Is contaminated groundwater an important cause of viral gastroenteritis in the United States? (Features).

Introduction

Groundwater may be subjected to fecal contamination from a variety of sources, including effluents of sewage treatment plants; discharges from onsite septic-waste treatment; runoff from urban, agricultural, and natural lands; and leachates from sanitary landfills, According to the 1990 U.S. Census, 25 million U.S. households use onsite wastewater treatment systems (U.S. Environmental Protection Agency [U.S. EPA], 2000). These households contribute one trillion gallons of septic-tank waste per year to the ground (Canter, 1984). Concerns about contamination of underground water supplies and human health risks have prompted a number of studies of virus occurrence in groundwater. The authors of this article reviewed those studies and the epidemiological evidence that enteric-virus contamination of groundwater is a likely source of human disease in the United States. In addition, the authors evaluated the feasibility of conducting studies to estimate the magnitude of endemic waterborne disease associated with viru s contamination of groundwater.

Review of Waterborne Outbreaks

Waterborne outbreaks continue to be reported in the United States. From 1991 to 1998, 126 waterborne outbreaks were reported in public and individual water systems in 39 states and two U.S. territories (Craun, Calderon, & Nwachuku, 2002). According to this passive reporting of outbreaks, water systems that use groundwater sources thought to be at low risk for waterborne-disease transmission accounted for 85 outbreaks, 74 of which occurred in public water systems. Community groundwater systems that served permanent resident populations accounted for only 22 of the 74 outbreaks. One of the 22 outbreaks was identified as caused by a viral agent, and in another outbreak, a viral agent was suspected but was not confirmed by laboratory analyses. Noncommunity water systems, which primarily serve transient populations, accounted for 52 of the 85 outbreaks reported in groundwater systems. The authors also evaluated the causes of waterborne outbreaks that have been reported to the Centers for Disease Control and Preven tion (CDC) and U.S. EPA since 1971 (data obtained from Rebecca Calderon, U.S. EPA, personal communication, 2000) (Table 1). It was found that only 42 enteric virus--related outbreaks were reported in groundwater systems over this 28-year period (Table 2). Only hepatitis A and Norwalk viruses were identified as etiological agents in these outbreaks (U.S. EPA, personal communication, 2000), and only 14 virus-related outbreaks occurred in community groundwater systems over the 28-year period (Table 3). Although these data suggest that human illnesses have not commonly been traced to the virus contamination of groundwater in public water systems, in a large number of the reported waterborne outbreaks, no etiological agent could be identified. It has been suggested that a significant fraction of these outbreaks may be caused by virus contamination of drinking water. In fact, one study (Kapikian, Estes, & Chanock, 1996) estimated that 25 percent of 96 waterborne outbreaks of unidentified etiology reported to CDC be tween 1975 and 1981 were caused by Norwalk virus. The probability that viral outbreaks will be detected is expected to be low in groundwater systems, since it is difficult to detect viruses either in infected humans or in drinking water near the time of an outbreak and disease-surveillance systems are relatively insensitive to increases in mild gastroenteritis. The authors do not know how many waterborne outbreaks go unrecognized and the extent to which viral outbreaks maybe underestimated in the United States (Frost, Craun, & Calderon, 1996).

Review of Enteric-Virus Epidemiology

A wide range of enteric viruses are capable of causing human illnesses, These viruses include rotaviruses, Norwalk and other caliciviruses, enteric adenoviruses, astroviruses, and coronaviruses. Rotavirus is highly contagious and is a major cause of childhood illness, hospitalizations, and even death. More than 90 percent of children develop antibody to rotavirus by the age of three years (Kapikian et al, 1975). In contrast, Norwalk virus and other caliciviruses are important causes of outbreaks in schools, families, hospitals, recreational camps, and cruise ships (O'Ryan et al., 2000). A high fraction of older children have antibodies to enteric calicivirus (O'Ryan et al., 1998). Enteric adenovirus and astrovirus are important causes of gastroenteritis in children but have limited pathogenicity for adults.

