Outbreaks in drinking--water systems, 1991-1998.

Introduction

In this article, the authors review information about outbreaks in public and individual drinking-water systems that were reported to the Centers for Disease Control and Prevention (CDC) and the U.S. Environmental Protection Agency (U.S. EPA) during 1991-1998. CDC, U.S. EPA, and the Council of State and Territorial Epidemiologists maintain a collaborative water-disease-outbreak surveillance system (Barwick, Levy, Craun, Beach, & Calderon, 2000; Herwaldt, Craun, Stokes, & Juranek, 1991; Kramer, Herwaldt, Craun, Calderon, & Juranek, 1996; Levy, Bens, Craun, Calderon, & Herwaldt, 1998; Moore et al., 1993). Because the surveillance system is voluntary and does not include data for sporadic cases of disease that may be waterborne, the statistics do not reflect the true incidence of waterborne outbreaks or disease. The surveillance system collects information on the types of water systems, their deficiencies, their water quality, and the respective etiological agents associated with the outbreaks; these data are us eful for evaluating the relative degrees of risk associated with different types of source water and systems, problems in current technologies and operating conditions, and the adequacy of current regulations.

Definitions

For an incident to be classified as a waterborne outbreak, at least two people must have experienced a similar illness after the ingestion of drinking water or after exposure to water used for recreational purposes, and epidemiological evidence must implicate water as the probable source of the illness. An exception is made for a single case of chemical poisoning (e.g., methemoglobinemia) when water-quality data indicate contamination by the chemical.

The surveillance system reports outbreaks associated with drinking water, recreational waters (swimming pools, water parks, and fresh and marine surface waters), and whirlpool- and hot tub--associated outbreaks of dermatitis, but only outbreaks related to drinking water are included in the analysis here. Although most of the reported outbreaks were associated with water used or intended for drinking or domestic purposes, several were associated with the ingestion of water not intended for consumption (e.g., the use of springs and creeks by backpackers and campers) and miscellaneous deficiencies, such as water contaminated at its point of use (e.g., water faucets, serving containers, and ice).

Public-water systems are regulated by U.S. EPA and are classified as either community or noncommunity systems. A community water system serves year-round residents of a community, subdivision, or mobile-home park that has 15 or more service connections or serves an average of 25 or more residents. A noncommunity water system is used by the general public for 60 days or more per year and has at least 15 service connections or serves an average of 25 or more people. Of the approximately 170,000 public-water systems in the United States, 55,000 (32 percent) are community systems and 115,000 (68 percent) are noncommunity systems. Community water systems serve some 243 million people. Millions of people use noncommunity systems while traveling or working in the United States. Some 24 million people (9 percent) rely on individual water systems, which are not owned or operated by a water utility and serve less than 15 connections or less than 25 people. The responsibility for regulating individual water systems rest s with the state or county.

Outbreak Statistics

During 1991-1998, 126 drinking-water outbreaks and more than 429,000 cases of illness were reported in 41 states and three U.S. territories (Table 1). Most outbreaks (87 percent) were reported in public water systems. Sixty-two outbreaks (49 percent) and 6,562 illnesses were reported in noncommunity water systems. Forty-seven outbreaks (37 percent) and 422,298 illnesses were reported in community water systems. A single outbreak of cryptosporidiosis in Milwaukee was responsible for an estimated 403,000 cases of illness, 440 hospitalizations, and 50 deaths in 1993 (MacKenzie et al., 1994). This outbreak is the largest reported in the United States since the collection of statistics began in 1920. Seventeen outbreaks (14 percent) and 161 illnesses were reported in individual water systems and among persons who consumed water not intended for drinking.

More than 156,000 public systems (92 percent) rely primarily on groundwater sources, and slightly more than 13,000 systems (8 percent) use surface water. Except for 11 public water systems where the outbreak was caused by contaminated ice, water taps, or storage containers, or where the water source was not identified, water systems were classified either as surface water or as groundwater. Seventy-four (68 percent) of the outbreaks occurred in public systems that used groundwater; 24 outbreaks (22 percent) occurred in surface-water systems.

A bacterial, viral, or protozoan etiology was identified in 51 outbreaks (41 percent) (Table 2). In 52 outbreaks (41 percent), an infectious etiological agent was suspected but the etiology was not identified. Of the 31 community-system outbreaks with a known or suspected infectious etiology, 15 (48 percent) occurred in surface-water systems and 13 (42 percent) in groundwater systems. Of the 61 infectious-disease outbreaks reported in noncommunity systems, 52 (85 percent) occurred in groundwater systems, and only two (3 percent) occurred in surface-water systems. Eight (73 percent) of the 11 infectious-disease outbreaks in individual water systems occurred in groundwater systems.

