Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
111 Microbiology of Drinking Water The principal microbiological contaminants found in drinking water of the United States are bacteria, viruses, and pathogenic protozoa. Each is considered in a separate section of this chapter. Helminths are included along with the protozoa. Little information is available on mycoplasma, pathogenic yeast, and pathogenic fungi in drinking water. Microbiologi- cal contaminants, such as fungi and algae, do not seem to be important causes of waterborne disease, although they are sometimes associated with undesirable tastes and odors. EPIDEMIOLOGY The average annual number of waterborne-disease outbreaks in the United States reported since 1938 is shown in Figure III-1 (Center for Disease Control, 1976b). There was a decrease in the number of outbreaks during the late 1930's and 1940's, but this trend was reversed in the early 1950's. There has been a pronounced increase in the outbreaks reported by the Center for Disease Control (CDC) in Atlanta, Georgia, since 1971. The reason for this apparent increase is not entirely clear, but it could be either the result of improved reporting or an overloading of our treatment plants with source water of increasingly lower quality. Since 1971, the CDC, the Environmental Protection Agency (EPA), state epidemiolo- gists, and engineers in state water-supply surveillance agencies have cooperated in the annual reporting of outbreaks. The purposes of such 63
64 DRINKING WATER AND H"LTH 50 _ CC m 40 in ~ 30 of LU 20 I: I: - o 1938- 1941- 1946- 1951 - 1956- 1961 - 1966- 197 1 1940 1945 1950 1955 1960 1965 1970 1974 YEARS FIGURE III-l Average annual number of waterborne disease outbreaks, 1938-1975. reports are to control disease by identifying contaminated water sources and purifying them, and to increase knowledge of disease causation. The roles of many microbial agents, including, for example, Yersinia enterocolitica and mycoplasma, remain to be clarified. The most important waterborne infectious diseases that occurred in 1971-1974 are listed in Table III-1. The etiologic agent was determined in only 53% of 99 disease outbreaks that involved 16,950 cases (Craun et al. 1976~. The remainder were characterized as "acute gastrointestinal illness of unknown etiology." Shigellosis was the most commonly identified bacterial disease (2,747 cases) in 1971-1974. Most of the cases were associated with non-municipal water systems. Four typhoid fever outbreaks affected 222 people and involved semipublic and individual water systems. In 1974, 28 waterborne-disease outbreaks, comprising 8,413 cases, were reported to the Center for Disease Control (1976a). The largest was an outbreak of giardiasis that occurred in Rome, N.Y., with an estimated 4,800 cases. The second largest involved about 1,200 cases caused by Shigella sonnet. In the third largest, which involved 615 cases of acute gastrointestinal illness, the etiologic agent was not definitely determined, but Yersinia enterocolitica was suspected. The fourth largest was caused by Shigella sonnet and involved 600 persons. Nineteen states reported at
Microbiology of Drinking Water 65 least one outbreak. Craun et al. (1976) stated that "this probably reflects the level of interest in investigating and reporting in different states rather than the true magnitude of the problem within the state." Semipublic water systems were associated with 55% of the outbreaks and accounted for 32% of the total cases in 1971-1974. Municipal systems accounted for 31% of the outbreaks, but 67% of the cases. Individual systems accounted for 14% of the outbreaks and only 1% of the cases, but outbreaks associated with individual systems probably are under-report- ed, as opposed to those associated with municipal and semipublic systems. Deficiencies in treatment and contamination of groundwater were responsible for a majority of the outbreaks (onto) and cases (onto) in 1971- 1974. Inadequate or interrupted chlorination was involved in 31% of the outbreaks and 44% of the cases. Craun et al. (1976) have drawn attention to the large number of waterborne disease outbreaks involving travelers. In 1971-1974, 49 (onto) of the 68 outbreaks that occurred in connection with semipublic and individual systems affected travelers, campers, visitors to recreational areas, or restaurant patrons; and 86% of the 49 outbreaks occurred during April-September. Outbreaks on cruise ships are excluded from the above tabulations, but they are of interest and should be mentioned because they involve the traveling public. For example, in June 1973, about 90% of 655 passengers and 35% of 299 crew were affected by an outbreak of acute gastroenteri- tis. An epidemiological investigation identified Shigella flexneri type 6 among early cases, and contaminated water and ice aboard the ship were implicated as vehicles of transmission (Center for Disease Control, 19731. In 1975, outbreaks of diarrhea on 8 ships affected between 9% and 61% of the passengers. In most of these outbreaks the causal agents and vehicles TABLE III-1 Etiology Of Waterborne Outbreaks and Cases, 1971-1974 Disease OutbreaksCases Gastroenteritis 467,992 Giardiasis 125,127 Shigellosis 132,747 Chemical poisoning 9474 Hepatitis-A 13351 Typhoid fever 4222 Salmonellosis 237 TOTAL 9916,950
66 DRINKING WATER AND H"LTH of transmission were unknown; water was identified as the vehicle in one of them (Center for Disease Control, 1976b). In 1975, 24 waterborne disease outbreaks involving 10,879 cases were reported to the Center for Disease Control (1976b). No etiologic agent was found for the two largest outbreaks (Sewickley, Pa. 5,000 cases and Sellersburg, Ind. 1,400 cases). The third largest outbreak, involving over 1,000 persons, occurred at Crater Lake National Park, Oreg. Enterotoxi- genic E. cold was isolated from residents of the park who became ill, and from the park's water supply. Seventeen of the 24 outbreaks and about 9070 of the cases reported to CDC were designated as "acute gastrointestinal illness." This category includes cases characterized by gastrointestinal symptoms for which no specific etiologic agent was identified. Cases resulting from water treatment deficiencies (2,695) or deficiencies in the water distribution system (6,961) accounted for almost 89% of the total cases in 1975. As in the past, most of the cases occurred in the spring and summer. The reported numbers of outbreaks and illnesses represent only a portion of the true totals. Craun et al. (1976) called attention to the outbreak at Richmond Heights, Fla., in 1974 as an example of why good disease surveillance is necessary and of the way in which many illnesses may go unnoticed. Initially, only 10 cases of shigellosis in this outbreak were recognized by authorities. An epidemiologic investigation revealed that approximately 1,200 illnesses actually occurred. This large outbreak might not have been detected if local health authorities had not been conducting shigellosis surveillance. In another outbreak, some 1,400 residents of Sellersburg, Ind. (31% of the town's population) experienced gastroenteritis. The high attack rate, rapid onset of the outbreak, review of water sampling data, and the town-wide survey suggested that the illness was waterborne, but no bacterial or viral pathogens or chemical toxins were found in the town water supply. Until improved detection and reporting systems are in use, the available epidemiological data will represent only a small fraction of the waterborne-disease problems in this country. BACTERIA The principal bacterial agents* that have been shown to cause human intestinal disease associated with drinking water are: Salmonella typhi, *Nomenclature in this report follows the 8th edition of Bergey's Manual of Determinative Bacteriology (Buchanan and Gibbons, 1974).
