition of dominant transmission routes. Am. J. Epidemiol. 120:449–55.

Camper, A. K. 1996. Factors Influencing Microbial Growth in the Distribution System: Laboratory and Pilot Experiments. Denver, CO: AwwaRF.

Cantor, K. P., C. F. Lynch, M. E. Hildesheim, M. Dosemeci, J. Lubin, M. Alavanja, and G. Craun. 1998. Drinking water source and chlorination byproducts—I. Risk of bladder cancer. Epidemiology 9:21–28.

Carnahan, A. M., and M. Altwegg. 1996. Taxonomy. In: The genus Aeromonas. S. Joseph (ed.). New York: John Wiley & Sons.

Carson, L. A., L. A. Bland, L. B. Cusick, M. S. Favero, G. A. Bolan, A. L. Reingold, and R. A. Good. 1988. Prevalence of nontuberculous mycobacteria in water supplies of hemodialysis centers. Appl. Environ. Microbiol. 54:31223125.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Chaidez, C., and C. Gerba. 2004. Comparison of the microbiological quality of point-of-use (POU)-treated water and tap water. Int. J. Environ. Health Res. 14:253–261.

Characklis, W. G. 1988. Bacterial Regrowth in Distribution Systems. Denver, CO: AWWA and AwwaRF.

Characklis, W. G. and K. C. Marshal. 1990. Biofilms. New York: John Wiley & Sons, Inc.

Chauret, C., C. Volk, R. Creason, J. Jarosh, J. Robinson, and C. Warnes. 2001. Detection of Aeromonas hydrophila in a drinking-water distribution system: a field and pilot study. Can. J. Microbiol. 47:782–786.

Clark, R. M., E. E. Geldreich, K. R. Fox, E. W. Rice, C. H. Johnson, J. A. Goodrich, J. A. Barnick, and F. Abdesaken. 1996. Tracking a Salmonella serovar typhimurium outbreak in Gideon, Missouri: role of contaminant propagation modeling. Journal of Water Supply Research and Technology - Aqua. 45(4):171–183.

Colford, J. M., Jr., T. J. Wade, S. K. Sandhul, C. C. Wright, S. Lee, S. Shaw, K. Fox, S. Burns, A. Benker, M. A. Brookhart, M. van der Laan, and D. A. Levy. 2005. A randomized, controlled trial of in-home drinking water intervention to reduce gastrointestinal Illness. Am. J. Epidemiol. 161(5):472–482.

Colford, J. M., J. R. Rees, T. J. Wade, A. Khalakdina, J. F. Hilton, I. J. Ergas, S. Burns, A, Benker, C. Ma, C. Bowen, D. C. Mills, D. J. Vugia, D. D. Juranek, and D. A. Levy. 2002. Participant blinding and gastrointestinal illness in a randomized, control trial of an in-home drinking water intervention. Emerging Infectious Diseases 8(1):29–36.

Colbourne, J. S., D. J. Pratt, M. G. Smith, S. P. Fisher-Hoch, and D. Harper. 1984. Water fittings as sources of Legionella pneumophila in hospital plumbing system. Lancet 1(8370):210–213

Collins, C. H., J. M. Grange, and M. D.Yates. 1984. Mycobacteria in water. J. Appl. Bacteriol. 57:193–211.

Covert, T. C., M. R. Rodgers, A. L. Reyes, and G. N. Stelma. 1999. Occurrence of non-tuberculous mycobacteria in environmental samples. Appl. Environ. Microbiol. 65(6):2492–2496.

Craun, G. F., and R. L. Calderon. 2001. Waterborne disease outbreaks caused by distribution system deficiencies. J. Amer. Water Works Assoc. 93(9):64–75.

Craun, G. F., ed. 1986. Waterborne Diseases in the United States. Boca Raton, FL: CRC Press, Inc.

Cunliffe, D. A. 1990. Inactivation of Legionella pneumophila by monochloramine. J. Appl. Bacteriol. 68(5):453–459.

De Victoria, J., and M. Galvan. 2001. Pseudomonas aeruginosa as an indicator of health risk in water for human consumption. Water Sci. Technol. 43:49–52.

