Water: Drinking Water Standards
GROUND WATER RULE: Public Health Concerns
EPA has the responsibility to develop a ground water rule which not only specifies the appropriate use of disinfection but, just as important, addresses other components of ground water systems to assure public health protection. Section 1412(b)(1)(A) of the Safe Drinking Water Act (SDWA) requires EPA to establish National Primary Drinking Water Regulations for contaminants that may have an adverse public health effect and that present a meaningful opportunity for health risk reduction. This general provision is supplemented with an additional requirement under Section 1412(b)(8) that EPA also develop regulations specifying the use of disinfectants for ground water systems as necessary. To meet these requirements EPA is working with stakeholders to develop a Ground Water Rule (GWR) proposal by March 1999, and a final rule by November 2000.
Assurance that the drinking water is not contaminated by human or animal fecal waste is the key issue for any drinking water system. Human waste is the source of pathogenic viruses and bacteria; animal waste is a source of some bacterial pathogens. Fecally contaminated waters are likely to contain enteric pathogens, which can cause gastrointestinal illness (diarrhea, fatigue, cramps, etc.), hepatitis A, meningitis, ulcers, myocarditis, typhoid fever, cholera, and a number of other diseases. These diseases can manifest themselves in the community at either background level of illnesses (referred to as endemic illnesses) or in outbreaks affecting several people within a short time period. The pathogens of concern for the GWR are viruses and bacteria. (The presence of protozoan pathogens indicates that the drinking water source is under the direct influence of surface water and therefore not regulated by the GWR). Preliminary data from 8 ground water microbial studies of Public Water Supply (PWS) system wells indicate pathogen occurrence rates from 6% to 12% and pathogen indicator occurrence rates of 12% to 38%. Recent data from the Safe Drinking Water Information Systems database for the period of 1993 to 1997 indicates ground water systems account for the vast majority of all total coliform MCL violations.
Significant outbreaks are investigated by the Centers for Disease Control and Prevention (CDC) with EPA assistance and reported in detail in the CDC publication, Morbidity and Mortality Weekly Report (MMWR). Other outbreaks are summarized as reported to CDR in MMWR but are not independently investigated by CDC or EPA. Information on bacterial and viral pathogens and illnesses comes primarily from textbooks supplemented by recent scientific papers. Information on TCR MCL violations comes from the EPA Safe Drinking Water Information Systems (SDWIS) database.
DATA SUMMARY AND ANALYSIS
Outbreaks in Ground Water Systems
The CDC maintains a database of waterborne and foodborne disease outbreaks in the United States (US). The database is compiled based upon responses to a voluntary and confidential survey form that is completed by some state and local public health officials. CDC and EPA believe that many disease outbreaks are not reported to the CDC and, therefore, are under represented by the survey. Craun (1996) reports that outbreaks are not recognized unless approximately 1% or more of the population becomes ill.
For the period 1971-1994, 356 outbreaks have occurred associated with ground water source or distribution system water in Public Water Supply (PWS) system wells (Craun, unpublished report, 1996, see attached tables). Of these, at least 158 (44%) occurred in untreated groundwater systems. Fifty (of 356) outbreaks resulted from pathogen entry into the distribution system, in either treated, inadequately treated or untreated systems. About twice as many of those outbreaks occurred in Non-Community systems as in Community systems. Almost all (89%) outbreaks in Non-community systems were reported to be the result of source water contamination. More outbreaks occurred in Community systems (31%) than in Non-Community systems (11%) as a result of pathogen entry into the distribution system. Most outbreak etiologic agents were unidentified. Agents were identified in only 106 of 329 (32%) outbreaks. Hepatitis A (virus or illness) was identified in 17 outbreaks and the Norwalk virus was identified in 16 gastrointestinal illness outbreaks. All other outbreaks with recognized etiology were caused by bacterial pathogens. However, most outbreaks are the result of unidentified organisms, commonly assumed to be viruses (e.g. Gerba and Rose, 1990).