Enteroviruses and hepatitis A and E viruses are enteric viruses with a very wide range of health effects. Enteroviruses account for an estimated 10 to 15 million symptomatic infections in the United States each year (Strikas, Anderson, & Parker, 1986). There are currently 66 enterovirus serotypes than can cause over 20 clinically recognized syndromes (Oberste, Maher, Kilpatriack, & Pallansch, 1999). The syndromes include gastrointestinal symptoms, poliomyelitis, perinatal enterovirus disease, myocarditis, pericarditis, pleurodynia, respiratory illnesses, conjunctivitis, hepatitis, aseptic meningitis, encephalitis, hand-footand-mouth disease, and possibly diabetes (Morens & Pallansch, 1995).

Most enterovirus infections do not result in illness. Between 90 and 95 percent of poliovirus infections are likely to be asymptomatic, and approximately 50 percent of echovirus and coxsackievirus infections are asymptomatic (Melnick, 1997).

The reasons for the uncommonness of waterborne virus outbreaks in groundwater systems are unclear. Natural waters may inactivate the viruses. This phenomenon has been demonstrated for polioviruses and coxsackieviruses in Rio Grande River water (O'Brien & Newman, 1977). In addition, the infectious dose for certain enteric viruses may be relatively high (Schiff et al., 1984). If so, low levels of groundwater contamination by certain enteric viruses may not result in doses sufficient to cause illness. Alternatively, detection of infrequent illnesses from waterborne enteric viruses is likely to be difficult. For enteric viruses that are highly infectious, such as rotavirus and Norwalk virus, person-to-person spread is common, and the occurrence of low-level groundwater contamination by viruses may simply reflect the occurrence of a previous person-to-person outbreak rather than suggesting that drinking water was the source of transmission.

Potential for Virus Occurrence in Groundwater

Outside the human host, enteric viruses are dormant and remain infectious in soil and water for a longer time than can coliform bacteria. Enteric viruses are excreted in feces in very high numbers (e.g., [10.sub.10] to [10.sub.12] per gram of feces of infected individuals) (Flewett, 1982). Enteric viruses are commonly recovered from domestic wastewater, even after disinfection. Once in the environment, enteric viruses may survive for several months (Yates & Yates, 1987).

The distance a virus moves through soil depends upon the nature of the soil, the amount of soil moisture, soil pH, rainfall events, and the type of virus (Yates & Yates, 1987). Virus removal by soil is largely dependent upon absorption of the virus to the soil or subsurface matrix. Under optimal conditions, viruses have been observed to travel more than 100 meters through the subsurface (Keswick, 1982).

The different survival of viruses and bacteria in the environment leads to the prediction that commonly used bacterial indicators of water contamination are not useful for assessing viral contamination. Groundwater sources containing coliform bacteria are usually disinfected, and groundwater sources free of coliform bacteria are usually considered free from enteric bacterial pathogens. Sewage contamination adds high levels of coliform bacteria to the water, and coliform bacteria survive for a period similar to that survived by bacterial pathogens (Berg, 1973). Some enteric viruses, however, may travel further and survive longer in the environment than coliform bacteria, and waterborne viral illness may occur in the absence of detectable coliform bacteria (Berg, Clark, Berman, & Chang, 1967). This is a significant concern for public-health officials. The microbial safety of drinking water is usually ascertained by the presence or absence of coliform bacteria. As the amount of human sewage discharged into the g round increases, however, a negative coliform test may offer less assurance that groundwater is free of viral contamination. Even when groundwater is disinfected, increasing concentrations of human enteric viruses in groundwater may increase the risk of illness. Coliform bacteria may not serve as good indicators of fecal risks; studies indicate that aggregations of viruses and some types of viruses are more resistant to chlorine disinfection than are indicator coliform bacteria (Sobsey, 1983; Trask, Melnick, & Wenner, 1945; Young & Sharp, 1979).

Virus Occurrence in Groundwater

In early studies of groundwater virus occurrence, groundwater study sites were selected to maximize the chance of virus detection (Bagdasarjan et al., 1979; Bitton & Gerba, 1984; Hejkal et al., 1982; Mack, Yue-Shoung, & Coohan, 1972; Vaughn & Landry, 1983). At some locations, there was evidence of sewage contamination or enteric illness among water consumers. Those studies demonstrated that in groundwaters known to be contaminated, viruses can be detected when samples are collected in a timely manner and if the analyses are sensitive.