Dodds, L., W. King, C. Woolcott, and J. Pole. 1999. Trihalomethanes in public water supplies and adverse birth outcomes. Epidemiology 10:233–237.

Donlan, R., R. Murga, J. Carpenter, E. Brown, R. Besser, and B. Fields. 2002. Monochloramine disinfection of biofilm-associated Legionella pneumophila in a potable water model system. Pp. 406–410 In: Legionella. R. Marre, Y. Abu Kwaik, C. Bartlett, N. P. Cianciotto, B. S. Fields, M. Frosch, J. Hacker, and P. C. Luck (eds.). Washington, DC: American Society for Microbiology.

Dott, W., and D. Waschko-Dransmann. 1981. Occurrence and significance of actinomycetes in drinking water. Zbl. Bakt. Hyg. I., Abt. Orig. B, 173:217–232.

Doyle, T. J., W. Zheng, J. R. Cerhan, C.-P. Hong, T. A. Sellers, L. H. Kushi, and A. R. Folsom. 1997. The association of drinking water source and chlorination by-

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

products with cancer incidence among postmenopausal women in Iowa: a prospective cohort study. Am. J. Public Health 87:1168–1176.

duMoulin, G. C., and K. D. Stottmeir. 1986. Waterborne mycobacteria: an increasing threat to health. Am Soc. Microbiol. News 52:525–529.

duMoulin, G. C., K. D. Stottmeier, P. A. Pelletier, T. A. Tsang, and J. Hedley-Whyte. 1988. Concentration of Mycobacterium avium by hospital hot water systems. J. Amer. Medical Assoc. 260:15991601.

Environmental Protection Agency (EPA). 1989. Drinking water; national primary drinking water regulations; total coliforms (including fecal coliforms and E. coli); final rule. Federal Register 54:27544–68.

EPA 1998. Announcement of the drinking water contaminant candidate list; notice. Federal Register 63:10274–87.

EPA. 1999. Legionella: human health criteria document. EPA-822-R-99-001. Washington, DC: EPA Office of Science and Technology.

EPA. 2002. Potential contamination due to cross-connections and backflow and the associated health risks: an issues paper. Available on-line at: http://www.epa.gov/safewater/tcr/pdf/ccrwhite.pdf. Accessed May 11, 2006.

EPA. 2004. Drinking water contaminant candidate list 2; notice. Federal Register 69(64):17406–17415.

Falkinhan, J. O., III, C. D. Norton, and M. W. LeChevallier. 2001. Factors influencing numbers of Mycobacterium avium, Mycobacterium intracellulare, and other mycobacteria in drinking water distribution systems. Appl. Environ. Microbiol. 67(3):1225–1231.

Ferroni, A., L. Nguyen, B. Pron, G. Quesne, M. C. Brusset, and P. Berche. 1998. Outbreak of nosocomial urinary tract infection due to Pseudomonas aeruginosa in a pediatric surgical unit associated with tap-water contamination. J. Hosp. Infect. 39:301–307.

Fewtrell, L. and J. M. Colford, Jr. 2004. Water, Sanitation and Hygiene: Interventions and Diarrhoea - A Systematic Review and Meta-analysis. 2004. Washington, DC: World Bank.

Fields, B. S., G. N. Sanden, J. M. Barbaree, W. E. Morril, R. M. Wadowsky, E. H. White, and J. C. Feeley. 1989. Intracellular multiplication of Legionella pneumophila in amoebae isolated from hospital hot water tanks. Current Microbiol. 18:131–137.

Fields, B. 2005. Legionella in distribution system biofilms. January 13, 2005. Presented to the NCR Committee on Public Water Supply Distribution Systems. Irvine, CA.

Fischeder, R. R., R. Schulze-Robbecke, and A. Weber. 1991. Occurrence of mycobacteria in drinking water samples. Zbl. Hygiene 192:154158.

Flannery, B., L. B. Gelling, D. J. Vugia, J. M. Weintraub, J. J. Salerno, M. J. Conroy, V. A. Stevens, C. E. Rose, M. R. Moore, B. S. Fields, and R. E. Besser. 2006. Reducing Legionella colonization of water systems with monochloramine. Emerg. Infect. Dis. [serial on the Internet]. Available on-line at http://www.cdc.gov/ncidod/EID/vol12no04/05-1101.htm. Accessed April 13, 2006.