Chemical contaminants were responsible for 23 (18 percent) of all outbreaks; 16 outbreaks (34 percent) occurred in community systems. Most community-system chemical outbreaks (69 percent) were associated with distribution system contamination. High copper levels due to corrosive water were identified in seven outbreaks (44 percent). Cross-connections and back-siphonage (nitrate, soap concentrate, and fluoride) caused four outbreaks (25 percent). Five outbreaks (31 percent) were associated with inadequate control of chemicals added during water treatment (fluoride, sodium hydroxide, and chlorine).

Only one chemical outbreak (copper) was reported in a noncommunity water system, but six of the outbreaks (35 percent) in individual systems were associated with chemicals. Three single-case outbreaks of methemoglobinemia were due to high levels of nitrate in well water. In one outbreak, nitrate levels exceeded maximum limits even though a reverse-osmosis membrane filter had reduced levels in the water source (Moore et al., 1993). Three single-case outbreaks were detected through a leadscreening program (Kramer et al., 1996). Water at the homes of infants with high blood lead levels was corrosive, leaching lead from fittings and lead-soldered seams in water storage tanks.

Because of the large outbreak in Milwaukee, almost all (99 percent) of the illnesses in community-system outbreaks occurred in surface-water systems. Outbreaks in noncommunity systems caused 6,562 illnesses, almost all (94 percent) of which occurred in groundwater systems.

Water System Deficiencies

For most water systems, an investigation identified the deficiency that likely caused the outbreak. Water treatment deficiencies included inadequate, interrupted, or no disinfection of groundwater; inadequate disinfection of unfiltered surface water; and inadequate or interrupted filtration of surface water (Table 3). Outbreaks also were caused by contaminated, untreated groundwater (Table 4) and contaminants that entered the distribution system (Table 5).

Outbreaks in community water systems were associated with distribution system deficiencies (36 percent), inadequately treated groundwater (19 percent), inadequately treated surface water (17 percent), miscellaneous or undetermined causes (17 percent), and inadequate control of chemical feed (11 percent). In noncommunity water systems, outbreaks were associated with inadequately treated groundwater (63 percent), miscellaneous or undetermined causes (23 percent), distribution system deficiencies (13 percent), and inadequately treated surface water (1 percent). Outbreaks in individual water systems were associated with inadequately treated groundwater (59 percent), distribution system deficiencies (24 percent), miscellaneous or undetermined causes (11 percent), and inadequately treated surface water (6 percent).

Deficiencies Associated with Source Water Contamination

Inadequate water treatment in community systems caused 17 outbreaks (36 percent), most (53 percent) of which were associated with contaminated wells (Table 3). Almost all (91 to 97 percent) of the outbreaks associated with source-water contamination and inadequate treatment in individual and noncommunity systems occurred in well water systems. In noncommunity systems, slightly fewer outbreaks (46 percent) were associated with contaminated, untreated well water than were caused by inadequate and interrupted disinfection of well water (54 percent). Two outbreaks were associated with deficiencies in UV-light disinfection systems, and in the remaining outbreaks, chlorine disinfection was either inadequate or interrupted.

Sources of contamination were identified in 71 percent of the outbreaks associated with untreated well water (Table 4). Sewage contamination of wells was identified in eight out-breaks (26 percent). Other infectious-disease outbreaks were associated with wells drilled in limestone and fissured rock (16 percent), wells vulnerable to surface-water runoff or floods (13 percent), and improperly constructed or located wells (10 percent).

Distribution System Deficiencies

In community water systems, distribution system contamination caused as many outbreaks as source-water contamination. A distribution system deficiency was identified in 17 (36 percent) of the outbreaks reported in community systems (Table 5). Outbreaks occurred when contaminants entered the distribution system either through a cross-connection (47 percent) or during storage or main repair (12 percent); corrosive water was responsible for seven outbreaks (41 percent) (Table 6). Although noncommunity and individual water systems generally have small, limited water distribution networks, they are also subject to distribution system contamination and corrosive water. Four (24 percent) of the outbreaks in individual water systems and eight (13 percent) of the outbreaks in noncommunity systems were associated with distribution system deficiencies, primarily cross-connections and corrosive water.