Several methods have been used to detect viruses in groundwater. Early methods employed cell culture techniques, which are expensive and time consuming. Cell culture requires the virus to infect a cell, replicate, and kill the cell. Therefore, a positive finding should indicate a high probability that there have been viable viruses in the drinking water. Unfortunately, false positives and false negatives occur. A particular cell line must be selected, but not all human enteric viruses will grow in one selected cell line. Some viruses cannot be cultivated. As a result, the occurrence of viruses will be underestimated. The water may also contain substances that are toxic to the selected cells, making the test result impossible to interpret. Since human enteroviruses are ubiquitous, laboratory contamination of cell cultures can easily occur.

Recent studies have used new molecular methods such as reverse-transcriptase polymerase chain reaction (RT-PCR), which decreases the time and cost of the analysis and increases detection. The PCR test detects only the specific viral DNA or RNA selected for testing. Detection of this genetic material does not mean that viable viruses are present in the water. Fragments of genetic material or the genetic material without the appropriate protein coating may not be infectious. RT-PCR also can be subject to laboratory contamination.

Lieberman and co-authors (1994) designed a national occurrence study to compare RT-PCR results with traditional tissue-culture findings and microbial indicators of water quality. Initially samples were collected from 96 potentially vulnerable wells used by public water systems in 22 states and two U.S. territories; 23 of the wells were selected for monthly sampling. A final report has not yet been published, but preliminary information indicates that only seven (7.3 percent) of the 96 wells were positive for human enteric viruses (U.S. EPA, 2000).

Other studies have been motivated by general water quality concerns rather than evidence that groundwater was contaminated by fecal material. The Missouri Groundwater Alluvial Study (Davis & Witt, 1998) was conducted to assess water quality in newly constructed and older community water systems in the Ozark Plateau region of Missouri. A total of 109 wells, mostly constructed in the past 15 years, were tested in the first study. Thirteen wells (12 percent) were found to be PCR-positive for human enteric viruses, and a number of these wells were also found to have bacterial evidence of contamination. Another study tested 106 older wells using cell culture methods but failed to detect that any were virus positive (U.S. EPA, 2000). A third study of Missouri wells (Davis & Witt, 1998) examined wells subject to flooding. Although total and fecal coliforms were detected in several samples tested, only one sample showed evidence of enteric virus contamination (U.S. EPA, 2000).

U.S. EPA tested 30 wells in eight states for enteric viruses (U.S. EPA, 2000). These wells were selected to be representative of 10 hydrogeological settings. Analyses included PCR and cell culture tests for enteroviruses and PCR tests for hepatitis A virus, rotavirus, and Norwalk virus. Only one well was found positive for hepatitis A virus, by PCR methods. Four wells were positive for total coliforms, and two were coliphage positive.

Several other regional enteric virus studies were conducted in the 1990s. On Oahu, Hawaii, 32 deep wells were tested for male-specific coliphage (U.S. EPA, 2000). None of the wells were positive. A California study tested monthly samples from 18 wells for enteric viruses using cell culture techniques. Six wells were positive for human enteric viruses. A virus occurrence study of 25 wells serving public water supplies in Wisconsin found three to be positive for human enteric viruses using cell culture (U.S. EPA, 2000).

The most comprehensive study (Abbaszadegan, Steward, & LeChevallier, 1999) was used by U.S. EPA to estimate the national occurrence of groundwater virus contamination. The study used both RT-PCR and cell culture methods to test wells from 35 states and from different hydrogeologic settings. Wells were selected at the following kinds of sites:

* groundwater sites with high concentrations of minerals, metals, or total organic carbon;

* sites at which any virus or bacteria had previously been detected in the groundwater source;

* sites with potential exposure to contaminants because of agricultural activities or septic tanks near the well; and

* sites with different pH values, temperatures, depths, production capabilities, and aquifer types.