Fout, G. S., B. C. Martinson, M. W. N. Moyer, and D. R. Dahling. 2003. A multiplex reverse transcriptase-PCR method for detection of human enteric viruses in groundwater. Appl. Environ. Microbiol. 69: 3158–3164.

Frost, F. J., M. Roberts, T. R. Kunde, G. Craun, K. Tollestrup, L. Harter, and T. Muller. 2005. How clean must our drinking water be: the importance of protective immunity. Journal of Infectious Diseases 191:809–814.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Geldreich, E. E., K. R. Fox, J. A. Goodrich, E. W. Rice, R. M. Clark, and D. L. Swerdlow. 1992. Searching for a water supply connection in the Cabool, Missouri disease outbreak of Escherichia coli 0157:H7. Water Research 26(8):1127–1137.

Geldreich, E. E. 1996. Microbial Quality of Water Supply in Distribution Systems. Boca Raton, FL: CRC Press, Inc.

Geldreich, E. E., and M. W. LeChevallier. 1999. Microbial water quality in distribution systems. Pp. 18.1–18.49 In: Water Quality and Treatment. 5th edition. R. D. Letterman (ed.). New York: McGraw-Hill.

Gerba, C. P., J. B. Rose, and C. N. Haas. 1996. Sensitive populations: who is at the greatest risk? International Journal of Food Microbiology 30(1–2):113–123.

Glover, N. A., N. Holtzman, T. Aronson, S. Froman, O. G. W. Berlin, P. Dominguez, K. A. Kunkel, G. Overturf, G. Stelma, Jr., C. Smith, and M. Yakrus. 1994. The isolation and identification of Mycobacterium avium complex (MAC) recovered from Los Angeles potable water, a possible source of infection in AIDS patients. International J. Environ. Health Res. 4:6372.

Graves, C. G., G. M. Matanoski, and R. G. Tordiff. 2001. Weight of evidence of an association between adverse reproductive and developmental effects and exposure to disinfection byproducts: a critical review. Regul. Toxicol. Pharmacol. 34:103–24.

Haas, C. N., M. A. Meyer, and M. E. Paller. 1983. The ecology of acid-fast organisms in water supply, treatment, and distribution systems. J. Amer. Water Works Assoc. 75:39144.

Haudidier, K., J. L. Paquin, T. Francais, P. Hartemann, G. Grapin, F. Colin, M. J. Jourdain, J. C. Block, J. Cheron, O. Pascal, Y. Levi, and J. Miazga. 1988. Biofilm growth in drinking water networks: a preliminary industrial pilot plant experiment. Water Sci. Technol. 20:109–115.

Havelaar, A. H., A. E. Hollander, P. F. Teunis, E. G. Evers, H. J. Van Kranen, J. F. Versteegh, J. E. Van Koten, and W. Slob. 2000. Balancing the risks and benefits of drinking water disinfection: disability adjusted life-years on the scale. Environmental Health Perspective 108(4):315–321.

Hawkins, M. A., S. M. DeLong, R. Marcus, T. Jones, S. Shallow, S. M. Zansky, K. G. McCombs, A. K. Courtney, C. Medus, B. C. Imhoff, and the EIP FoodNet Working Group. 2002. The Burden of Diarrheal Illness in FoodNet, 2000–2001 (Draft report). Atlanta, GA: CDC.

Heffelfinger, J. D., J. L. Kool, S. K. Fridkin, V. J. Fraser, J. C. Carpenter, J. Hageman, J. Carpenter, and C. G. Whitney. 2003. Risk of hospital-acquired Legionnaires ’ disease in cities using monochloramine versus other water disinfectants. Infection Control and Hospital Epidemiol. 24(8):569–574.

Hellard, M. E., M. I. Sinclair, A. B. Forbes, and C. K. Fairley. 2001. A randomized, blinded, controlled trial investigating the gastrointestinal health effects of drinking water quality. Environ. Health Perspect. 109:773–778.