Microbiology of Drinking Water 67 typhoid fever; Salmonella paratyphi-A, paratyphoid fever; Salmonella (other species and a great number of serotypes), salmonellosis, enteric fever; Shigella dysenteriae, S. pexneri, and S. sonnet, bacillary dysentery; Vibrio cholerae, cholera; Leptospira sp., leptospirosis; Yersinia enterocoli- tica, gastroenteritis; Francisella tularensis, tularemia; Escherichia cold (specific enteropathogenic strains), gastroenteritis; and Pseudomonas aeruginosa, various infections. Several other organisms have been associated with gastroenteritis, such as those in other genera of the Enterobacteriaceae: Edwardsiella, Proteus, Serratia, and Bacillus. Number of Cells Required to Infect In attempting to assess the hazards in drinking water, it is important to know how many viable pathogenic cells are necessary to initiate an infection. McCullough and Eisele (19Sla,b,c,d) found that a dose of 106-108 salmonellae per person was necessary for most strains, although 105 cells of some strains could infect. More recent studies by Dupont, Hornick, and associates on selected enteric bacterial pathogens are summarized in Table III-2. Some enteric pathogens are highly virulent, causing infection when relatively few cells are administered (e.g., Shigellaflexneri and S. dysenteriae), whereas others require large numbers to infect (e.g., Salmonella typhosa and Vibrio cholerae). Virulence is a genetic trait and can vary markedly from strain to strain (Meynell, 19611. Phenotypic variation in virulence can occur within a given clone. A small percentage of the cells in a population may be unusually virulent (Meynell, 1961; Meynell and Meynell, 1965~. Thus, it does not always follow that because large numbers of cells are required for infection in feeding trials, that large numbers in drinking water are necessary to cause infection. Some few individuals may become infected by small numbers of unusually virulent cells. Recent evidence also indicates the possibility of genetic transfer of virulence from invading microbes into the resident intestinal population, providing another me.ans by which small numbers of organisms might initiate a disease state. The consequences of an increasing prevalence in livestock and their excrete of coliform organisms containing infectious plasmids and giving rise to clinical conditions were not examined in detail because of time constraints and their lack of immediate relevance to standard setting. Similarly, the consequences of adding antibiotics to animal and poultry feed and the enhanced hazards-of spreading drug-resistant organisms were not examined. The infecting dose also varies with the age and general health of the
68 DRINKING WATER AND H"LTH TABLE III-2 Infective Doses For Man of Bacterial Enteric Pathogens Subjects Infected/Total Tested Enters Pathogen Dose: Viable Cells 10i 102 103 104 105 106 107 108 109 Shigella dysenteriae Strain M131 1/10 2/4 7/10 5/6 Strain A-1 1/4 2/6 Shigella flexr~eri Seam 2A# 6/33a 33/49b66/87 15/24 Strain 2A# # 1/4 3/4 7/8 13/19 7/8 Salmonella typhi Strain Quailes 0/14 32/116 16/32 8/9 40/42 Vibrio cholerae Strain Inaba With NaHCO3 11/13 45/52 2/2 No NaHCO3 0/2 0/4 0/4 2/4 1/2 Enteropathogen~c E. cold Strain 4608 0/5 0/5 4/8 SOURCES: Shigella dysenteriae: Levine et al., 1973; Shigella flexneri: Dupont et al., 1972b; Dupont et al., 1969; Salmonella typhi: Cornice et al.. 1970; Vibrio cholerae: Cash et al., 1974; Enteropathogen~c E. coli: Dupont et al., 1971. aD.ose 1.8 x 102. bDose: 5 x 103. host population (MacKenzie and Livingstone, 1968~. Infants and the aged may be particularly susceptible. Previous exposure to a given pathogen is important, in that coproantibodies may prevent infection with a strain that is generally present in the population, whereas a new serotype introduced into the water supply may present an increased hazard. Not all strains of Shigella are highly virulent. Shaughnessy et al. (1946) determined infecting doses of four strains of Shigella while studying immunization in volunteers. They found that infectivity in mice could not be directly correlated with infectivity in humans and that doses of 109 organisms or higher `.vere needed to produce human infection. In their extensive studies to develop a Shigella vaccine, Hornick, DuPont, and associates observed the infective dose for several strains. With S. flexneri 2A, 30 of 39 volunteers became ill from a dose of 105-108 organisms
Microbiology of Drinking Water 69 (DuPont et al., 1969~. They showed that Shigella must penetrate the intestinal mucosa to produce symptoms of classic dysentery and that addition of bicarbonate facilitated this process. Two vaccine strains of S. exneri 2A, a hybrid of a Shigella mutant, E. colt, and a streptomycin- dependent strain, could be safely administered orally in doses of 10~° organisms or higher (DuPont et al., 1972a). A virulent strain could cause symptoms in doses of as few as 180 organisms (DuPont et al., 1972b). With Shigella dysenteriae 1 (Shiga strain)-an organism that has two pathogenic modes, invasiveness and enterotoxin elaboration the infecting dose in man was shown to be as low as 10 organisms (Levine et al., 19731. With such high infectivity of Shigella, why are waterborne outbreaks not more common? One possibility is that Shigella survives poorly in water. Wang et al. (1956) pointed out that, in a number of bacillary dysentery outbreaks involving water, the organism was not, or could not be, isolated. Over several years of studying irrigation water in Colorado, Wang et al. (1956) and Dunlop et al. (1952) were not successful in isolating shigellae, although salmonellae were frequently isolated. The survival of shigellae in water appears to be shorter than that of many other bacteria; Dolivo-Dobrovolskiy and Rossovskaya (1956) found Shigella survival times of only 0.5-4.0 h during the warmest time of the year. However, enteric pathogens may survive much longer times in lake or river sediment than in free waters, and resuspension of such pathogen-loaded sediments at a later time may introduce a "slug" of bacteria into the waters that is not completely removed by treatment systems. Estimation of Disease Potential by Direct Quantitation of Bacterial Pathogens The detection of bacterial pathogens in water polluted with human or animal fecal matter is relatively easy when large numbers of organisms are present (American Public Health Association, 1975~. Pathogenic bacteria have been isolated from relatively clean reservoirs, rivers, streams, and groundwater; large samples, concentration techniques, and often elaborate laboratory procedures are used. However, detecting the presence of these pathogenic organisms in processed and disinfected water is far more difficult. Scientific literature presents a vast array of media and methods for direct pathogen detection in finished water (Geldreich, 1975~. The greatest emphasis has been on the Salmonella-Shigella group of enteric organisms. Numerous modifications of well-known media are used for
70 DRINKING WATER AND H"LTH pre-enrichment, enrichment, selective inhibition, and isolation, and there are many recommended modifications of incubation temperature and time. Some methods use the classic most-probable-number (MPN) procedure for quantification; others use membrane filtration. Reviews of proposed procedures may be found in the Journal of the Water Pollution Control Federation (Geldreich, 1968, 1969, 1970b; Van Donsel, 1971; Reasoner, 1972, 1973, 1974, 1975~. A recent review appeared in the fourteenth edition of Standard Methods (American Public Health Association, 1975~. There are serious limitations to the use of direct isolation of specific pathogenic bacteria for evaluating water quality. First, there is no single procedure that can be used to isolate and identify all these microorgan- isms. Second, only for salmonellae are the available procedures sufficiently accurate; the methods for other major pathogens such as Shigella, Vibrio, and Leptospira-are inadequate. Third, none of the available procedures is applicable to quantitative isolation of small numbers of pathogens in drinking water. Fourth, even if procedures could be recommended, it is doubtful whether laboratories doing routine bacteriologic studies of water would have the expertise to carry out the procedures reliably. In outbreaks caused by gross contamination, the standard procedures would be of value. Recently, Reasoner and Geldreich (1974) reviewed several of the rapid-detection methods proposed for water and concluded that the cost per test, although perhaps higher than for conventional procedures, must be tolerated for potable-water quality assessment in emergency situations created by natural disasters, treatment breakdown, or rupture in the distribution network. None of these procedures would provide protection to the public as great as that provided by the currently used indicator organism, the coliform. Indicator Organisms The term "indicator organism," as used in water microbiology, means: a microorganism whose presence is evidence that pollution (associated with fecal contamination from man or other warm-blooded animals) has occurred. Indicator organisms may be accompanied by pathogens, but do not necessarily cause disease themselves. As noted above, pathogens are usually more difficult to grow, isolate, and identify than indicator organisms, and often require special media and procedures. Indicator organisms, rather than the actual pathogens, are used to assess water quality because their detection is more reliable
Microbiology of Drinking Water 71 and less time-consuming. Pathogens appear in smaller numbers than indicator organisms and are therefore less likely to be isolated. An indicator organism should have the following characteristics: · Applicable to all types of water. · Present in sewage and polluted waters when pathogens are present. · Number is correlated with the amount of pollution. · Present in greater numbers than pathogens. · No aftergrowth in water. · Greater survival time than pathogens. · Absent from unpolluted waters. · Easily detected by simple laboratory tests in the shortest time consistent with accurate results. · Has constant characteristics. · Harmless to man and animal. No organism or group of organisms meets all these criteria, but the "coliform group" of organisms fulfills most of them. ESCHERICHIA COLI AND THE COLIFORM GROUP Escherichia cold is commonly found in the human intestine. It is not normally a pathogen, although pathogenic strains are known. Physiologi- cally, E. cold and members of the genera Salmonella and Shigella are quite similar. All are classified as enteric bacteria of the family Enterobacteria- ceae (Cowan, 1974~. They are facultatively anaerobic, and are able to ferment sugars with the production of organic acid and gas. These three genera carry out a type of fermentation called "mixed-acid fermenta- tion," but differ in a number of physiological characteristics. Many physiological differences between various enteric bacteria are known (Ewing and Martin, 1974), but at the beginning of the twentieth century this was not so. In the early days of water bacteriology, some simple operational distinctions were necessary. The lactose-fermentation test became the prime diagnostic tool: E. cold ferments lactose with the formation of acid and gas; Salmonella and Shigella do not ferment lactose. One source of confusion is the necessity to distinguish between E. cold and the "coliform group" of bacteria. Although the taxonomy of bacteria is constantly undergoing revision (see Buchanan and Gibbons, 1974, for the latest version), the genus Escherichia is well defined. It is distinguished from other mixed-acid fermenters of the Enterobacteriaceae primarily on
72 DRINKING WATER AND H"LTH the basis of sugar-fermentation reactions, motility, production of indole from t~yptophan, lack of urease, inability to utilize citrate as sole carbon source, and inhibition of growth by potassium cyanide. However, the "coliform group" is not so precisely defined. The "coliform group," as defined in Standard Methods (American Public Health Association, 1975), comprises all "aerobic and facultative anaerobic, gram-negative, non- spore-forming, rod-shaped bacteria which ferment lactose with gas formation within 48 hr at 35 C." This is not a taxonomic grouping, but an operational one that is useful in water-supply and sewage-treatment practice. It includes organisms in addition to E. colt, most importantly Klebsiella pneumoniae and Enterobacter aerogenes, which are not m~xed- acid fermenters. The entry of the term "coliform" into sanitary bacteriology was associated with a policy established by H. E. Jordan when he became editor of the Journal of the American Water Works Association; he stated that he would substitute "coliform" for "E. colf' in papers submitted to him (Jordan, 1937~. Although most isolates classifiable as Escherichia by modern methods ferment lactose, about 5-9% of them do not (Ewing and Martin, 1974~. No isolates of the genus Salmonella, either in the species S. typhi or in other species, produce gas from lactose (Ewing and Martin, 1974~; therefore, a water sample containing Salmonella and a lactose-negative E. cold would be negative on the coliform test and would probably be discarded without further examination, because of the definition of"coliform." Even if glucose were substituted for lactose in a coliform analysis, a significant fraction of organisms would be missed, inasmuch as about 9% of isolates of Escherichia do not form gas from glucose (Ewing and Martin, 1974~. Because there are two procedures the multiple-tube-dilution or most- probable-number (MPN) technique, and the membrane-filter (MF) technique-the coliform group of organisms requires two definitions (American Public Health Association, 1975~. On the basis of the MEN technique, the group consists of all aerobic and facultatively anaerobic, gram-negative, non-spore-forming, rod-shaped bacteria that ferment lactose with formation of gas within 48 h at 35°C. On the basis of the ME technique, the group consists of all organisms that produce a dark colony (generally puIplish-green) with a metallic sheen within 24 h of incubation on the appropriate culture medium; the sheen may cover the entire colony or appear only in a central area or on the periphery. These two groups are not necessarily the same, but they have the same sanitary . ·^ slgnlucance. If the coliform group is to be used as an indicator of fecal pollution of water, it is important to know that the coliforms do not lose viability in the water environment faster than pathogenic bacteria, such as salmonel