The study team consulted with local utility officials and public-health officials to identify the specific wells to be tested. Wells were sometimes selected because of local concerns that a well might be contaminated with a virus or might be more likely to be contaminated. The study found that almost 5 percent of the wells were positive for enteric virus by cell culture, 15 percent were positive for bacterial indicators, 21 percent were positive for coliphage indicators, and 32 percent were positive by RT-PCR. The number of wells that tested negative for coliforms but were positive for virus was not reported. The number of wells not disinfected also was not reported. Thus, it is difficult to assess the number of people exposed to groundwater contaminated by enteric viruses and the potential health risks.

In summary, findings of the virus occurrence studies show that enteric viruses can be detected in groundwater sources used by public water systems. In many cases, however, the water systems were selected for testing because of known or suspected water quality problems. With the exception of some local studies (e.g., the Missouri and New England studies), available data likely overestimate the occurrence of groundwater virus contamination in the United States because high-risk water systems were surveyed. In addition, it is unclear whether wells found to be virus-positive pose a significant risk of disease or infection because information about disinfection and other treatment often is not reported.

Epidemiological Approaches and Difficulties

Studies to determine an association between illness and drinking water present a challenge for state and local public-health officials. Most states do not conduct active surveillance for gastroenteritis (Frost, Calderon, & Craun, 1995). Because viral gastroenteritis is usually a self-limiting disease of short duration, people seldom report their illness to medical practitioners. Even when reported, physicians seldom order tests to identify an enteric virus in the stool of a potential viral gastroenteritis case. The primary reason that stools are not often tested for enteric virus is that there is seldom a benefit to the patient.

The high fraction of asymptomatic infections for many enteroviruses also adds to the difficulty in identifying clusters and a common source, such as drinking water. Even if drinking water is contaminated, only people at risk and sufficiently exposed to contaminated water will become infected, and only a fraction of those people will experience an illness from the infection.

For viral agents that are highly infectious, such as rotavirus or Norwalk virus, an initial waterborne infection may result in many additional cases caused by person-to-person transmission. Because these clusters occur commonly and usually have few serious consequences, except for rotavirus infection in young children, identifying the mode of virus transmission is seldom a priority for state and local public-health officials. Household, school, daycare, or community clusters of illnesses are often assumed to result from person-to-person transmission.

An alternative to studying viral illness is studying viral infection through serological surveys of populations. Serological tests have the advantage of detecting persons who are asymptomatically infected, thus easily identifying those at risk. Serological surveys were essential for understanding the epidemiology of polioviruses, since only a small fraction of infected individuals became seriously ill (Melnick, 1997). By measuring antibody levels to the poliovirus at two points in time, it was possible to estimate both the size of the at-risk population (e.g., people not previously infected) and the number of infections during the interval between the two tests.

For other enteric viruses, however, serological surveys have been less useful in expanding the understanding of enteric-virus epidemiology. One problem is the large number of enteroviruses. There is no general serological screening tool to detect seroconversions against all enteric viruses. Such a test would require 66 different antigens just for the enteroviruses (Melnick, 1997). For selected enteroviruses, serological studies have been conducted, but the large number of different enteroviruses makes tracking patterns of infection both difficult and expensive. Serologic studies of rotavirus and the Norwalk-like viruses have revealed the age patterns of infection and, for the Norwalk-like viruses, found temporal differences and geographic variation in the specific strains of virus circulating in the population.

Cross-reactions between many enteroviruses can complicate the interpretation of serological tests. Since many enteroviruses are closely related, infection by one enterovirus can result in an apparent serological response to other, related viruses. Use of paired sera--that is, sera collected from the same person at different times--can prevent some of these problems, but paired sera are both difficult and expensive to obtain. Furthermore, since many of the infections are asymptomatic, knowing that a person was infected yields no medical benefit to the serum donor. Finally new regulations to protect patient privacy prohibit testing unidentified surplus sera unless the donor has explicitly given permission for the test.