Herikstad, H., S. Yang, T. J. Van Gilder, D. Vugia, J. Hadler, P. Blake, V. Deneen, B. Shiferaw, F. J. Angulo, and The FoodNet Working Group. 2002. A population-based estimate of the burden of diarrhoeal illness in the United States: FoodNet, 1996–97. Epidemiol. Infect. 129:9–17.

Herwaldt, B. L., G. F. Craun, S. L. Stokes, and D. D. Juranek. 1991. Waterborne-disease outbreaks, 1989–1990. MMWR 40(No. SS-3):1–21.

Horsburgh, C. R. 1991. Mycobacterium avium complex infection in the acquired immunodeficiency syndrome. New England Journal of Medicine 324:13321338.

Hulley, S. B., and S. R. Cummings. 1988. Designing Clinical Research: An Epidemiologic Approach. Baltimore, MD: Williams and Wilkins.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Hunter, P. R., R. M. Chalmers, S. Hughes, and Q. Syed. 2005. Self-reported diarrhea in a control group: a strong association with reporting of low-pressure events in tap water. Clinical Infectious Diseases 40:e32–34.

Imhoff, B., D. Morse, B. Shiferaw, M. Hawkins, D. Vugia, S. Lance-Parker, J. Hadler, C. Medus, M. Kennedy, M. R. Moore, and T. Van Gilder. 2004. Burden of self-reported acute diarrheal illness in FoodNet surveillance areas, 1998–1999. For the Emerging Infections Program (EIP) FoodNet Working Group. Clinical Infectious Diseases 38 Suppl 3:S21926.

Jacquemin, J. L., A. M. Simitzia, and N. Chaneoau. 1981. Free-living amoebae in fresh water: a study of the water supply of the town of Poiters. Bull. Soc. Pathol. Exot. 74:521–534.

Jo, W. K., C. P. Weisel, and P. J. Lioy. 1990. Routes of chloroform exposure and body burden from showering with chlorinated tap water. Risk Analysis 10(4):575–580.

Keinanen, M. M., P. J. Martikainen, and M. H. Kontro. 2004. Microbial community structure and biomass in developing drinking water biofilms. Can. J. Microbiol. 50(3):183–191.

Kelley, J., G. Kinsey, R. Paterson, D. Brayford, R. Pitchers, and H. Rossmoore. 2003. Identification and control of fungi in distribution systems. Denver CO: AWWA and AwwaRF.

King, W. D., and L. D. Marrett. 1996. Case-control study of bladder cancer and chlorination by-products in treated water (Ontario, Canada). Cancer Causes Control 7:596–604.

King, W. D., L. Dodds, and A. C. Allen. 2000. Relation between stillbirth and specific chlorination by-products in public water supplies. Environ. Health Perspect. 108:883–886.

Kirmeyer, G. J., M. Friedman, K. Martel, D. Howie, M. LeChevallier, M. Abbaszadegan, M. Karim, J. Funk, and J. Harbour. 2001. Pathogen intrusion into the distribution system. #90835. Denver, CO: AWWA and AwwaRF.

Klotz, J. B., and L. A. Pyrch. 1999. Neural tube defects and drinking water disinfection by-products. Epidemiology 10:383–390.

Koivusalo, M., E. Pukkala, T. Vartiainen, J. J. K. Jaakkola, and T. Hakulinen. 1997. Drinking water chlorination and cancer – a historical cohort study in Finland. Cancer Causes Control 8:192–200.

Kool, J. L., J. C. Carpenter, and B. S. Fields. 1999. Effect of monochloramine disinfection of municipal drinking water on risk of nosocomial Legionnaires’ disease. Lancet 353(9149):272–277.

Kramer, M. H., and T. E. Ford. 1994. Legionellosis: ecological factors of an environmentally 'new' disease. Zentralbl Hyg Umweltmed. 195(5–6):470–482.

Kramer, M. H., B. L. Herwaldt, G. F. Craun, R. L. Calderon, and D. D. Juranek. 1996. Surveillance for waterborne-disease outbreaks—United States, 1993–1994. MMWR 45(No. SS-1):1–33.