For the most recent reporting period 1993-1994, 10 (1568 illnesses, 7 deaths) outbreaks were associated with untreated ground water from Public Water Supply wells. Of these, 5 (50%) outbreaks (474 illnesses) resulted from contaminated source water in undisinfected systems and 5 (50%) resulted from a distribution failure including inadequate treatment. Etiologic agents were identified in 5 (50%) outbreaks. All were bacterial pathogens.
In the period after 1994, at least one source water outbreak associated with sewage contamination has been studied and reported. A total of 82 cases of shigellosis were identified resulting from source water contamination of a well at a resort in Island Park, Idaho in 1995 (MMWR, 1996). Nearby septic tanks or a leaky sewer line are the possible sources of contamination.
It is rare that investigations identify deaths associated with a gastrointestinal illness outbreak because mortality rates for waterborne agents are typically around 1 death per thousand illnesses or fewer. However two recent outbreaks in Cabool and Gideon, Missouri are notable because the outbreak investigation identified mortality associated with the outbreak. In 1989, 4 deaths (243 illnesses) occurred in Cabool as a result of distribution system contamination by E. coli 0157:H7 (Swerdlow et al., 1992). In 1993, 7 deaths (650 illnesses) occurred in Gideon as a result of distribution system contamination by Salmonella typhimurium. (Angulo, 1993).
The largest documented ground water outbreak occurred in Georgetown, Texas in 1980 (Hejkal et al., 1982). About 7900 people became ill. Both Coxsackievirus and Hepatitis A virus were found in the raw well water and the outbreak was the result of source water contamination. The contamination source was never identified but the site is within an area of karst (solution-enhanced) limestone with cavernous zones which allows contamination to travel long distances in short periods with relatively little natural attenuation. In Racine, Missouri, 46 cases of hepatitis A were identified in 1992 at a non-community water supply in a karst limestone setting. A nearby septic system was identified as the contaminant source. In Lancaster County, Pennsylvania, 49 cases of hepatitis A occurred in 1980 as the result of septic tank contamination of a well in a karst limestone area. An outbreak resulting in 1200 cases of shigellosis occurred in 1974 (Wiessman, 1976) as the result of septic tank contamination of source water in a likely karst area in Florida. Craun (1984) reports a total of five outbreaks associated with untreated groundwater in karst or fractured rock areas for the years 1971-1979, including outbreaks in Missouri and Arkansas.
An outbreak at a Sedona, Arizona resort in 1990 resulted in about 900 cases of illness due to Norwalk virus (Lawson et al., 1991). Source water contamination was identified despite the fact that the depth to water was about 400-600 feet.. A sewage treatment plant failure resulted in discharge of inadequately treated sewage to the soil. A thin soil and a fractured rock hydrogeologic setting allowed pathogens to reach the well intake depth quickly and with minimal attenuation. Thin soils and fractured rock were also in part responsible for a 39 case outbreak at a northern Michigan resort island in 1991. An inadequate septic system was identified as the source that contaminated the raw water of the public water supply well.
A septic tank sited uphill from a well was the cause of 135 cases of illness due to Norwalk virus in 1986 at a South Dakota campground (MMWR, 1988).
Bacterial pathogens may be divided into two broad groups; those that result in illness in most individuals and those that are opportunistic pathogens, only causing illness in sensitive sub-populations. Opportunistic bacterial pathogens are believed to associated with biofilm growth in water system distribution lines. Under certain optimal conditions it may be possible for these pathogens to thrive and reproduce within a pipe. All other pathogens, including opportunistic pathogens, require an entry into the water system, either as the result of source water contamination or as the result of direct entry through a failure in the distribution system.