To avoid conducting many serological tests to study the transmission of enteroviruses, it may be possible to combine a virus detection program, in which stools are examined for the presence of enteroviruses, with a serological study. The combination could be accomplished with family volunteers similar to those used in the 1960s for the Virus Watch Program. (Kogon et al., 1969). Routine groundwater monitoring and virus surveillance programs that detect enteric viruses in human stools would be needed in areas at high risk for groundwater contamination. Families should be followed as well, to detect the serological response to enteric virus detected in the stools. Information would be available about outbreaks of specific enteroviruses and the occurrence of groundwater contamination.

Discussion

The high concentration of enteric viruses in sewage, the large number of onsite sewage disposal systems in the United States, and the trillion gallons of sewage discharged into the ground each year all suggest that there is potential for enteric-virus contamination of groundwater. The results of current studies into enteric-virus occurrence in groundwater, however, have not provided evidence of widespread exposures or significant health risks. Since the studies have focused on high-risk wells, there is a need to estimate the occurrence of groundwater virus contamination in the absence of bacterial indicators of sewage contamination and in systems where groundwater is not adequately disinfected, Current epidemiological studies have not identified populations at high risk for infection from virus-contaminated groundwater. Although serological studies for selected enteric viruses have been conducted successfully outside the United States and risk factors for infection have been identified (O'Ryan et al., 1998, 2 000), it has become increasingly difficult and expensive to conduct these studies in the United States. The requirement that informed consent must be obtained for testing of surplus sera increases privacy protection but also substantially increases the costs of conducting serological studies. The extensive safety testing for donated blood uses so much blood that there are rarely any surplus sera available, even if informed consent to conduct the tests is obtained. Future serological studies will likely require considerably more expense and time than prior studies. Nevertheless, accurate epidemiological information on the risk of groundwater transmitting enteric viruses would be valuable because it could prevent costly under- or over-reaction to the risks from these pathogens.

TABLE 1

Etiology of Reported Waterborne Infectious-Disease Outbreaks, 1971-1998

Etiology                                 Outbreaks
                          Drinking-Water  Recreational  Total
                            Systems *      Waters **

Acute gastroenteritis of       347             32        379
 undetermined etiology
Giardia                        119             15        134
Shigella                        43             31        74
Cryptosporidium                 13             24        37
Hepatitis A                     28             2         30
Norwalk virus                   22             7         29
E. coli 0157                    8              11        19
Campylobacter                   17             0         17
Naegleria                       0              17        17
Salmonella                      13             1         14
Schistosoma                     0              7          7
Typhoid                         7              0          7
Leptospiria                     0              5          5
Pseudomonas                     0              5          5
V. cholerae non-01              2              0          2
Yersinia                        2              0          2
Adneovirus                      0              2          2
Enterovirus                     0              2          2
Rotovirus                       1              0          1
Cyclospora                      1              0          1
E. Histolytica                  1              0          1
P. shigelloides                 1              0          1
Total                          625            161        786

* Community, noncommunity, and individual water systems.

** Lakes, rivers, and swimming pools.
TABLE 2

Enteric Virus-Related Waterborne Outbreaks Reported in the United
States, 1971-1998, by Water Source

Etiology                            Outbreaks
                 Groundwater  Surface-Water  All Drinking-Water
                   Systems       Systems          Systems

Hepatitis A          25             3                28
Norwalk viruses      17             5                22
Rotavirus             0             1                1
Total                42             9                52
TABLE 3

Enteric Virus-Related Waterborne Outbreaks Reported in U.S. Groundwater
Systems, 1971-1998, by Type of Water System

Etiology                           Outbreaks
                   Community    Noncommunity    Individual
                 Water Systems  Water Systems  Water Systems

Hepatitis A            9              9              7
Norwalk viruses        5             12              0
Total                 14             21              7

Acknowledgements: The study was funded by the American Water Works Association Research Foundation, Contract #2505.

REFERENCES

Abbaszadegan, M.P.W., Steward, P.W., & LeChevallier, M.W. (1999). A strategy of detection of viruses in groundwater by PCR. Applied and Environmental Microbiology, 65(2), 444-449.