LeChevallier, M. W. 2004. Control, treatment, and disinfection of Mycobacterium avium complex in drinking water. Pp. 143–168 In: Pathogenic Mycobacteria in Water. S. Pedley, J. Bartram, G. Rees, A. Dufour, J. Cotruvo (eds.). ISBN: 1843390590. Geneva: World Health Organization.

LeChevallier, M. W., R. W. Gullick, M. R. Karim, M. Friedman, and J. E. Funk. 2003. The potential for health risks from intrusion of contaminants into distribution systems from pressure transients. J. Water and Health 1(1):3–14.

LeChevallier, M. W., and R. J. Seidler. 1980. Staphylococcus aureus in rural drinking water. Appl. Environ. Microbiol. 39:739–742.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

LeChevallier, M. W., T. J. Wade, S. E. Shaw, D. Levy, R. L. Calderon, J. M. Colford, Jr. 2004. Results of the big WET: an epidemiology study of the microbiological quality of drinking water in Davenport, Iowa. AWWA Water Quality and Technology Conference, San Antonio, Texas, November 14–17, 2004.

Lee, S. H., D. A. Levy, G. F. Craun, M. J. Beach, and R. L. Calderon. 2002a. Surveillance for waterborne disease outbreaks—United States, 1999–2000. MMWR 51(SS-8):1–47.

Lee, S., D. Levy, A. Hightower, B. Imhoff, and the EIP FoodNet Working Group. 2002b. Drinking water exposures and perceptions among 1998–1999 FoodNet survey respondents. International Conference on emerging Infectious Diseases. Atlanta, GA, March 2002.

Levy, D. A., M. S. Bens, G. F. Craun, R. L. Calderon, and B. L. Herwaldt. 1998. Surveillance for waterborne-disease outbreaks—United States, 1995–1996. MMWR 47(SS-5):1–34.

Liang, J. L., E. J. Dziuban, G. F. Craun, V. Hill, M. R. Moore, R. J. Gelting, R. L. Calderon, M. J. Beach, and S. L. Roy. 2006. Surveillance for Waterborne Disease and Outbreaks Associated with Drinking Water and Water not Intended for Drinking—United States, 2003–2004. MMWR (in print).

Lin, Y. E., R. D. Vidic, J. E. Stout, and V. L. Yu. 1998. Legionella in water distribution systems. J. Amer. Water Works Assoc. 90:112–121.

Maluccio, J. A., and R. Flores. 2005. Impact evaluation of a conditional cash transfer program. The Nicaraguan Red de Protección Social. Research Report No. 141. Washington, DC: International Food Policy Research Institute.

Martiny, A. C., T. M. Jorgensen, H. J. Albrechtsen, E. Arvin, and S. Molin. 2003. Long-term succession of structure and diversity of a biofilm formed in a model drinking water distribution system. Appl. Environ. Microbiol. 69(11):6899–6907.

McMillan, M., T. F. Jones, A. Banerjee, D. Vugia, A. Cronquist, S. Segler, P. Ryan, C. Medus, P. Smith, B. Shiferaw, and F. Angulo. 2004. The Burden of Diarrheal Illness in FoodNet, 2002–2003. In: Proceedings of the International Conference on Emerging Infectious Diseases, Atlanta, GA.

Mead P.S., L. Slutsker L, V. Dietz, L. F. McCaig, J. S. Bresee, C. Shapiro, J. M. Griffin, and R. V. Tauxe. 1999. Food-related illness and death in the United States. Emerg. Infect. Dis. 5(5):607–25.

Moore, A. C., B. L. Herwaldt, G. F. Craun, R. L. Calderon, A. K. Highsmith, and D. D. Juranek. 1993. Surveillance for waterborne disease outbreaks—United States, 1991–1992. MMWR 42(SS-5):1–22.

Moore, M. R., M. Pryor, B. Fields, C. Lucas, M. Phelan, and R. E. Besser. 2006. Introduction of monochloramine into a municipal water system: Impact on colonization of buildings by Legionella spp. Appl Environ Microbiol 72(1):378–383.

Morris, R. D., and R. Levine. 1995. Estimating the incidence of waterborne infectious disease related to drinking water in the United States. Assessing and Managing Health Risk from Drinking Water Contamination: Approaches and Application. Proceedings of the Rome Symposium, September 1994. IAHS 233: 75–88.