Microbiology of Drinking Water 73 TABLE III-3 Comparative Die-Off Rates (Half-TimeJa of Fecal Indicator Bacteria and Enteric Pathogens Bacteria Half-t~me Number O of strains 1 Indicator Bacteria Coliform (avg.) 17.~17.5 29 Enterococci (avg.) 22.0 20 Streptococci (from sewage) 19.5 S. equines 10.0 1 S. bovis 4~3 1 Pathogenic bacteria Shigella dysenteriae 22.4 1 S. sonnet 24.5 1 S. pexneri 2S.8 1 S. enteritidis, paratyphi A&D 16.0 19.2 2 S. enteritidis, typhimurium 16.0 1 S. typhi 6.0 2 V. cholerae 7.2 3 S. enteritidis, paratyphi B 2.4 1 aTime required for 50~O reduction in the population. From McFeters et al. (1974). lee and shigellae. Little information exists on the survival of bacteria in finished water, and the data on other types of water are scattered and fragmentary. McFeters et al. (1974) recently reviewed previous work and presented their own data on die-o~ of intestinal pathogens in well water. As seen in Table III-3, die-o~ rates for pathogens and coliforms are approximately the same. Earlier work on the survival of salmonellae in water was reviewed by McKee and Wolf (1963~. Another factor to be considered is the relative sensitivity of coliforms and bacterial pathogens to disinfection. Although this subject has been studied little recently, the older work (Butterfield et al., 1943; Butterfield and Wattle, 1946; Wattle and Chambers, 1943) indicated that there was essentially no difference between these different organisms in sensitivity to disinfection. This is not true when the coliform group is compared with viral pathogens. Viruses survive longer than bacterial pathogens (Colwell and Hetrick, 1976~. SOME DEFICIENCES OF COLIFORMS AS INDICATOR ORGANISMS Coliforms meet many of the criteria for an ideal indicator organism previously listed; however, there are some deficiencies. There is after
74 DRINKING WATER AND H"LTH growth in water-some strains do not disappear, but adapt to the new environment, and may become part of the natural flora. They are often found in waters, having entered through storm runoff. And they do not have constant characteristics; it is this property more than any other that has recently caused some water bacteriologists to question the continued use of coliforms as indicator organisms. False-negative and false-positive test results are not uncommon. The following summarizes the objections which have been raised by workers to the continued use of coliforms: · Atypical lactose reactions, the concern of bacteriologists as early as 1899 (Parr, 1939), have occurred. · Coliforms may be suppressed by high populations of other organ- isms, especially in untreated groundwater (Allen and Geldreich, 1975) or where there is no free residual chlorine (Geldreich et al., 1972~. · Coliforms do not represent a homogeneous group; it has been suggested (Dutka, 1973) that the definition of"coliform" include any organism defined by Edwards and Ewing (1972) as Enterobacteriaceae (this might require the addition of the ox~dase test to the definition of the group, because all Enterobacteriaceae are oxidase-negative). · The genus Aeromonas is a common cause of false-positive results in warm weather (Bell and Vanderpost, 1973; Ewing et al., 1961; Ptak et al., 1973~. These can be eliminated by the use of the oxidase test, since Aeromonas is oxidase-positive and coliforms are ox~dase-negative. · False-negative results by strains that are unable to ferment lactose can give an unwarranted sense of security. In the final analysis, testing for coliform, while not perfect bacteriologi- cally, is still the most reliable indicator of the possible presence of fecal contamination and therefore of the pathogens that may be present in water. OTHER INDICATOR ORGANISMS Because of certain limitations of the coliform group as general indicators of water quality, workers have continually searched for better indicator organisms. No other organism has been found that is better than the coliform group, but it is pertinent to mention briefly some of these other indicators (Geldreich, 1975~. Fecal coliforms (defined as those organisms which develop on media incubated at 44.5°C) have frequently been used in stream and lake pollution work, but are not suitable as indicators of drinking water quality because the number of fecal coliforms is considerably lower in source waters than total coliforms, making the test
Microbiology of Drinking Water 75 less sensitive. Fecal streptococci (defined as those organisms able to grow on medium containing sodium azide) have been used in water pollution work, but have not proven suitable for drinking water analysis because of low recovery rates, poor agreements between various methods, and uncertainty as to their significance in water. Several other organisms have been suggested as indicators but have found even less acceptance: Clostridium perfringens, Bi~dobacterium, Pseudomonas, Staphylococcus. It would be undesirable and extremely risky to substitute any organism for the coliform group now, although research studies that compare other indicator organisms with coliforms are warranted. RAPID METHODS FOR COLIFORM COUNTS There is great need for a rapid method of coliform counts which could give results in a shorter time than the 18-24 h required by the membrane- filter method. A rapid method would permit early detection of system problems, and would considerably increase public health protection. Correlation of any rapid method used with accepted standards methods is essential, at least to the degree that the ME and MEN methods correlate. Unfortunately, no suitable rapid method for drinking water exists at present, but we will mention several methods briefly, in order to encourage research in this area. One theoretically-feasible method would be a direct microscopic method, in which coliform cells would be concentrated on a membrane filter and stained by a specific staining procedure. The most commonly suggested staining procedure has been the use of fluorescent antibodies specific for the coliform organisms (Danielsson, 1965; Ginsburg et al., 1972~. Unfortunately, this method is unlikely to succeed because there are at least 145 serotypes of E. colt, plus several more for the Klebsiella and Enterobacter group, so that a polyvalent antibody containing all of these antibodies would be necessary. Also, the number of cells in drinking water is likely to be so low that inconveniently long observation times would be necessary for quantifying the organisms present. Another disadvantage is that fluorescent antibodies might not distinguish living cells from dead, so that nonviable cells (perhaps resulting from a disinfection treatment) would also be counted. Another rapid method that was suggested involves the use of radioactively labeled lactose and measurement of the radioactive CO2, liberated as a result of metabolism of the coliforms. This procedure was first suggested by Levin et al. (1961) and has been studied further by Scott et al. (1964~. The EPA in Cincinnati (Reasoner and Geldreich, 1974) is currently attempting to improve this method. One of the desirable
76 DRINKING WATER AND H"LTH features of such a method is that it could be automated and run on-line. A disadvantage of the system is that, at present, it is not sensitive enough to detect 10 or fewer organisms after 6 h of incubation. It seems unlikely that such a method could ever be made sensitive enough to detect single coliform cells in 100-ml water samples, so that it probably will never replace standard methods, but it might prove useful in process control in large water systems. The EPA at Cincinnati has also developed a 7-h fecal coliform test, employing a membrane filter technique (Reasoner and Geldreich, 1974~. Several other rapid methods which have been under study can be found in: Cady and Dufour (19741; DeBlanc et al. (1971~; and Newman and O'Brien (1975~. SAMPLING FOR THE COLIFORM TEST Water bacteriologists and sanitary engineers agree that the weakest link in the chain of water-monitoring and -testing is the collection of the sample. Too often this is left to the semiskilled or untrained worker. The USEPA Interim Regulations (Federal Register, Dec. 24, 1975i follow the USPHS 1962 Drinking Water Standards: the number of required samples per month is based on the population served. Prior to 1925, when water quality standards were under the jurisdiction of the Treasury Department, bacterial sampling was left to the discretion of the local health and water utility officials and varied widely according to local practices and the capacity of the laboratory. This, of course, was unsatisfactory, and was so recognized by responsible water utility people. In 1941, a conference was called by the USPHS to revise the drinking water quality standards, and among the subjects under consideration were the frequency of sampling, the location of the sampling points, and the increased number of samples from the distribution system rather than from the plant final effluent. The deliberations of this 1941 conference, which resulted in the adoption of new Standards in 1942, can be found in the Journal of the American Water Works Association: 1941, 33:1804; 1942, 35: 135; 1942, 35:93; and 1941, 35:2215-2226. The principal changes adopted by the U.S. Public Health Service in 1941 were as follows: 1. The number of samples to be examined monthly from the distribution system would depend on the population served. During this discussion, the quality of the source water, the treatment procedure, the sanitary condition of the distribution lines, and the daily volume delivered to the consumers were also considered. Apparently, these latter
Microbiology of Drinking Water 77 factors made the final decision so complicated that in the final standard, only the size of the population served was used to decide on sampling frequency. 2. The detection of potential hazards in the distribution system due to faulty plumbing, cross-connection, post-contam~nation, and faulty plant and distribution system operation was included in the standards. Samples were to be taken from representative points throughout the distribution system, with the frequency to be such as to determine the bacteriological quality of the water. The minimum number of samples per month was to be determined from the graph appearing in the Journal of the American Water Works Association (35:93-104,1942~. Apparently, this graph became the basis of the 1962 standards and the present interim regulations. According to F. Donald Maddox, Chief, Water Supply Systems Section, Region V, USEPA (personal communication to Walter Ginsburg), the subject of sample frequency was again discussed by the Committee that revised the 1962 Drinking Water Standards, and no changes were made. It is Maddox's opinion that the original 1942 curve was based on a number of water supply systems of different sizes which were known to have good treatment facilities and to be sampling at what was considered to be adequate frequency. The curve was clotted using 1 C, these selected cities as a base line. Richard L. Woodward, who served on the 1946 and 1962 Standards revision committees (personal communication to Walter Ginsburg) bears out the contention that the decision as to the number of samples to be examined was based on these elements of expert judgement. One inherent weakness of this frequency grape has been discussed in the Federal Register (Dec. 24, 1975, n. 595681. It concerns the question of preparing monthly coliform averages from monthly percentages of positive samples. When four or fewer samples are examined each month, and one sample exceeds 4/100 inl, there is no way that the monthly average can conform to the recommended standards, even though the other three samples are negative. The regulations give the states the authority to average over a 3-montn period when four or fewer samples are the required rate, but this hardly affords a community the protection it needs when the water system is doing a poor job of treatment. The Community Water Supply Survey by the US~PA n 1969 (McCabe et al., 1970) showed that 85% of the systems surveyed did not collect the required minimum number of samples. Such improper sampling frequency is one of the most abused requirements. This can cause problems in tile distribution system to go undetected (Geldreich, 1971; Miller, 1975~.