Miscellaneous and Undetermined Deficiencies

Seventeen outbreaks (71 percent) were associated with the use of contaminated ice or water at its point of use, consumption of water not intended for drinking, and contaminated bottled water (Table 7). These deficiencies are not related to the adequacy of water treatment or operation and should be corrected by food safety regulations and health education. An outbreak of acute gastroenteritis among hikers was associated with the use of water from an outside tap at a wastewater plant. The tap was not marked as nonpotable, and water from the tap had more than 23 fecal coliforms per 100 milliliters (mL)(Levy et al., 1998). In another outbreak, consumption of water from an outdoor faucet by attendees at a day camp resulted in 77 primary cases of cryptosporidiosis and 24 probable secondary cases in household members (Levy et al., 1998). In Saipan, an outbreak of non-O1 Vibrio cholerae was associated with bottled water from two commercial bottling plants (Kramer et al., 1996). Bottled water tested positive for fecal coliforms, but the source of contamination was not determined.

In seven outbreaks (29 percent), a water system deficiency was not identified, primarily because of an incomplete investigation (Table 7). In an extensively investigated outbreak in Clark County, Nevada, however, no water treatment deficiencies or breakdowns were discovered (Goldstein et al., 1994), and water quality was found to be excellent before, during, and after the outbreak (Roefer, Monscvitz, & Rexing, 1996). Although investigators concluded that cryptosporidiosis among HIV-positive adults was associated with consumption of municipal water, there are concerns that the observed waterborne association may be due to recall bias (Craun & Frost, 2002).

Water Quality

The chemical or pathogen responsible for the outbreak was detected in the water source or water system during 21 (91 percent) of 23 chemical outbreaks and 21 (78 percent) of 27 infectious-disease outbreaks in which samples were collected and analyzed. Cryptosporidium, Giardia, Shigella, and E. coli O157:H7 were isolated from water samples collected during eight, six, three, and two outbreaks, respectively. Enteric viruses and Salmonella were isolated during one outbreak each. Water samples were analyzed for coliform bacteria during the investigation of 81 outbreaks of infectious disease (Table 8).

U.S. EPA's Total Coliform Rule (TCR) requires public water systems to routinely monitor their tap water for total coliforms (U.S. EPA, 1989a, 1990). Coliform bacteria are members of the Enterobacteriaceae (e.g., species of Enterobacter, Klebsiella, Citrobacter, and Escherichia) and have been used for decades to assess the microbiological quality of drinking water. These bacteria are present in the normal intestinal flora of humans and other warm-blooded animals and are found in large numbers in fecal wastes. Most species of coliforms are also free-living in the environment, and their presence in drinking water does not necessarily represent fecal contamination. During 50 outbreaks, samples were collected and analyzed for fecal coliforms, a marker of human fecal contamination.

Water samples were usually collected during the early stages of an outbreak. In several investigations, samples were not collected until two to four weeks after the beginning of the outbreak. In one outbreak, water samples were collected two months after the beginning of the outbreak. In seven outbreak investigations, water samples were collected from the water sources (wells) of noncommunity systems rather than from the distribution systems. In these seven outbreaks, information from source-water samples was used as a substitute for information from tap water samples. Because the well water was not treated, it was presumed that tap water samples, if they bad been collected, would likely be of similar quality.

During many outbreak investigations, more coliform samples were collected than would be required by the TCR for routine monitoring purposes, and the samples were collected during relatively short periods of time, usually one- to two-week periods. As many as 60-samples (median = 4) and as many as 16 samples (median = 3) were collected during the investigation of outbreaks in community and non-community systems, respectively.

Total coliforms were detected during the investigation of 59 outbreaks (73 percent) in which samples were collected (Table 9). Coliforms were found in all individual systems and most noncommunity systems (83 percent) but in only 46 percent of the community systems. In 12 noncommunity-, five individual-, and three community-system outbreaks, the only water sample that was collected was found to be coliform positive. In six outbreaks, one coliform-positive sample was found, with two to 20 samples collected from each outbreak. Fecal coliforms were found in 41 (82 percent) of the outbreaks in which water samples were analyzed for fecal coliforms (Table 9).

Total coliforms were detected during the investigation of all 16 outbreaks of bacterial etiology (100 percent) and during the investigation of 35 outbreaks (81 percent) caused by viruses and unidentified agents; however, coliforms were detected during only eight outbreaks (36 percent) caused by Cryptosporidium or Giardia (Table 10). In community systems, coliforms were detected in five outbreaks caused by protozoa (33 percent). In noncommunity systems, coliforms were not found during the four protozoan outbreaks in which samples were collected, but in individual systems, coliforms were detected during the investigation of all three protozoan outbreaks in which samples were collected.

Discussion

Outbreaks of Undetermined Etiology

Success in identifying an etiological agent in waterborne outbreaks has improved only slightly in the past 25 years. During 1991-1998, an etiological agent was determined in 59 percent of waterborne outbreaks compared with 45 percent and 55 percent of outbreaks reported during 1971-1980 and 1981-1990, respectively. Investigations are not always timely and clinical specimens are not always collected. Even when specimens are collected, the analysis may not be comprehensive. Although methods for detection of enteric viruses have improved greatly in recent years (Ando et al., 1995; Barwick et al., 2000), diarrheal stool specimens generally are examined for bacterial pathogens, not for viruses. In many laboratories, testing for Cryptosporidium is done only if requested and may not include routine stool examinations for ova and parasites (Boyce, Pemberton, & Addiss, 1996; Levy et al., 1998).