Murga, R., T. S. Forster, E. Brown, J. M. Pruckler, B. S. Fields, and R. M. donlan. 2001. Role of biofilms in the survival of Legionella pneumophila in a model potable water system. Microbiology 147:3121–3126.

Nagy, L. A., and B. H. Olson. 1982. The occurrence of filamentous fungi in drinking water distribution systems. Canadian Journal Microbiol. 28:667–671.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Nahapetian, K., O. Challemel, D. Beurtin, S. Dubrou, P. Gounon and F. Squinazi. 1991. The intracellular multiplication of Legionella pneumophila in protozoa from hospital plumbing systems. Res. Microbiol. 142:677–685.

National Research Council (NRC). 1983. Risk Assessment in the Federal Government: Managing the Process. Washington, DC: National Academy Press.

NRC. 2004. Review of the Army’s Technical Guides on Assessing and Managing Chemical Hazards to Deployed Personnel. Washington, DC: National Academy Press.

Niemi, R. M., S. Knuth, and K. Lundstrom. 1982. Actinomycetes and fungi in surface waters and in potable water. Appl. Environ. Microbiol. 43:378–388.

Nightingale, S. D., L. T. Byrd, P. M. Southern, J. D. Jockusch, S. X. Cal, and B. A. Wynne. 1992. Mycobacterium avium-intracellulare complex bacteremia in human immunodeficiency virus positive patients. Journal of Infectious Disease 165:10821085.

Nieuwenhuijsen, M. J., M. B. Toledano, N. E. Eaton, J. Fawell, and P. Elliott. 2000. Chlorination disinfection byproducts in water and their association with adverse reproductive outcomes: a review. Occup. Environ. Med. 57:73–85.

Norton, C. D., and M. W. LeChevallier. 2000. A pilot study of bacteriological population changes through potable treatment and distribution. Appl. Environ. Microbiol. 66(1):268–276.

Payment, P., J. Siemiatycki, L. Richardson, G. Renaud, E. Franco, and M. Prevost. 1997. A prospective epidemiological study of gastrointestinal health effects due to the consumption of drinking water. Intern. J. Environ. Health Res. 7:5–31.

Payment, P., L. Richardson, J. Siemiatycki, R. Dewar, M. Edwardes, and E. Franco. 1991. Randomized trial to evaluate the risk of gastrointestinal disease due to consumption of drinking water meeting microbiological standards. Am. J. Pub. Health 81(6):703–708.

Piriou, P., K. Helmi, M. Jousset, N. Castel, E. Guillot, and L. Kiene. 2000. Impact of biofilm on C. parvum persistence in distribution systems. In: Proceedings of an International Distribution System Research Symposium, June 10–11. Denver, CO: AWWA.

Pryor, M., S. Springthorpe, S. Riffard, T. Brooks, Y. Huo, G. Davis, and S. A. Satter. 2004. Investigation of opportunistic pathogens in municipal drinking water under different supply and treatment regimes. Water Sci. Technol. 50(1):83–90.

Quignon, F., M. Sardin, L. Kiene, and L. Schwartzbrod. 1997. Poliovirus-1 inactivation and interaction with biofilm: a pilot-scale study. Appl. Environ. Microbiol. 63(3):978–982.

Raina, P. S., F. L. Pollari, G. F. Teare, M. J. Goss, D. A. Barry, and J. Wilson. 1999. The relationship between E. coli indicator bacteria in well-water and gastrointestinal illnesses in rural families. Can. J. Public Health 90(3):172–175.

Regli, S., J. B. Rose, C. N. Haas, and C. P. Gerba. 1991. Modeling risk from Giardia and viruses in drinking water. J. Amer. Water Works Assoc. 83(11):76–84.

Rogers, J., A. B. Dowsett, P. J. Dennis, J. V. Lee, and C. W. Keevil. 1994. Influence of materials on biofilm formation and growth of Legionella pneumophila in potable water systems. Appl. Environ. Microbiol. 60:1842–1851.