78 DRINKING WATER AND H"LTH In light of this discussion, it is clear that considerable research is necessary, with modern statistical methods, to develop better sampling protocols for water systems that serve different populations. COLIFORM STANDARDS United States Standards The first U.S. national standards for bacteriological water quality were established in 1914 (Public Health Reports, 1914~. These standards were specifically applicable to water used on interstate carriers, but were adopted quite early (formally or informally) by many states. Morse and Wolman (1918) concluded that -the standards are not precise and accurate indices of quality, but simply a convenient mode of analysis for comparative purposes that must be used with considerable caution. It is obvious that with drinking water that meets the standards there is no absolute assurance of the absence of pathogens, only confidence that their presence is unlikely; hence, the probability of waterborne-disease transmission is decreased. The decline in morbidity and mortality from some diseases, such as cholera, typhoid fever, salmonellosis, and shigellosis, provides some evidence of the validity of this confidence, although some of the decline may be due to the generally better health of the population, making people less susceptible to infection. The latest standards adopted by the PHS were those of 1962 (U.S. Public Health Service, 1962~. Although none of the bacteriological numerical values was changed, a major procedural revision was made; the membrane-filter technique was accepted as an equivalent alternative to the multiple-tube-dilution (MPN) technique that had been in use since 1914. The ME standard was set at one coliform/ 100 ml. The interim regulations of the EPA (Federal Register, Dec. 25,1975) have broadened the applicability of the standards to all public water systems, specified a sample size of 100 ml when the ME technique is used, and modified the required frequency of sampling. A monthly mean of less than one coliform/100 ml is still the standard. International Standarcis Although many countries have their own drinking-water standards, two standards have international status the World Health Organization European Standards (World Health Organization, 1970) and the Interna- tional Standards (World Health Organization, 1971~. In their preparation, individual national standards were considered.
Microbiology of Drinking Water 79 The International Standards are proposed as "minimal standards which are considered to be within the reach of all countries throughout the world at the present time" (World Health Organization, 1970~. They distinguish between piped supplies (roughly equivalent to the EPA definition of community public systems) and individual or small community supplies (comparable with the EPA definition of noncommu- nity systems) and between water leaving the treatment plant and that in the distribution system. For treated, disinfected supplies, water entering the distribution system should be of such quality that no coliform bacteria can be demonstrated (by the specified procedures) in 100 ml of water. For undisinfected water, the samples should yield no E. cold and three or fewer coliforms/100 ml. Within the distribution system, the standards are specified that "~1) Throughout any year, 95% of samples should not contain E. cold in 100 ml; (2) No sample should contain more than 10 coliform organisms per 100 ml; and, (3) Coliform organisms should not be detectable in 100 ml of any two consecutive samples." In nonpiped systems, the coliform count should not exceed 10/100 ml. The European Standards are comparable to the International stan- dards, but do not distinguish-in terms of quality between disinfected and undisinfected water. The standard of 95% of 100-ml samples showing no coliform bacteria throughout a year "corresponds to an average density of about one coliform organism in 2 liters of water." Despite the numerical differences between the U.S. and the international standards, it is the intent of both sets of standards that coliform bacteria be absent from drinking water, to provide protection against disease. STATISTICAL LIMITS In discussing numerical standards of bacteriological quality of drinking water, the accuracy and the statistical limits of the tests must be considered. In using the classical, multiple-tube-fermentation test for coliform bacteria, it has been recognized that the procedure itself has a large inherent error (Morse and Wolman, 1918~. A more recent discussion of these limits (Prescott et al., 1946) concluded that "because of the marked inaccuracy of the dilution MEN method . . . any tendency toward fictitious accuracy in expressing the result should be discouraged." Table III-4, reproduced from Standard Methods for the Examination of Water and Wastewater (American Public Health Associa- tion, 1975), clearly shows this. The membrane filter technique, because it is equivalent statistically to a plate count, has a much smaller error. Thomas and Woodward (1955)
80 DRINKING WATER AND H"LTH TABLE III-4 MPN Index and 95% Confidence Limits for Various Combinations of Positive and Negative Results, Using Five l~ml Samples No. of Tubes 95% Confidence Giving PositiveMPN Limits Reaction out ofIndex 5 of 10 ml Each per 100 ml Lower Upper 0 < 2.2 0 6.0 1 2.2 0.1 12.6 2 5.1 0.5 19.2 3 9.2 1.6 29.4 4 16. 3.3 52.9 5 >16. 8.0 Infinite From American Public Health Association ( 1975). compared the two methods and concluded "that, on the average, the former MPN gave higher indications of density by a factor of 1.0-1.9 with an average of 1.3 for the specific technics used.... However, the difference is not regarded as important from a practical viewpoint because of the inherent lack of precision of the individual MPN value." and that, "With nearly all of the samples listed, the precision attained with a single filter was found to be two to five times greater than that of a 5-5-5-tube MPN" (note that the drinking water standards require only 5 tubes at one dilution with a corresponding greater lack of precision). THE HEALTH SIGNIFICANCE OF THE COLIFORM TEST Several pathogens, notably those in the genus Shigella, are able to initiate infection in humans even when introduced in very low numbers. Because it is not feasible to assay for bacterial pathogens directly in water, it is important to consider the utility of the coliform test in ensuring the bacteriologic safety of drinking water. A direct approach to assessing the significance of the coliform count would be to obtain e-idence of a correlation between numbers of coliforms and numbers of pathogenic bacteria (e.g., salmonellae or shigellae). One attempt to seek such a correlation was the study of Kehr and Butterfield (19431. Although imperfect, this is the only study found that attempts to relate the coliform count directly to disease incidence. In the discussion below, this study is analyzed in some detail in order to present a picture of the approach necessary to place the coliform
Microbiology of Drinking Water 81 standard on a firmer scientific basis. (It is perhaps of historical interest that the motivation for their study was an adverse decision by the Illinois Supreme Court as to the value of the coliform count in indicating that water is unsafe.) . _ C, am_ ~ The approach used by Kehr and Butterfield involved two aspects. First, data on the relative proportions of Salmonella typhosa and coliforms in various types of water (such as river, sewage, and sludge) were obtained from the literature, and the numbers of S. typhosa per 106 coliforms were calculated. These data, obtained from cities throughout the world, were then plotted on the ordinate on log-log paper, with typhoid fever morbidity on the abscissa (Figure III-2. An approximately straight line was obtained. The point here was that the excretion rate of coliforms would be the same in a healthy as in ~ sick population, but that the latter would also excrete typhoid bacteria. Kehr and Butterfield then considered the stability of the E. coli-S. typhosa ratio and showed that S. typhosa and coliforms died on at approximately the same rate during sewage treatment, during self-purification in streams and lakes, and during drinking-water purification. The stability of this ratio is not . . ~. ~_,., . ~ w_ ~ in, , , - surprising, when it is considered that S. typhosa and coliforms are members of the same group of bacteria and are likely to show similar tolerance and sensitivities to environmental influences. From the data in Figure III-2 and from recorded waterborne outbreaks of typhoid fever, Kehr and Butterfield estimated a minimal infecting dose of S. typhosa for the general population and the percentage of persons infected by that dose. In doing this, these workers considered only epidemics of typhoid fever of a disuse nature, i.e., characterized by a low attack rate but spread over a fairly large population. Kehr and Butterfield wrote that such epidemics were common at the time. Outbreaks with high attack rates, in which infection could arise more or less directly from carrier or patient discharge, were not considered, in order to avoid situations not likely to involve drinking water. In the disuse typhoid epidemics, a common observation is the additional widespread occur- rence of nontyphoid gastrointestinal disturbances. Considering the frequency of occurrence of the diffuse pattern of epidemics, and the data on concentrations of S. typhosa found in sewage and polluted waters (as given in Figure III-2), Kehr and Butterfield concluded that it would be unlikely in such an outbreak for a person to drink more than a single typhoid bacterium, or at most only a few, and that a single typhoid organism could produce infection in a small percentage of the general population. This conclusion is consistent with studies in experimental animals, which have shown that infection can be initiated by single bacterial cells (Meynell, 1961; Meynell and Meynell, 1965~.