During 1993-1996, illnesses associated with at least six of 14 outbreaks of undefined etiology had incubation periods, durations, and symptom complexes that were consistent with viral syndromes, but limited or no viral testing was done (Kramer et al., 1996; Levy et al., 1998). Testing in four outbreaks of suspected viral etiology found that stool specimens were negative for bacterial pathogens, and testing in three outbreaks found that stool specimens were negative for both bacterial and parasitic pathogens (Kramer et al., 1996; Levy et al., 1998). In one outbreak of undetermined etiology in 1995, the state laboratory identified rotavirus in two stool specimens, but no testing was done for Norwalk-like viruses, which have been more commonly associated with waterborne outbreaks (Levy et al., 1998).

Additional knowledge about etiological agents can provide more information about the effectiveness of regulatory and other efforts to reduce waterborne outbreak risks. Investigators are encouraged to collect clinical specimens and submit them for complete testing either to CDC or to state laboratories that conduct viral and protozoan, as well as bacterial analyses.

Water System Deficiencies

Almost half (44 percent) of all outbreaks in public water systems were caused by the inadequate disinfection of groundwater and use of untreated, contaminated groundwater. These statistics suggest that additional regulations and increased surveillance activities may be needed for groundwater systems. Wells and springs should be better protected from sources of contamination, especially surface-water runoff and underground seepage from septic-tank and other sewage discharges. Periodic sanitary surveys and hydrogeologic assessments can help identify systems with the highest possibility of fecal contamination. Some groundwaters may require disinfection. Disinfection should be adequate in terms of contact time and concentration, not interrupted, and sufficient to maintain a chlorine residual throughout the system.

Distribution system deficiencies were responsible for a significant number of outbreaks in public water systems, especially community systems. Increased attention should be paid to maintaining the integrity of the distribution system and preventing contamination from entering the system through cross-connections and back-siphonage, and during water main repairs and construction. Water system managers also should assess the corrosive potential of their source water, identify possible sources of contamination for distribution system storage facilities, and take appropriate steps to remedy any deficiencies. Chlorine-residual monitoring can help warn of potential contamination. When decreased residual levels are detected, operators should initiate investigations to identify sources of contamination.

Giardia and Cryptosporidium continue to pose waterborne risks, especially in groundwater systems, where 55 percent of 11 outbreaks caused by Cryptosporidium and 40 percent of 15 outbreaks caused by Giardia were associated with groundwater contamination. The inadequate treatment of surface water was responsible for 27 percent of cryptosporidiosis outbreaks and 40 percent of giardiasis outbreaks. The continued occurrence of protozoan outbreaks in surfacewater systems underscores the importance of meeting U.S. EPA's turbidity standards and other provisions for treating surface-water sources (U.S. EPA 1989b, 1991a, 1994). It is also important to assess the potential for sewage contamination of groundwater sources, especially seepage from on-site systems and surface runoff. Groundwaters contaminated by protozoa are considered to be under the direct influence of surface water and are subject to U.S. EPA's filtration requirements.

Coliforms in Water During Outbreaks

Coliforms were found during the investigation of all outbreaks of bacterial etiology, but in less than half of the outbreaks caused by Giardia or Cryptosporidium. Similar findings were reported in a previous study of waterborne outbreaks reported during 1983-1992 (Craun, Berger, & Calderon, 1997). Although coliforms can be adequate indicators for bacterial pathogens, they are inadequate indicators for the presence of protozoa. Protozoa are more resistant to water disinfectants than are bacteria, and disinfection may be sufficient to inactivate coliforms but not protozoa.

As in earlier findings (Craun et al., 1997), coliforms were detected during most outbreaks of undetermined etiology. If unidentified viruses caused half of these outbreaks, as suggested by recent outbreaks, coliforms may sometimes serve as an indicator for waterborne viruses. Coliforms may not always, however, indicate the presence of waterborne viruses. Some viruses survive longer in water than do coliform bacteria and are more resistant than coliforms to water disinfectants normally used in water treatment (Sobsey, 1989; National Academy of Sciences, 1980). Also, coliforms were detected during only two of the four viral outbreaks reported during 1991-1998.