Savitz, D., and C. L. Moe. 1997. Water: chlorinated hydrocarbons and infectious agents. Pp. 89–118 In: Topics in Environmental Epidemiology. N. K. Steenland and D. A. Savitz (eds.). Oxford: Oxford University Press.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

Schoenen, D. 1986. Microbial growth due to materials used in drinking water systems. In: Biotechnology, Vol. 8. H. J. Rehm and G. Reed (eds.). Weinheim: VCH Verlagsgesellschaft.

Schulze-Robbecke, R., and R. Fischeder. 1989. Mycobacteria in biofilms. Zbl. Hyg. 88:385–390.

Schwartz J., R. Levin, and K. Hodge. 1997. Drinking water turbidity and pediatric hospital use for gastrointestinal illness in Philadelphia. Epidemiology 8:615–620.

Shepherd, J. L.; R. L. Corsi, and J. Kemp. 1996. Chloroform in indoor air and wastewater: the role of residential washing machines. Journal of the Air & Waste Management Association 46(7):631–642.

Singh, N., and V. Yu. 1994. Potable water and Mycobacterium avium complex in HIV patients: is prevention possible? Lancet 343:1110–1111.

Smith, D. B., A. F. Hess, and S. A. Hubbs. 1990. Survey of distribution system coliform occurrences in the United States. In: Proceedings of the Water Quality Technology Conference. Denver, CO: AWWA.

Smith, C. A., C. Phiefer, S. J. Macnaughton, A. Peacock, R. S. Burkhalter, R. Kirkegaard, and D. C. White. 2000. Quantitative lipid biomarker detection of unculturable microbes and chlorine exposure in water distribution system biofilms. Water Research 34(10):2683–2688.

Stanwell-Smith, R., Y. Anderson, and D. Levy. 2003. National surveillance systems. Pp. 25–40 In: Drinking Water and Infectious Disease: Establishing the Links. P. R. Hunter, M. Waite, and E. Ronchi (eds.). Boca Raton, FL: CRC Press

Steenland, K., and C. Moe. 2003. Epidemiology and drinking water—are we running dry? Epidemiology 14(6):635–636.

Steinmaus, C., Y. Yuan, M. N. Bates, and A. H. Smith. 2003. Case-control study of bladder cancer and drinking water arsenic in the western United States. American Journal of Epidemiology 158:1193–1201.

Stout, J. E., V. L. Yu, and M. G. Best. 1985. Ecology of Legionella pneumophila within water distribution systems. Appl. Environ. Microbiol. 49:221–228.

Strauss, B., W. King, A. Ley, and J. R. Hoey. 2001. A prospective study of rural drinking water quality and acute gastrointestinal illness. BMC Public Health 1:8.

Taylor, R. H., J. O. Falkinham, III, C. D. Norton, and M. W. LeChevallier. 2000. Chlorine, chloramine, chlorine dioxide, and ozone susceptibility of Mycobacterium avium. Appl. Environ. Microbiol. 66(4):1702–1705.

Thofern, E., D. Schoenen, and G. J. Tuschewitzki. 1987. Microbial surface colonization and disinfection problems. Off Gesundh.-wes., 49:Suppl:14.

Thomas, V., K. Herrera-Rimann, D. S. Blanc, and G. Greub. 2006. Biodiversity of amoebae and amoeba-resisting bacteria in a hospital water network. Appl. Environ. Microbiol. 72(4):2428–2438.

Trautman, M., T. Michalsky, H. Wiedeck, V. Radlosavljevic, and M. Ruhnke. 2001. Tap wate colonization with Pseudomonas aeruginosa in a surgical intensive care unit (ICU) and relation to Pseudomonas infections of ICU patients. Infect. Control. Hosp. Epidemiol. 22:49–52.

Vaerewijck, M. J. M., G. Huys, J. P. Palomino, J. Swings, and F. Portaels. 2005. Mycobacteria in drinking water distribution systems: ecology and significance for human health. FEMS Microbiology Review 29:911–934.

von Reyn, C. F., R. D. Waddell, T. Eaton, R. D. Arbeit, J. N. Maslow, T. W. Barber, R. J. Brindle, C. F. Gilks, J. Lumio, J. Lahdevirta, A. Ranki, D. Dawson, and J. O. Falkinham, III. 1993. Isolation of Mycobacterium avium complex from water in the

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×

United States, Finland, Zaire, and Kenya. Journal of Clinical Microbiology 31:32273230.

von Reyn, C. F., J. N. Maslow, T. S. Barber, J. O. Falkinham, III, and R. D. Arbeit. 1994. Persistent colonisation of potable water as a source of Mycobacterium avium infection in AIDS. Lancet 343:11371141.