82 DRINKING WATER AND HEALTH 120 100 E o . _ o c o . _ E ~ 10 - J U] At O A: U) Tjitpoes '0 River - ~ ^~> ~ Thames R iver _ _ ~ ~ Q` ~ ~ ~ 6~~ ~ ~ ~In' London Sewage`: .' \ Palo Alto Sewage I W'' - - .' O A' 0<<~, '~0 Belfast Sewage Bandoeng Sewage $ Uncorrected for Recovery Ratios · Corrected for Recovery Ratios 1 1 1 1' .01 .1 1.0 10 100 TYPHOID FEVER MORBIDITY (cases/1000 population/year) FIGURE III-2 Ratio of Salmonella tvphi per million coliforms for varying typhoid fever morbidity rates. From Kehr and Butterfield (1943). But this conclusion is not unequivocal. Populations of bacteria, even if derived from the same clone, can have a range of infectivities, and populations of humans have a range of susceptibilities. Thus, although a single cell may initiate an infection, not every cell-host contact will lead to infection. When the pathogenic bacterial population is diluted by a large volume of drinking water and spread over a large population, there is a probability that an appropriately virulent cell will reach a susceptible person. This is the situation that Kehr and Butterfield considered in their analysis of disuse waterborne typhoid epidemics. If the hypothesis that a single typhoid bacterium is infective can be accepted, then it is possible to consider the significance of the typhi:coliform ratio in drinking water. Assume that a water plant is treating source water with a typhi:colitorm ratio of 10:106, corresponding roughly to the ratio found in many polluted surface waters. Assuming equal destruction of typhoid bacteria and coliforms during treatment, if the finished water contained one coliform/100 ml (a reasonable possibili
Microbiology of Drinking Water 83 ty, even in many well-run plants), then the probability of consuming one typhoid bacterium when drinking a liter of water would be 1O-5; or, put another way, 10 people in a population of 1,000,000 could be infected. In an attempt to verify this, Kehr and Butterfield analyzed a number of epidemics of waterborne disease of the disuse type, and concluded that the observed incidence of infection was consistent with this hypothesis. Thus, it was concluded that water which does not meet the coliform standard (one coliform/100 ml) can be responsible for waterborne disease, both gastroenteritis and typhoid fever. Even more important, this analysis raises the question of whether water that meets the standard might bear disease-causing organisms. It is a simple exercise in arithmetic to convince oneself that this could be the case, inasmuch as drinking water with less than one coliform/100 ml, thus meeting the standards, could well have pathogenic organisms. Suppose that, instead of analyzing 100 ml of water, 1 liter were analyzed, and one coliform was found. This would be a coliform count of 0.1/100 ml, equivalent to one typhoid infection per 1,000,000 people (assuming the same typhi:coliform ratio as in the previous paragraph). Since the incidence of nonspecific gastroenteritis can be expected to be higher than that of typhoid, water that meets the present coliform standard may be the bearer of disease. However, existing epidemiologic data and reporting systems would not permit detection of such waterborne incidents, because the number of organisms would be below the detection limits of current surveillance methods. The epidemiologic work of Stevenson (1953) added much weight to the rationale of establishing a coliform standard for drinking-water sources. His analyses showed that if raw water has fewer than 1,000 coliforms/100 ml, then it would be very likely that the salmonellae in finished water would be below infective levels. Gallagher and Spino (1968) challenged the validity of the standards and stated that "summarized data from several stream surveys over the past few years showed little apparent correlation between quantities of total or fecal coliforms and the probable isolation of salmonellae." Geldreich (1970a) challenged their conclusion, on the basis of fecal coliform detections that showed a correlation between coliform numbers and salmonellae isolations. He showed from numerous previous studies that, when fecal coliforms were 200/100 ml or more, there was a finite probability of isolating salmonellae. Smith and Twedt (1971) corroborated these data in a study of two Michigan rivers. There are well-known epidemiologic histories of the presence of bacterial pathogens when the coliform index was low. Boring et al. (1971) reported that Salmonella typhimurium outnumbered coliforms by a factor
84 DRINKING WATER AND H"LTH of 10 in the Riverside, California outbreak. Similarly, a report by Seligmann and Reitter (1965) showed that index organisms can be low in the presence of pathogens. These sporadic reports of the failure of the index-organism concept emphasize the need for more research on pathogen detection. CONCLUSION ON COLIFORM STANDARD Thus, we conclude that the current coliform standard is of value in protecting against frank outbreaks of bacterial diseases, but it may not protect against low levels of virulent pathogens. It is clear that further protection could be achieved by analyzing larger samples for coliforms. More important than analysis of larger water samples for the protection of the public health more frequent sampling, especially at more points throughout the distribution system should be considered. It has been reported repeatedly in the literature that the presence of any type of coliform organism in drinking water is undesirable. The regulations essentially demand that colifo~-free water be distributed to consumers. Wolf (1972) has ably summarized: "The drinking water standard presently in use (approximately one coliform per 100 ml) is, in a sense, a standard of expedience. It does not entirely exclude the possibility of acquiring an intestinal infection. It is attainable by the economic development of available water supplies, their disinfection, and, if need be, treatment in purification works by economically feasible methods. It is not a standard of perfection." It is not clear that the colifo~n~ standard provides comparable protection against virus disease. In fact, available information indicates that viruses may survive considerably longer than coliform bacteria outside the human host, and that the infectious dose may be very small (see section on viruses). Good engineering and public health practices emphasize the need for using raw water of the highest possible quality, and for providing adequate sanitary-survey information. Bacteriological testing or the imposed use of bacteriological standards are adjuncts, not replacements for good-quality raw water, proper water treatment, and integrity of the distribution system. The present coliform standards appear adequate to protect public health when: raw water is obtained from a protected source, is appropriately treated, and is distributed in a contamination- free system. Current coliform standards are not applicable for water reclaimed directly from wastewater.
Microbiology of Drinking Water 85 The Standard Plate Count The standard plate count (SPC) for drinking water, as described in Standard Methods (American Public Health Association, 1975), is the plating of small quantities (usually 1.0 or 0.1 ml) of a properly collected water sample in a nutrient agar medium (plate count tryptone glucose extract agar) and incubating aerobically for a fixed period at a prescribed temperature (35°C for 24 h or 20°C for 48 h).* The SPC is often referred to by other names, such as "total count," "viable count," "plate count," "bacterial count," "water plate count," and "total bacterial count." The organisms that develop into colonies on the agar plates represent a portion of the total population of bacteria in the water. This portion contains the bacteria that grow in the prescribed time in the specific environment provided. Although the method described is standard, there is no universal agreement on the acceptable concentration of organisms in drinking water. The most common allowable bacterial numbers used by health departments, water-supply agencies, and local jurisdictions vary from 100/ml to SOO/ml of colony-forming units. Geldreich (1973) has summarized several ways in which a standard based upon fecal-organ~sm detection alone may not provide adequate health protection. Information gained from continued surveillance of water supplies by a standard plate count would provide an added degree of health protection. Even though water treatment is adequate and chlorine disinfection is provided, quality could deteriorate in the distribution system as a result of growth of organisms other than detectable conforms. Finished water containing flee, or unfiltered turbid waters, may carry organisms past the disinfection treatment, or the organisms may be protected by association with larger forms of life, such as nematodes (Chang et al., 19601. Changes in water pressure in the distribution lines may cause release of organisms from protected areas, dead ends, or protecting materials in the system. Geldreich et al. (1972) and Geldreich (1973) also reviewed the phenomenon of suppression of fecal organisms by the presence of large populations of organisms in common genera whose members normally show up in standard plate counts. These genera include Pseudomonas, Bacillus, Streptomyces, Micrococcus, Flavobacterium, Proteus, and various yeasts. It was concluded that the presence of 1,000 noncoliform organisms *In most tap waters, the number of organisms capable of growing at 20°C is considerably higher than the number growing at 35°C. It is unlikely that most organisms growing at 20°C will also grow at 3S-37°C, and hence could not be pathogenic to man. A 20°C count could still be of value in water works practice, in providing information about filtration efficiency and about possible contamination of groundwater supplies.