Bagdasarjan, G.A., Berg, G., Yunte, H., Ren, K., Lund, E., Melnick, J.L., Rao, V.C., Shuval, H.I., Slade, J.S., & Hui, S.E. (1979). Human viruses in water, wastewater and soil. Report from a meeting of the WHO Scientific Group on Human Viruses in Water, Wastewater and Soil, Geneva, 23-27 October, 1978. World Health Organization technical report series (Report #639, pp. 5-50). Geneva, Switzerland: World Health Organization.

Berg, G. (1973). Reassessment of the virus problem in sewage and in surface and renovated water. Progress in Water Technology, 3(2), 87-94.

Berg, G., Clark, R.M., Berman, D., & Chang, L. (1967). Aberrations in survival curves. In G. Berg (Ed.), Transmission of viruses by the water route (pp. 235-240). New York: Interscience Publishers.

Bitton, G., & Gerba, C.P. (1984). Groundwater pollution microbiology. New York: John Wiley & Sons.

Canter, L., & Knox, R.C. (1984). Evaluation of septic tank system effects on groundwater quality (No. EPA-600/2-84-107). Ada, OK:

Robert S. Kerr Environmental Research Laboratory, U.S. Environmental Protection Agency.

Craun, G.F., & Calderon, R.L., & Nwachuku, N. (2002). Causes of reported waterborne outbreaks in the United States, 1991-1998. In P.R. Hunter, M. Waite, & E. Ronchi (Eds.), Drinking water and infectious disease: Establishing the links. Boca Raton, FL: CRC Press.

Davis, J.V, & Witt, E.C. (1998). Microbiological quality of public water supplies in the Ozark Plateaus aquifer system Missouri (USGS Fact Sheet 028-98). Rolla, MO: U.S. Department of Interior/U.S. Geological Survey.

Flewett, T.H. (1982). Clinical features of rotavirus infections. In D.A. Tyrell & A.Z. Kapikian (Eds.) Virus infections of the gastrointestinal tract (pp. 125-146). New York, NY: Marcel-Kekker, Inc.

Frost, F.J., Calderon, R.L., & Craun, G.F. (1995). Waterborne disease surveillance: Findings of a survey of state and territorial epidemiology programs. Journal of Environmental Health, 58(5), 6-11.

Frost, F.J., Craun, G.F., & Calderon, R.L. (1996). Waterborne disease surveillance. Journal of the American Water Works Association, 88(9), 66-76.

Hejkal, T.W., Keswick, B., LaBelle, R.L., Gerba, C.P., Sanchez, Y., Dressman, G., Hafkin, B., & Melnick, J.L. (1982). Viruses in a community water supply associated with an outbreak of gastroenteritis and infectious hepatitis. Journal of the American Water Works Association, 74, 318-321.

Kapikian, A.Z., Cline, W.L., Mebus, C.A., Wyatt, R.G., Kalica, A.R., Chanock, R.M., & Kim, H.W. (1975). New complement-fixation test for the human Reovirus-like agent of infantile gastroenteritis. Lancet, 1(7915), 1056-1061.

Kapikian, A.Z., Estes, M.K., & Chanock, R,M. (1996). Norwalk group of viruses. In B.N. Fields, D.M. Knipe, & P.M. Howley (Eds.) Fields Virology (3rd ed., pp. 315-327). Philadelphia, PA: Lippincott.

Keswick, B.H. (1982). The use of microorganisms as groundwater tracers: A review. Groundwater, 20(2), 142-149.

Kogon, A., Spigland, I., Frothingham, T.E., Elveback, L., Williams, C., Hall, C.E., & Fox, J.P. (1969). The virus watch program: A continuing surveillance of viral infections in metropolitan New York families. VII: Observations on viral excretion, seroimmunity, intrafamilial spread and illness associated in coxsackie and echovirus infections. American Journal of Epidemiology, 89(1), 51-61.

Lieberman, R.J., Shadix, L.C., Newport, B.S., Frebis, M.W.N., Moyer, S.E., Safferman, R.S., Stetler, R.E., Lye, D., Fout, G.S., & Dahling, D. (1994). Source water microbial quality of some vulnerable public groundwater supplies. In Proceedings, Water Quality Technology Conference, San Francisco, CA, October 1994 [CD-ROM]. Denver, CO: American Water Works Association.