Although the total-coliform test may indicate the presence of bacterial pathogens during an outbreak, it is important to remember that coliform analyses during an outbreak investigation may not be relevant in assessing the vulnerability of water systems to an outbreak. Water quality monitoring is often intense during an outbreak investigation, and water samples are being collected during or just after a contamination event. In some instances, contamination was sufficiently large that a single sample was found to be positive for coliforms. In other instances, no coliforms were detected even though a large number of samples (16 to 44) were collected. This finding emphasizes that water contamination sufficient to cause an outbreak can be intermittent and short-lived. Thus, the timing of the sample collection, the number of samples collected, and the selection of appropriate indicators are all important considerations in the development of a monitoring program that can assess a system's vulnerability to an outbr eak. For example, under the current TCR minimum-sample-collection requirements, a large or continuous source of contamination would be necessary to detect coliforms in the few samples collected by small systems. Community systems that serve 1,000 or fewer persons are required to collect only a single coliform sample each month. For systems serving 1,001 to 4,900 persons, only two to five samples per month are required. Only for communities of more than 25,000 people are 30 or more samples required each month. Most noncommunity systems need collect only one sample per month or quarter.

Routine coliform surveillance records for 45 outbreaks during 1991-1998 showed that only 22 percent of community and 9 percent of noncommunity systems experiencing an outbreak had violated U.S. EPA's maximum contaminant limit (MCL) for total coliforms in the 12-month period before the outbreak. This finding suggests that current routine coliform-monitoring programs alone are not able to identify water systems that are vulnerable to an outbreak. Similar findings were reported in an evaluation of the previous TCR (Craun et al., 1997; U.S. EPA, 1975). In that study only 38 percent and 19 percent of community systems reporting an outbreak had exceeded the total-coliform MCL during the three and 12 months before the outbreak, respectively. The greater frequency of coliform detection in water systems during an outbreak investigation may be due to more intensive monitoring, insufficient coliform-monitoring requirements, or both.

Conclusions

Most waterborne outbreaks were caused by inadequate or no treatment of groundwater and the contamination of water during its storage and distribution to the customer, especially through cross-connections and corrosion. Additional regulations and increased surveillance activities, along with appropriate corrective measures, are needed to reduce waterborne-disease risks posed by these sources of contamination. Outbreak data underscore the importance of U.S. EPA's filtration and turbidity requirements for surface-water systems and for groundwater systems under the direct influence of surface water. U.S. EPA (1991b) has promulgated regulations to control corrosion by-product, and water utilities should pay more attention to cross-connection control regulations and inspections, monitoring and treatment for corrosive water, and maintaining distribution system integrity. To reduce waterborne-disease risks in groundwater systems, U.S. EPA (1999) has recently proposed requirements for sanitary surveys, hydrogeologic a ssessments, and disinfection. Protozoan outbreaks in groundwater systems emphasize the importance of assessing the potential for sewage contamination of wells and springs and identifying sources that are under the direct influence of surface water.

Outbreak data suggest that coliform-monitoring requirements are inadequate to identify water systems that are vulnerable to an outbreak. For most public systems that reported an outbreak, the TCR required the collection of only one to three coliform samples each month. The TCR focuses on detection of coliforms in the distribution system after treatment, It may be more effective for groundwater systems to monitor source-water quality and assess the potential for contamination from sewage discharges and surface water. Although coliforms adequately indicate the presence of bacterial pathogens in drinking water during an outbreak, they are not a good indicator for Cryptosporidiwn, Giardia, and waterborne viruses. Consequently other indicators should be considered. Activities such as frequent sanitary surveys should be conducted in addition to monitoring.

Although the reporting of outbreaks is incomplete and the accuracy of case counts varies, waterborne-outbreak-surveillance data have helped identify the types of water systems, their deficiencies, and the respective etiologic agents associated with the outbreaks. Outbreak detection activities and investigations should be improved so that more complete information is obtained about the causes and etiologies of waterborne outbreaks. Epidemiologists should collaborate with engineers and environmental health specialists during investigations to evaluate water systems and sources of contamination. These data are important for evaluating source water protection strategies, water treatment technologies, and drinking-water regulations and for influencing research priorities.