86 DRINKING WATER AND H"LTH per ml could suppress the growth of coliforms. Geldreich 0973) speculated that this type of suppression of coliforms below detectable concentrations may have occurred in the Riverside, Calif., Salmonella typhimurium outbreaks, where coliform analyses did not reveal the contamination (Boring et al., 1971~. Plate counts over SOO/ml also seemed to make detection of salmonellae and shigellae very difficult in a bacteriologic study of potable water in Karachi, Pakistan (Ahmed et al., 1967~. Although the genera of organisms detected by the SPC may not be harmful or dangerous to normal humans when present in drinking water in low numbers, under special circumstances (such as during therapy), these organisms are known to produce severe acute or chronic human infections (Geldreich et al., 19721. Geldreich (1973) proposed that a SOO/ml limit be placed on the standard plate count and that immediate investigation of water treatment and distribution systems be undertaken whenever the limits were exceeded. He recommended that water supplies be monitored routinely at least every 3 months to maintain the baseline data on the general bacterial population. A summary from Geldreich's (1973) paper follows. It should be clearly understood that the standard plate count is not a substitute for total coliform measurements of the sanitary quality of potable water. Rather, the use of a standard plate count limitation will: 1. Provide a method of monitoring for changes in the bacteriological quality of finished water in storage reservoirs and distribution systems. 2. Indirectly limit the occurrence and magnitude of Pseudomonas, Flavabacterium and other secondary pathogenic invaders that could pose a health risk in the hospital environment. 3. Reduce problems in the detection of low densities of total coliforms due to interference by noncoliform bacteria. 4. Monitor the electiveness of chlorine throughout the distribution network and provide a warning of filter effluent-quality deterioration and the occurrence of coliform breakthrough. 5. Indicate the existence of sediment accumulation in the distribution network that provides a protective habitat for the general bacterial population. Finally, the noncoliform bacterial population can be controlled by removing sediment and slime deposits from the distribution network followed by continuous application of chlorine in sufficient dosage to insure the maintenance of a free residual throughout the system. It should be emphasized that a standard plate count of SOO/ml is attainable by water systems. Control of the general bacterial population
Microbiology of Drinking Water 87 in a variety of public water-supply distribution systems was demonstrated by Geldreich (1973), with a chlorine residue of approximately 0.3 mg/liter. In 60~7o of 923 water systems, standard plate count densities of 10/ml or less were obtained. Although it is difficult to document, it is probably true that among the reasons for the decline in use of the standard plate count for drinking water are the cost of the test, the lack of trained people to perform the number of tests needed, the difficulty of implementing and monitoring the laboratories needed, the resistance that would occur among the small and isolated water-plant operators, and similar issues. Although these are valid criticisms, they do not speak to the basic question of whether a regular system of plate-count surveillance would provide a tool for assessing the health hazards of water. CONCLUSIONS ON STANDARD PLATE COUNT The standard plate count is a valuable procedure for evaluating the bacterial quality of drinking water. Ideally, standard plate counts should be performed on samples taken throughout systems. The SPC has major health significance for surface-water systems that do not use flocculation, sedimentation, filtration, and chlorination, and for those groundwater systems that do no chlorination. When it is used, the sampling frequency should be at least 10~7o of the frequency of the coliform analysis, except that at least one sample should be collected and analyzed each month. The scientific information available makes it reasonable to establish the upper limit of the SPC initially at 500/ml, as developed in 35°C, 48-h plate count procedure (using the procedures prescribed by Standard Methods (American Public Health Association, 1975.) RECOMMENDATIONS FOR RESEARCH ON BACTERIAL CONTAMINANTS A research program is needed to increase the value of the relatively simple bacteriological tests in controlling the sanitary quality of drinking water. The program should include: 1. Epidemiological studies of water quality and health, with applica- tion of more sensitive methods for detecting pathogens in drinking water and better reporting of outbreaks of waterborne disease. 2. Development of membrane-filtration methods to allow testing of larger samples and to reduce interference by overgrowth and disinfec- tants. 3. Improvement of procedures for making total plate counts and study
88 DRINKING WATER AND H"LTH of the utility of such tests for assessing the health hazards of drinking water. 4. Research on more rapid and sensitive methods for detecting pathogens and the use of such methods for monitoring the quality of water. VIRUSES Viruses diner fundamentally from other microrganisms that may occur in water. They are transmitted as submicroscopic, inert particles that are unable to replicate or adapt to environmental conditions outside a living host. These particles, or virions, have the potential to produce infections, and sometimes disease, in people who ingest them with drinking water. A viral particle eventually loses its infectivity with the passage of time, and with exposure to the rigors of its environment. The viruses important to human health that are most likely to be transmitted by drinking water are the enteric viruses. These are primarily parasites of a portion of the intestinal tract. The stomach and duodenum are seldom affected by viruses, partly because of unfavorable conditions. Acid and proteolytic secretions predominate in the stomach; these and bile empty into the duodenum. Only viruses that can withstand such conditions will remain infectious and thus able to implant further down the digestive tract. The most important human enteric viruses are the enteroviruses (i.e., acid-stable picornaviruses), reoviruses, parvoviruses, and adenoviruses (Fenner et al., 1974~. The virions of all these groups are roughly spherical, acid-stable, and lack envelopes. All are relatively stable in the environ- ment outside the host organism. Enterovirus particles are small (20-30 rim) and contain single-stranded ribonucleic acid (RNA). Reovirus particles are medium-sized (70-80 nm) and contain double-stranded RNA. Parvovirus particles are small and contain single-stranded deoxyri- bonucleic acid (DNA). Adenovirus particles are medium-sized and contain double-stranded DNA. Hepatitis A (infectious hepatitis) is transmitted by particles that closely resemble those of the enteroviruses (provost et al., 1975~. Unlike bacteria, the viruses are obligate intracellular parasites and cannot replicate in a cell-free medium. Many viruses can be grown in cultured cells in vitro. Some transmitted to man by water, most notably the virus of hepatitis A, have not yet been cultivated in vitro. Ordinarily, viruses are detected and enumerated on the basis of their infectivity in cell culture or experimental animals. Testing of viruses for
Microbiology of Drinking Water 89 infectivity was greatly advanced by the finding that animal cells grown in vitro would support the replication of human viruses (Enders et al., 1949~. Infectivity resides in the nucleic acid portion of the virus particle. When a suspension of infectious particles is inoculated into a culture of susceptible cells, the particles are engulfed by, or penetrate, host cells, and the cells produce progeny virus. Death of the cells often results. If diffusion of the progeny virus particles is restricted by gelling the medium in which the cells are maintained, cell death occurs in a localized area called a "plaque." Such a plaque can be initiated by one viral particle, so that plaque enumeration has become an important method to measure animal viruses (Dulbecco, 1952; Dulbecco and Vogt, 1954~. However, a plaque is not always initiated by a single viral particle, and many viral suspensions contain more aggregates than single particles (Sharp et al., 1975~;so the plaque-forming unit (PFU) is not an absolute basis for determining the numbers of viral particles. Knowledge of viruses was acquired much more rapidly after the advent of cell-culture techniques. Nevertheless, transmission of enteric viruses by water had already been surmised by the time cell cultures became available. History of the Enteric Viruses The first reported epidemic of poliomyelitis in the United States occurred in New England (Putnam and Taylor, 1893~. Caverly's ~ 1 896) observation that many cases had occurred in the Otter Creek Valley in Vermont suggested that the disease might be waterborne. The poliomyelitis virus had been thought to infect people by the nasopharyngeal route. Kling (1929) found more virus in patients' stools than in their throats, but his findings were not generally accepted. Later, Harmon (1937) reported isolating virus by inoculating monkeys with colonic washings of four patients whose nasopharynxes yielded no virus. It was evident that infantile paralysis was not a strictly neurotropic disease and that the causative agent was strongly associated with the human intestines. The picture was complicated by the presence of yet-unidentified agents other than the polioviruses in stools. The criterion of neuronophagia, seen in the spinal cords of inoculated monkeys, did not reliably distinguish the polioviruses from these other agents. In 1948, Dalldorf and Sickles reported isolating viruses from the stools of children by inoculation into suckling mice. The original stool specimens came from Coxsackie, N.Y., but "coxsackieviruses" were soon being discovered in many other locations.