Mack, W.N., Yue-Shoung, L., & Coohon, D.B. (1972). Isolation of poliomyelitis virus from a contaminated well. Health Services Report, 87(3), 271-274.

Melnick, J.L. (1997). Poliovirus and other enteroviruses in viral infections of humans. In A.S. Evans & R.A. Kaslow (Eds.) Viral infections in humans (4th ed., pp. 583-657). London: Plenum Medical Blood Company.

Morens, D.M., & Pallansch, M.A. (1995). Epidemiology In H.A. Rotbard (Ed.) Human enterovirus infections (pp. 3-24). Washington, DC: American Society for Microbiology Press.

O'Brien, T.R., & Newman, J.S. (1977). Inactivation of polioviruses and coxsackieviruses in surface water. Applied and Environmental Microbiology, 33(2), 334-340.

O'Ryan, M.L., Mamani, N., Gaggero, A., Avendano, L.F., Prieto, S., Pena, A., Jiang, X., & Matson, D.O. (2000). Human caliciviruses are a significant pathogen of acute sporadic diarrhea in children of Santiago, Chile. Journal of Infectious Diseases, 182(5), 1519-1522.

O'Ryan, M.L., Vial, P.A., Mamani, N., Jiang, X., Estes, M.K., Ferrecio, C., Lakkis, H., & Matson, D.O. (1998). Seroprevalence of Norwalk virus and Mexico virus in Chilean individuals: Assessment of independent risk factors for antibody acquisition. Clinical Infectious Diseases, 27(27), 789-795.

Oberste, M.S., Maher, K., Kilpatrick, D.R., & Pallansch, M.S. (1999). Molecular evolution of human enteroviruses: Correlation of serotypes with VP1 sequence and application to picornavirus classification. Journal of Virology, 73(3), 1941-1948.

Schiff, G.M., Stefanovic, G.M., Young, E.C., Sander, D.S., Pennekamp, J.K., & Ward, R.L. (1984). Studies of echovirus-12 in volunteers: Determination of minimal infectious dose and the effect of previous infection on infectious dose. Journal of Infectious Diseases, 150(6), 858-866.

Sobsey, M.D. (1983). Transport and Fate of viruses in soils. In L.W. Canter, E.W. Akin, J.F. Kreissel, & J.F. McNabb (Eds.) Microbiological health considerations of soil disposal of domestic wastewaters (USEPA 600/9-83-017, pp. 283-285). Cincinnati, OH: U.S. Environmental Protection Agency.

Strikas, R.A., Anderson, L., & Parker, R.A. (1986). Temporal and geographic patterns of isolates of non-polio enteroviruses in the United States, 1970-1983. Journal of Infectious Diseases, 153(2), 346-351.

Trask, J.D., Melnick, J.L., & Wenner, H.A. (1945). Chlorination of human, monkey-adapted and mouse strains of poliomyelitis virus. American Journal of Hygiene. 41(1), 30-40.

U.S. Environmental Protection Agency: Proposed rules, national primary drinking water regulations: Ground water systems; waterborne pathogens from fecal contamination; public health risk reduction (00-11675), 65(91) Federal Register 30193-30274 (2000) (codified at 40 C.F.R. [section] 141, 142).

Vaughn, J.M., & Landry E.F. (1983). Viruses in soils and groundwater. In G. Berg (Ed.) Viral pollution of the environment (pp. 177-187). Boca Raton, FL: CRC Press.

Yates, M.V., & Yates, S.R. (1987). A comparison of geostatistical methods for estimating virus inactivation rates in groundwater. Water Research, 21(9), 1119-1125.

Young, D.C., & Sharp, D.G. (1979). Partial reactivation of chlorine-treated echovirus. Applied and Environmental Microbiology, 37(4), 766-773.

Corresponding Author: Floyd J. Frost, Epidemiologist, Lovelace Respiratory Research Institute, 2425 Ridgecrest Drive, SE, Albuquerque, NM 87108. E-mail: <[e-mail]>.