TABLE 1

Waterborne Outbreaks and Cases of Illness, by Type of Water System,
1991-1998

Type Water        Outbreaks   Cases    Emergency   Hospitalizations
System                                Room Visits

Cummunity water      47      422,298       8             599
systems *

Noncommunity         62       6,562        8              41
water systems

Individual Water     17        161         0              13
systems

Total               126      429,021      16             653

Type Water        Deaths
System

Cummunity water     58
systems *

Noncommunity         0
water systems

Individual Water     0
systems

Total               58

* 403,000 cases illness reported in Milwaukee.
TABLE 2

Etiology of Waterborne Outbreaks, by Type of Water System, 1991-1998

Etiological Agent             Community Water         Noncommunity Water
                                  Systems                  Systems

                           Outbreaks    Cases     Outbreaks     Cases

Undetermined                   7       10,295        42         5,140

Chemical                      16         513          1           2

Giardia                       10        1,986         3          128

Cryptosporidium *              7       407,714        2          578

Norwalk-like virus             2         742

Campylobacter                  1         172          3          58

Salmonella, nontyphoid         1         625

E. coli                        1         157          1           3

Shigella                       1         625

Y. cholerae                    1         11

Hepatitis A virus                                     1          46

Plesiomonas shigelloides                              1          60

Total                         47       422,298       62         6,562

Etiological Agent            Individual Water               All
                                  Systems                 Systems

                          Outbreaks    Cases     Outbreaks   Cases

Undetermined                  3          51         52      15,486

Chemical                      6          8          23        523

Giardia                       2          12         15       2,126

Cryptosporidium *             2          39         11      408,331

Norwalk-like virus                                   2        742

Campylobacter                                        4        230

Salmonella, nontyphoid                               1        625

E. coli                       4          66          6        226

Shigella                                             1        625

Y. cholerae                                          1        11

Hepatitis A virus             1          10          2        56

Plesiomonas shigelloides                             1        60

Total                        17         161         126     429,021

* 403,000 cases of illness reported in Milwaukee.
TABLE 3

Water System Deficiencies in Outbreaks Associated with Source-Water
Contamination, 1991-1998

Type of Contamination or
Water System Deficiency          Number of Outbreaks

                                  Community Systems
                               Surface-     Ground-
                                 Water       Water
                                Source       Source

Untreated water                    0           5

Inadequate or
 interrupted disinfection;
 disinfection-only treatment       3           1

Inadequate or interrupted          4           1
 filtration

Inadequate chemical removal        0           0

Total                              8           9

Type of Contamination or
Water System Deficiency              Number of Outbreaks

                                     Noncommunity Systems
                                     Surface-         Ground-
                                      Water            Water
                                      Source          Source

Untreated water                         0               18

Inadequate or
 interrupted disinfection;
 disinfection-only treatment            1               21

Inadequate or interrupted               0                0
 filtration

Inadequate chemical removal             0                0

Total                                   1               39

Type of Contamination or
Water System Deficiency                   Number of Outbreaks

                                 Individual Water Systems      All Water
                                                                 Systems
                                Surface-       Ground-       Surface-
                                 Water          Water          Water
                                 Source         Source        Source

Untreated water                    1              8              1

Inadequate or
 interrupted disinfection;
 disinfection-only treatment       0              1              5

Inadequate or interrupted          0              0              4
 filtration

Inadequate chemical removal        0              1              0

Total                              1             10             10

Type of Contamination or
Water System Deficiency       Number of
                              Outbreaks

                              All Water
                               Systems
                               Ground
                               Water
                               Source

Untreated water                  31

Inadequate or
 interrupted disinfection;
 disinfection-only treatment     25

Inadequate or interrupted         1
 filtration

Inadequate chemical removal       1

Total                            58
TABLE 4

Sources of Contamination, Outbreaks in Untreated Groundwater (Wells),
1991-1998

                             Community  Noncommunity  Individual
Deficiency                    Systems     Systems      Systems

Overflow, seepage                1           5            2
of sewage

Contamination through            1           4            0
limestone, fissured rock

Insufficient information         1           6            2

Improper construction/           0           1            2
location

Chemical                         0           0            2

Surface runoff, heavy rains      2           2            0

Total                            5          18            8

                               All
Deficiency                   Systems

Overflow, seepage               8
of sewage

Contamination through           5
limestone, fissured rock

Insufficient information        9

Improper construction/          3
location

Chemical                        2

Surface runoff, heavy rains     4

Total                          31
TABLE 5

Water System Deficiencies in Outbreaks Not Related to Source-Water
Contamination, 1991-1998

Type of Contamination of                Number of Outbreaks
Water System Deficiency

                          Community  Noncommunity  Individual    All
                            Water       Water        Water     Systems
                           Systems     Systems      Systems

Distribution system          17           8            4         29
contamination

Inadequate control of         5           0            0          5
chemical feed

Miscellaneous/unknown         8           14           2         24

Total                        30           22           6         58
TABLE 6

Distribution System Deficiencies Causing Outbreaks, 1991-1998

Deficiency                              Community  Noncommunity
                                         Systems     Systems

Cross-connection/back-siphonage             8           5

Corrosion/leaching of metals                7           0

Broken/leaking water mains/main repair      1           1

Contamination during storage                1           2

Total                                       17          8

Deficiency                              Individual    All
                                         Systems    Systems