90 DRINKING WATER AND H"LTH When cell cultures became available, tests of stool suspensions began to yield viruses that were neither polioviruses nor coxsackieviruses (Robbing et al., 1951~. The isolations of many such viruses were not clearly associated with any disease; these agents were eventually designated "enteric cytopathic human orphan" (ECHO) viruses. The National Foundation for Infantile Paralysis convened a committee to identify and classify these new viruses (Committee on the ECHO Viruses, 1955. The echoviruses were later classified with the polioviruses and coxsackieviruses to form the enterovirus group, members of which are numbered serially (Committee on the Enteroviruses, 1957, 1962~. The reoviruses were originally designated as members of the ECHO group. In 1959, Sabin suggested the term "reovirus" to apply to three serologic types of viruses that had common properties that differed from those of the echoviruses. The reoviruses are common in wastewater; 80C%o of the viruses that were isolated from wastewater by England et al. (1967) were reoviruses. The reoviruses have been recovered from persons with a wide variety of illnesses, but in no case has their etiologic role been established unequivocally (Horsfall and Tamm, 1965~. The adenoviruses (Rowe et al., 1953) have produced large outbreaks of acute respiratory disease in military populations, and were a serious problem during World War II (Klein, 1966~. They produce sporadic infections in the general population but are not often associated with overt disease. They occur frequently in sewage and wastewater and might thus contaminate drinking water. Their significance to the health of the general public is uncertain. Probably the most important viral disease sometimes transmitted by water is hepatitis A. The disease was first described by Lurman in 1885. Its most common route of transmission is fecal-oral, primarily by person- to-person contact. In addition to water and personal contact, the disease is sometimes transmitted by food. The largest epidemic of hepatitis known to have been transmitted by water occurred in 1956 in Delhi, India; it was estimated that more than 30,000 cases occurred after sewage contamination of drinking water(Anonymous, 1957; Melnick, 1957~.The virus has never been cultivated in cell culture, so that all available evidence about the transmission of infectious hepatitis by polluted drinking water is epidemiologic. Mosley (1967) suggested that the incidence of waterborne infectious hepatitis is grossly underreported, but estimated that less than 1% of the cases of this disease result from transmission by water. Viruses that are still unidentified may also be responsible for waterborne disease, since there are many episodes of waterborne gastroenteritis and diarrhea of unknown etiology. Taylor et al. (1972)
Microbiology of Drinking Water 91 listed gastroenteritis as the most common waterborne disease, on the basis of number of outbreaks during 1961-1970. It is not clear which of the enteric viruses, if any, produce this common illness. Epidemiology Knowledge that a virus is shed in feces and may persist for a short time in wastewater is not proof that it is transmitted by water. Clearly, not every virus that is present in feces is waterborne. Viruses transmitted by a fecal- oral cycle can always be transferred directly from person to person, but this requires close contact. Several of the enteric viruses are known to be transmitted by water, but only some of the time. Eight outbreaks of poliomyelitis in Europe and North America were eventually attributed to transmission by water, but Mosley (1967) believed that only one of them was adequately documented. This occurred in "Huskerville," Nebr., in 1952; at least 45 people were made ill after contamination of a municipal water system. The viral disease most frequently reported to be transmitted by water is hepatitis A (infectious hepatitis). Viral hepatitis became a reportable disease in the United States in January 1952, but the two types of hepatitis (i.e., infectious and serum hepatitis) were not separately notifiable until 1966. The viral etiology of infectious hepatitis and transmission of the virus through water were adduced by Neefe and Stokes in 1945. Published reports of 50 hepatitis A outbreaks, attributed to contaminated drinking water, were summarized by Mosley (1967), who found the evidence of transmission by water to be convincing in only 30 of the 50 outbreaks cited. Craun et al. (1976) reported 13 outbreaks of waterborne hepatitis that affected 351 people during 1971-1974. Although hepatitis A has been implicated epidemiologically in some 66 waterborne outbreaks since 1946 (Table III-5), there is no evidence of its transmission through correctly operated conventional water-treatment systems, except where defects in the distribution system have been found to be the source of contamina- tion (Craun, 1976~. About 40,358 cases of hepatitis were reported in 1974 (Center for Disease Control, 1974), but only a small fraction of these cases was transmitted by water. Epidemiologists have techniques that permit a decision as to whether an outbreak is waterborne or foodborne. These techniques require adequate and thorough reporting of the outbreak and a follow-up of infected individuals, to determine opportunities for contact with the
92 DRINKING WATER AND H"LTH TABLE III-5 Waterborne Outbreaks of Hepatitis A by Source of Contamination, 194~1974 Semipublic Municipal and Individual Cause Systems Systems Untreated surface water Untreated groundwater Inadequate or interruption of disinfection Contamination through distribution system Insufficient data to classify TOTAL 4 22 10 26 5 2 44 disease agent. Since most waterborne outbreaks involve breakdowns or deficiencies in water-treatment systems, they are usually signalled as waterborne by their localization in one city or district and by evidence from conform counts of unsafe water. Hepatitis A infections also result from the consumption of contaminat- ed shellfish (Dents, 1974; Dismukes et al., 1969; Dougherty and Altman, 1962; Gard, 1957; Mason and McLean, 1962; Roos, 1956; and Ruddy et al., 1969~. Much information involving shellfish has been derived from studies with enteric viruses other than the hepatitis agent. Clams (Hoff and Becker, 1969; Kohl et al., 1967; Liu et al., 1966b), oysters (DiGirolamo et al., 1970; Hedstrom and Lycke, 1964; and Metcalf and Stiles, 1965), and mussels (Bellelli and Leogrande, 1967; Bendinelli and Ruschi, 1969; and Dub, 1967) have been implicated. In water polluted with human feces, the shellfish accumulate enteric viruses (Liu et al., 1967; Metcalf and Stiles, 1965; and Mitchell et al., 1966), including the hepatitis virus. Humans have become infected by eating improperly cooked shellfish (DiGirolamo et al., 1970; Kohl and Sear, 1967; Kohl et al., 1967; and Mason and McLean, 1962~. The shellfish themselves do not become infected (Chang et al., 1971~; rather, the virus is confined largely to their digestive tracts (Liu et al., 1966b; Metcalf and Stiles, 1968~. If the shellfish are removed to water that is not polluted, they will eventually free themselves of the virus (Liu et al., 1966a; Metcalf and Stiles, 1968; and Seraichekas et al., 1968~. Neither the common cold nor gastroenteritis is a reportable disease; these are probably the two most common (in the order named) human illnesses. Gastroenteritis has been called "sewage poisoning," "summer flu," and 'unspecified diarrhea." Symptoms include nausea, vomiting, and diarrhea. No single etiologic agent has been identified; comparable
Microbiology of Drinking Water 93 illnesses may be caused by Shigella sonnet, Salmonella typhosa, enteropa- thogenic k~scherichia colt, and Ciardia lamblia, as well as nonbacterial agents (Craun and McCabe, 1973~. lThe"Norwalk agent" (Kapikian et al., 1972) and other viruses may eventually take their places among the known etiologic agents of gastroenteritis, but some questions as to how viruses may cause gastroenter~tis have not yet been answered (Blacklow et al., 1972~. Infants whose intestines are infected with oral poliomyelitis vaccine virus do not show the symptoms of gastroenteritis. There have been two outbreaks of gastroenteritis in the United States (one of which led to hepatitis) in which enteric viruses were recovered from the drinking water. In Michigan, restaurant patrons who drank unchlorinated well water became ill within 30 h (Mack et al., 1972~. Records showed that the well water had been contaminated with coliform organisms several times during the previous 6 months. No coliforms were detected initially in a 50-ml sample of the water. After 9.46 liters (2.5 gal) were concentrated by ultracentrifugation, both Escherichia cold and poliovirus 2 were recovered from the sample. No salmonellae or shigellae were recovered. A second 9.46-liter sample was tested after chlorination, but neither E. cold nor poliovirus could be recovered. In Dade County, Fla., the same migrant labor camp that had a water-associated typhoid outbreak in February 1973 (Pfei~er, 1973) was the site of an outbreak of hepatitis A in March 1975 (Wellings et al., 1976~. Samples of the chlorinated water taken at the camp nursery, rectal swabs from children in the nursery, and water from the evaporation pond all yielded echovirus. Poliovirus also was isolated from the evaporation-pond water. The present status of hepatitis A has been summarized as follows by Craun et al. (1976~: There were thirteen outbreaks of waterborne viral hepatitis, affecting 351 people during 1971-1974. Over the past several years, there has been considerable controversy regarding the existence of viruses in treated water supplies and the possible health consequences. Hepatitis A has been epidemiologically-implicated in 66 waterborne outbreaks since 1946, and the data can be examined to determine how these outbreaks occurred (Table III-5. Of the 22 outbreaks occurring in municipal systems, three resulted from either inadequate or interrupted disinfection, and five were related to the use of contaminated, untreated surface or groundwater. Half (eleven) of the outbreaks in municipal systems, however, occurred as the result of contamination of the distribution system, primarily through cross-connections and backsiphonage. There is no evidence, however, that the hepatitis A virus has been transmitted through correctly-operated, conventional water-treatment systems, except where distribu- tion defects have been found as the source of the contamination. For the semipublic and individual systems, the use of contaminated, untreated groundwa- ter was the important factor responsible for outbreaks of hepatitis-A.