VROM. 2005. Legionella in Europe: Problems and Prevention: Summary by the Chairman. International Congress September 28–29, Amsterdam, The Netherlands. Available on-line at http://international.vrom.nl/docs/internationaal/congres%20verslag%20Engels%2025-11-04.pdf. Accessed May 11, 2006.

Wade, T. J., S. K. Sandhu, D. Levy, S. Lee, M. W. LeChevallier, L. Katz, and J. M. Colford, Jr. 2004. Did a severe flood in the Midwest cause an increase in the incidence of gastrointestinal symptoms? Amer. J. Epidemiol. 159(4):398–405.

Wadowsky, R. M., and R. B. Yee. 1983. Satellite growth of Legionella pneumophila with an environmental isolate of Flavobacterium breve. Appl. Environ. Microbiol. 46:1447–1449.

Wadowsky, R. M., and R. B. Yee. 1985. Effect of non-legionellaceae bacteria on the multiplication of Legionella pneumophila in potable water. Appl. Environ. Microbiol. 49:1206–1210.

Waller, K., S. H. Swan, G. DeLorenze, and B. Hopkins. 1998. Trihalomethanes in drinking water and spontaneous abortion. Epidemiology 9:134-140.

Wolinsky, E. 1979. Nontuberculous Mycobacteria and associated diseases. Am. Rev. Respir. Dis. 119:107–159.

World Health Organization (WHO). 2002. Heterotrophic Plate Count Measurements in Drinking Water Safety Management. WHO/SDE/WSH/02.10. Geneva, Switzerland: WHO.

Yang, C. Y., H. F. Chiu, M. F. Cheng, and S. S. Tsai. 1998. Chlorination of drinking water and cancer mortality in Taiwan. Environmental Research 78:1–6.

Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 87
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 88
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 89
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 90
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 91
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 92
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 93
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 94
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 95
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 96
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 97
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 98
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 99
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 100
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 101
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 102
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 103
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 104
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 105
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 106
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 107
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 108
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 109
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 110
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 111
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 112
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 113
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 114
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 115
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 116
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 117
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 118
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 119
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 120
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 121
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 122
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 123
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 124
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 125
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 126
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 127
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 128
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 129
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 130
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 131
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 132
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 133
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 134
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 135
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 136
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 137
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 138
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 139
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 140
Suggested Citation:"3 Public Health Risk from Distribution System Contamination." National Research Council. 2006. Drinking Water Distribution Systems: Assessing and Reducing Risks. Washington, DC: The National Academies Press. doi: 10.17226/11728.
×
Page 141
Next: 4 Physical Integrity »
Drinking Water Distribution Systems: Assessing and Reducing Risks Get This Book
×
Buy Hardback | $94.00 Buy Ebook | $74.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Protecting and maintaining water distributions systems is crucial to ensuring high quality drinking water. Distribution systems -- consisting of pipes, pumps, valves, storage tanks, reservoirs, meters, fittings, and other hydraulic appurtenances -- carry drinking water from a centralized treatment plant or well supplies to consumers’ taps. Spanning almost 1 million miles in the United States, distribution systems represent the vast majority of physical infrastructure for water supplies, and thus constitute the primary management challenge from both an operational and public health standpoint. Recent data on waterborne disease outbreaks suggest that distribution systems remain a source of contamination that has yet to be fully addressed. This report evaluates approaches for risk characterization and recent data, and it identifies a variety of strategies that could be considered to reduce the risks posed by water-quality deteriorating events in distribution systems. Particular attention is given to backflow events via cross connections, the potential for contamination of the distribution system during construction and repair activities, maintenance of storage facilities, and the role of premise plumbing in public health risk. The report also identifies advances in detection, monitoring and modeling, analytical methods, and research and development opportunities that will enable the water supply industry to further reduce risks associated with drinking water distribution systems.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!