Cross-connection/back-siphonage             0         13

Corrosion/leaching of metals                3         10

Broken/leaking water mains/main repair      0          2

Contamination during storage                1          4

Total                                       4         29
TABLE 7

Miscellaneous Deficiencies Causing Outbreaks, 1991-1998

Deficiency                       Community  Noncommunity  Individual
                                  Systems     Systems      Systems

Insufficient information             4           3            0

Contaminated ice                     1           7            0

Contaminated container               1           3            0

Water not intended for drinking      0           0            2

Contaminated drinking fountain       1           0            0

Contamination of bottled water       1           1            0

Total                                8          14            2

Deficiency                         All
                                 Systems

Insufficient information            7

Contaminated ice                    8

Contaminated container              4

Water not intended for drinking     2

Contaminated drinking fountain      1

Contamination of bottled water      2

Total                              24
TABLE 8

Total-Coliform Data Collected During Waterborne-Outbreak Investigations,
1991-1998 *

Water System Type                Number of Outbreaks

                   Unidentified
                     Etiology    Bacterial  Viral       Protozoan

Community               5            4        2            16

Noncommunity            34          11        1             5

Individual              3            3        1             4

Totals                  42          18        4            25

Water System Type                Number of Outbreaks

                                   Total-Coliform  Fecal-Coliform
                   All Etiologies  Data Available  Data Available

Community                27              26              11

Noncommunity             51              46              33

Individual               11              9               6

Totals                   89              81              50

* Excluded from analysis are chemical outbreaks and outbreaks associated
with contaminated bottled water, ice, faucets, and containers.
TABLE 9

Total and Fecal Coliforms Detected During Outbreak Investigations,
1991-1998 *

Water System Type                    Number of Outbreaks
                   Total Coliforms  Total Coliforms Not  Fecal Coliforms
                      Detected           Detected           Detected

Community                12                 14                 10

Noncommunity             38                  8                 26

Individual                9                  0                  5

Totals                   59                 22                 41

Water System Type   Number of Outbreaks
                   Fecal Coliforms Not
                        Detected

Community                   1

Noncommunity                7

Individual                  1

Totals                      9

* Excluded from analysis are chemical outbreaks and outbreaks associated
with contaminated bottled water, ice, faucets, and containers.
TABLE 10

Total Coliforms Detected During Waterborne-Outbreak Investigations,
1991-1998

Total Coliforms              All Etiologies  Protozoan Etiology

Community water systems,        14 (54%)          10 (67%)
not detected

Community water systems,        12 (46%)          5 (33%)
detected

Noncommunity water systems,     8 (17%)           4 (100%)
not detected

Noncommunity water systems,     38 (83%)             0
detected

Individual water systems,          0                 0
not detected

Individual water systems,       9 (100%)          3 (100%)
detected

All water systems,              22 (27%)          14 (64%)
not detected

All water systems,              59 (73%)          8 (36%)
detected

Total Coliforms              Bacterial Etiology  Viral Etiology

Community water systems,             0              2 (100%)
not detected

Community water systems,          4 (100%)             0
detected

Noncommunity water systems,          0                 0
not detected

Noncommunity water systems,      10 (100%)          1 (100%)
detected

Individual water systems,            0                 0
not detected

Individual water systems,         2 (100%)          1 (100%)
detected

All water systems,                   0              2 (50%)
not detected

All water systems,               16 (100%)          2 (50%)
detected

Total Coliforms              Unknown Etiology

Community water systems,         2 (40%)
not detected

Community water systems,         3 (60%)
detected

Noncommunity water systems,      4 (13%)
not detected

Noncommunity water systems,      21 (87%)
detected

Individual water systems,           0
not detected

Individual water systems,        3 (100%)
detected

All water systems,               6 (15%)
not detected

All water systems,               33 (85%)
detected

REFERENCES

Ando, T., Monroe, S.S., Gentsch, J.R., Jin, Q., Lewis, D.C., & Glass, R.I. (1995). Detection and differentiation of antigenically distinct small round-structured viruses (Norwalk-like viruses) by reverse transcription--PCR and southern hybridization. Journal of Clinical Microbiology, 33(1), 64-71.

Barwick, R.S., Levy, D.A., Craun, G.F, Beach, M.J., & Calderon, R.L. (2000). Surveillance for waterborne-disease outbreaks--United States, 1997-1998. Morbidity and Mortality Weekly Report, 49(SS-4), 1-36.

Boyce, T.G., Pemberton, A.G., & Addiss, D.G. (1996). Cryptosporidum testing practices among clinical laboratories in the United States. Pediatric Infectious Diseases Jounral, 15(1), 87-88.