Several bacterial pathogens such as E. coli, Shigella sonnei, and Campylobacter jejuni produce toxins that can result in bloody diarrhea and abdominal cramps. E. coli 0157:H7 can also cause severe illness or death in children as the result of hemolytic uremic syndrome (kidney failure). A waterborne outbreak in Cabool Missouri in 1989 resulted in 2 cases of hemolytic uremic syndrome (Swerdlow et al., 1992) . Shigella sonnei, Shigella dysenteriae and Shigella flexneri are the most common sources of bacillary dysentery in the United States (MMWR, 1996). Campylobacter jejuni is another common bacterial pathogen, causing fever, abdominal pain and diarrhea, often with bloody stools. Recently, C. jejuni was recognized as a possible cause of Guillain-Barré paralysis (5% mortality rate, Havelaar et al, in press). Salmonella typhimurium is a common waterborne bacteria that causes gastroenteritis. A recent S. typhimurium outbreak in Gideon, Missouri resulted in seven deaths at a nursing home (Angulo et al., 1997). S. typhi causes typhoid fever and Vibrio cholerae causes cholera but are rare in the US.
Campylobacter, E. coli and Salmonella have animal as well as human reservoirs; Shigella is associated only with humans (Gelderich, 1996). Runoff contaminated by animal waste may enter groundwater by natural or induced infiltration, especially in arid regions without well developed soils. Campylobacter, Yersinia, Salmonella and Shigella are associated with reactive arthritides and Reiter's syndrome (Rose et al., 1996). Yersinia enterocolitica has caused waterborne outbreaks, but like Klebsiella sp. and Listeria sp, is more commonly a food-borne pathogen. Dose-response data from human feeding studies are available for Campylobacter jejuni, Salmonella typhi, Shigella flexneri, and Vibrio cholerae.
The most common waterborne bacterial pathogens are the opportunistic pathogens because they may enter the water system (see microbial occurrence fact sheet) or may be found in situ. The most important is the pneumonia and diarrhea-causing Pseudomonas aeruginosa (Rusin et al., 1997). Colonization of this organism in healthy individuals ranges from 2.6%-24%. Aeromonas hydrophila is a common diarrhea causing opportunistic bacterial pathogen and has been isolated from 0.7% of healthy individuals. Legionella pneumophila is the cause of Pontiac fever, Legionnaires disease and pneumonia but only the genus, not the species has been found in groundwater (Lye, 1997).
There are more than 100 viruses from 13 families that are capable of transmission in drinking water. Viruses and bacteria are easily transmitted by the fecal-oral pathway. Some common fecal-oral viruses are also transmitted by person-to-person spread, by respiratory inhalation and contact with infected surfaces resulting in a secondary spread of infection after initial infection via drinking water. Illness from these viral pathogens ranges from gastroenteritis caused by many viruses including Rota and Norwalk virus to meningitis caused by Echovirus to myocarditis caused by Coxsackie B.
Ground water samples during outbreaks from drinking water wells have been identified as containing Hepatitis A, Polio, Coxsackie B, Echo, Norwalk, SRSV (in Canada) and Reovirus. Routine sampling for occurrence of viruses in ground water has yielded Coxsackie B, Rota, Echo and Reovirus. Cell culture methods are available to identify Polio, Coxsackie A and B, Echo, and Reovirus. More specialized cell culture or other methods are necessary to identify all possible viruses in ground water. Dose-response data from human feeding studies are available for three polio strains, echo and rotavirus. Enteric Adeno and Hepatitis E virus have only been identified with outbreaks in association with non-US groundwater wells, but have been recognized as causing illness by exposure in the US, but the pathway by which infection occured is not yet identified. Hepatitis E outbreaks of 100,000 cases have occurred in China and, for pregnant women, the mortality rate in undeveloped countries can approach 20%.
The percentage of individuals with detectable antibodies indicating infection has been reported from Montreal, Canada (Payment et al., 1991): by age 19, 100% of the population had rotavirus antibodies, 44%-58% had coxsackievirus antibodies, and 40%-90% had echovirus antibodies. By age 60 or over, 36% had antibodies indicating exposure to Hepatitis A virus.
TCR MCL Violations
Since the effective date of the Total Coliform Rule (1993-1997), 90% of TCR MCL violations occured in groundwater systems. A small percentage of systems (15% of all groundwater systems) are responsible for most of these TCR MCL violations.
The next CDC two-year outbreak summary (1995-1996) is in preparation and should be available by the end of 1998. EPA will continue to evaluate microbial data as it comes available.