Craun, G.E, & Frost, F.J. (2002). Possible information bias in a waterborne outbreak. International Journal of Environmental Health Research, 12, 5-15.

Craun, G.F., Berger, P.S., & Calderon, R.L. (1997). Coliform bacteria and waterborne disease outbreaks. Journal of the American Water Works Association, 89(3), 96-104.

Goldstein, S.T., Juranek, D.D., Ravenholt, O., Hightower, A.W, Martin, D.G., Mesnik, J.L., Griffiths, S.D., Bryant, A.J., Reich, R.R, & Herwaldt, B.L. (1996). Cryptosporidiosis: An outbreak associated with drinking water despite state-of-the-art water treatment. Annals of Internal Medicine, 124(5), 459-468 [erratum appears in 125(20), 158].

Herwaldt, B.L., Craun, GE, Stokes, S.L., & Juranek, D.D. (1991). Waterborue-disease outbreaks, 1989-1990. Morbidity and Mortality Weekly Report, 40(SS-3), 1-43.

Kramer, M.H., Herwaldt, B.L, Craun, G.F., Calderon, R.L, & Juranek, D.D. (1996). Surveillance for waterborne-disease outbreaks--United States, 1993-1994. Morbidity and Mortality Weekly Report, 45(55-1), 1-33.

Levy DA., Bens, MS., Craun, G.F., Calderon, R.L., & Herwaldt, B.L (1998). Surveillance for waterborne-disease outbreaks--United States, 1995-1996. Morbidity and Mortality Weekly Report, 47(55-5), 1-33.

Mac Kenzie, WR., Hoxie, N.J., Proctor, M.E., Gradus, MS., Blair, K.A,, Peterson, D.E., Karmierezak, J.J., Addiss, D.G., Fox, K.R., Rose, J.B., & David, JR (1994). A massive outbreak in Milwaukee of Cryptosporidium infection transmitted through the public water supply. New England Journal of Medicine, 331(3), 161-167.

Moore, A.C., Hewwaldt, B.L., Craun, GE, Calderon, R.L., Highsmith, A.K., & Juranek, D.D. (1993). Surveillance for waterborne disease outbreaks--United States, 1991-1992. Morbidity and Mortality Weekly Report, 42(SS-5), 1-35.

National Academy of Sciences. (1980). Drinking water and health, vol. 2. Washington, DC: National Academy Press.

Roefer, P.A., Monscvitz, J.J., & Rexing, D.J. (1996). The Las Vegas cryptosporidiosis outbreak. Journal of the American Water Works Association, 88(9), 95-106.

Sobsey, M.D. (1989). Inactivation of health-related microorganisms in water by disinfection processes. Water Science & Technology, 21(3), 179-183.

U.S. Environmental Protection Agency Water programs: National interim primary drinking water regulations. 40 Federal Register 59566-74 (1975) (codified at 40 C.F.R. ss 141).

U.S. Environmental Protection Agency Drinking water: National primary drinking water regulations; total coliforms (including feral califorms and E. coli); final rule. 54 Federal Register 27544-68 (1989a) (codified at 40 C.F.R., ss 141 and ss 142).

U.S. Environmental Protection Agency Drinking water: National primary drinking water regulations; filtration, disinfection; turbidity Giardia lamblia, viruses, Legionella, and heterotrophic bacteria; final rule. 54 Federal Register 27486-541 (1989b) (codified at 40 C.F.R. ss 141 and ss 142).

U.S. Environmental Protection Agency Drinking water: National primary drinking water regulations; total coliforms; corrections and technical amendments; final rule. 55 Federal Register 25064-5 (1990) (codified at 40 C.F.R. ss 141 and ss 142).

U.S. Environmental Protection Agency (1991a). Guidance manual for compliance with the filtration and disinfection requirements for public water systems using surface water sources (Contract No. 68-01-6989). Washington, DC: Author.

U.S. Environmental Protection Agency. Drinking water: Regulations; total coliforms. Maximum contaminant level goals and national primary drinking water regulations for lead and copper; final rule. 56 Federal Register 26460-4 (1991 lb) (codified at 40 CER. ss 141 and ss 42).

U.S. Environmental Protection Agency National primary drinking water regulations: Enhanced surface water treatment requirements; proposed rule. 59 Federal Register 38832-58 (1994) (codified at 40 C.F.R. ss 141 and ss 142).

U.S. Environmental Protection Agency (1999). Ground water rule deliberative document, November 23, Washington, DC: Author.

Corresponding Author: Gunther F Craun, Gunther F Craun & Associates, Suite 207, 101 West Frederick Street, Staunton, VA 24401. E-mail: <[e-mail]>.