94 DRINKING WATER AND H"LTH The cited epidemiological and laboratory evidence presents some paradoxes. The virus of hepatitis A has been shown, epidemiologically, to be transmissible in water, but has not been isolated in the laboratory from water samples. Gastroenteritis is transmissible in water; however, at least some of the time the disease is not caused by virus. Clearly, more research is needed to resolve these problems. Recovery and Identification of Viruses FACTORS INFLUENCING RECOVERY More than 100 different serological types of human enteric viruses may appear in wastewater (Berg, 1973~. The chance that one or more of these viruses will appear in a community source of potable water is steadily increasing, as demands on available water make reuse increasingly necessary (Berg et al., 1976~. According to an EPA study of 155 cities that use surface-water supplies, 1 of every 30 gal that enter water-treatment plants has already passed through the wastewater system of a community that is upstream (Culp et al., 1973~. With the increasing use of renovated water, there is a need for methods that can detect the enteric viruses that might occur in raw or finished water. Detection methods based on virus infectivity are more sensitive than immunochemical procedures, because more virus is needed to produce a perceptible immunochemical reaction than would ordinarily be needed to infect a cell culture. The amount of virus present in polluted surface- waters may vary over a wide range. A calculated value of 38 plaque- forming units (PFU) per gal, based on virus-shedding by children under 15 yr old, was derived from data analyzed by Clarke et al. (1964~. A total of 330 PFU/gal was recovered from a stream receiving wastewater effluent (Grinstein et a/., 1970~. The available data suggest that raw potable water sources, subject to pollution from wastewater discharges, may contain significant numbers of enteric viruses. Conventional treatment of a water supply (coagulation, sedimentation, filtration, and disinfection) to produce finished water is expected to result in removal or inactivation of a minimum of 6 logic of virus (Clarke, 1976~. Barring gross contamination of finished water, only low numbers of viruses are likely to occur in properly treated supplies. Such small amounts of virus are likely to be detected only by highly efficient methods. The general quality of the water is important in virus recovery. Turbidity, organic matter, pH, salts, and heavy metals all influence virus recovery. Both organic and inorganic solids suspended in water can
Microbiology of Drinking Water 95 adsorb viruses (Schaub and Sagik, 1975~. In the testing of large volumes of water containing suspended solids, virus adsorption to these solids complicates the recovery process. Sample clarification that results in removal of the solids may also result in removal of virus. Deliberate use of solids for recovery of virus may not always succeed, because of failure to desorb the virus (Berg, 1973~. Organic material in water may influence the virus-recovery process by interfering with the ability of the virus to adsorb to a collecting surface Challis and Melnick, 1967a). The interference results in loss of the virus through penetration of the collecting surface. Organic matter can also protect the virus. Entrapment of enteric virus in fecal clumps would call for special procedures to disrupt the clumps, and thus release the virus. Acid pH and metal cations are known to enhance adsorption of the virus to clay particles (Carlson et al., 1968) and to membrane-filter surfaces (Walks and Melnick, 1967b). Desorption of virus is favored by a pH of 9-11 and by the presence of organic substances (sliver, 1967~. Knowledge of the influence of these factors can be used to avoid loss of the virus during sample collection, or can be applied to procedures to promote the recovery of the virus from water. In summary, detectability of waterborne viruses is limited by: the low concentrations in which the virus occurs, the presence of interfering substances in the water, adsorption or entrapment of the naturally occurring virus, and inadequacies in the laboratory host systems (cell lines, experimental animals) in which isolation must ultimately take place. CELL-CULTURE SYSTEMS FOR DETECTION The types and quantities of viruses that can be detected depend on which cell cultures are used and how virus activity is manifested (plaguing versus cytopathology). Primary-monkey-kidney, or human embryonic- kidney cell cultures were originally recommended for maximal efficiency in detecting enteric viruses (Lee et al., 1965~. One established primate cell line (BGM) was more sensitive than primary primate cultures in detecting viruses in water (Dahling et al., 1974~. In another study, 31 echoviruses were recovered in human, diploid-embryonic lung cells, but not in monkey-kidney cells (Zdrazilek, 19741. A cell line that may eventually reduce the need for suckling mice in detecting cosackievirus A has been evaluated by Schmidt et al.(l9751. Plaquing (agar overlay) and cytopathic (liquid overlay) methods in combination were recommended for isolation of as many enteric viruses as possible (Hatch and Marchetti, 1975~. The
96 DRINKING WATER AND H"LTH TABLE III-6 Enteric Viruses and Host Systems Required for Isolation of Virus Host System Primate Human Nicer Virus Group Number Primary Continuousb Primary' Continuous Enteroviruses Polioviruses 3 + + Coxsackievirus A' 23 Coxsackievirus B Echoviruses2 Reoviruses Adenoviruses Hepatitis A3 Gastroenter~tis viruses3 4 + + + + + + 31 + + + + + + + + + 31 3? + tCoxsackieviruses A-7, 9, 16, and perhaps others may be isolated in cell culture systems. 2Human diploid-embryonic lung cultures were reported useful for isolation. 3No acceptable host-cell system for routine laboratory use has been developed. Identified as acute infectious nonbacterial gastroenter~tis agents (AING). aPrimary monkey-kidney, African Green monolayers (PMK-AG). Buffalo-Green monkey-kidney monolayers (BGM). t primary human embryon~c-kidney monolayers (HEK). dDiploid human embryon~c-lung monolayers (HEL). eSwiss mice, 4 days of age or less. Ideal host system use for maximum virus recoveries includes: (1) PMK-AG, agar and liquid enterovirus, and reovirus, overlay media; (2) HEK, agar and liquid enterovirus overlay media; (3) HEL liquid overlay medium; and (4) suckling mice. enteric viruses and the host systems that are required to isolate them are shown in Table III-6. The best available combination of laboratory host systems will not detect the virus of hepatitis A and will miss other viruses. Cultures that are susceptible to a given virus may diner in sensitivity, and wild strains of a virus may fail to form plaques (or to produce cytopathology in the same way) as laboratory strains. RECOVERY PROCEDURES FOR VIRUS IN FINISHED WATER There are extensive reviews of methods for recovering waterborne viruses (Hill et al., 1971; Sobsey, 1976~. Present methods seem generally adequate
Microbiology of Drinking Water 97 for recovery of enteroviruses, but perhaps not of reoviruses and adenoviruses (Fields and Metcalf, 1975~. Representative recoveries of virus from seeded water-samples are shown in Table III-7. Although there is no universal method for concentrating all enteric viruses from water, some methods are quite effective for some of the enteric viruses in some types of water. Flowthrough-filter adsorption-elusion systems and ultrafiltration meth- ods with anisotropic, polymeric membranes in tangential-flow systems are the best for recovery of small quantities of enterovirus in large volumes of treated or finished water. Virus concentration from water by adsorption to precipitable salts or polyelectrolytes, aqueous polymer two- phase separation, reverse osmosis, hydroextraction, continuous-flow ultracentrifugation, and forced-flow electrophoresis methods can general- ly be applied only to samples of a few liters. Therefore, they are not suited to recovery of a low number of infectious units of virus from large samples of water. IDENTIFICATION OF VIRUSES RECOVERED FROM WATER Serologic methods for virus identification are often based on virus infectivity. In the serum-neutralization test, infectivity is blocked by the action of a homologous antibody. Neutralization tests, using combina- tion pools of enterovirus equine antisera, are one of the reliable ways of identifying an enterovirus isolate (Melnick et al., 1973~. These tests are usually completed within 48-72 h in cell cultures, after recovery of the isolate and passage in cell culture. The total time required for completion of all these procedures depends on virus growth-rates and may vary from a few days to weeks. Immunochemical methods have been evaluated as a more rapid means of virus identification. Immunofluorescence (Katzenelson, 1976) and '.mmunoenzymatic (Kurstak and Morisset, 1974) procedures have been tested. These procedures are based on antigen-antibody interactions whose results can be seen sooner than the visible cytopathic ejects required for interpretation of serum-neutraliza'ion tests. The concept of an indicator bacterium was discussed in some detail earlier in this report. Indicator bacteria are organisms that, although not pathogenic in themselves, are constantly present in intestinal discharges, so that their detection in the environment constitutes a signal for possible fecal pollution. Is it possible to consider certain viruses as indicator viruses? Such indicator viruses might provide a clue to the inadequacy of source-waters, or water-treatment systems, in terms of virus contamina
98 - a~ 4- 3 o an · _ ._ so o o _ ~ Cal V, .~ o .~ At: a' so ,4 ._ V, ._ s~ o 2 . _ ._ Ct ~: s~ 1 - LU m - a~ ;^ _ ~ O .E ;> ~: o ~ _ t: ' ~ ,1 O ~ C~ ~ V, ~ ;> O C O - _ ~ ~ _ O ~ O ~ O ;> ~ _ ~ Ct ~ ~ O O _ O O O I ~ O ~ O ~ O _ O ~ =\ _ G ou - Le o 3 ~ ., ~ Cd, _ Ct <: o o~ - ~ o _ o~\ . c oN .Q c - : ~ - . C 3 3 o 3 g - o ~ ~ o C~ oo ') - . - V _ 3 ~ 1 - - V L. s: ~ _ ~C o, O O r~ _ 0 1 o V) a~ 00 - w' , ~ . _ ~ _ :s ~ <,, oo ~ O ~ ~ _ v v Ct v C au _ ~ C~ ~ o V a~ - -V ~ O ~ -v ~ ~ ~ au 3 ~ 3 Z ~ U) V) oo ~ 5\ ~ ~ V) ~ 00 ~ 00 ~ r~ \0 X Xo g - X V~) ~o X o _ _ _ o _ C~- - ~ O ~ - oo CO - o ~ 5\ ~ _ _ ~ ~ _ ~ _ ~ - , _ ~ ~ _ ~P~ =~= m - = ~ == op O i ~ O . . . - 8 ~ - ~ ~. o o o ~ ~ I o _ I . . . . . - _ _ 0 0 r~ - O r~ oe Cd 3, . .' v . _ x o v ~C c ·E V _ ~ ~ 6 cL ~ c A ._ L~ - Ct ~ _~ ~ v 3 ~ ~ ~d ~ ~ · ct ~c: ~ 3 0 0 E m "
99 c: ~ - - ~ ~ C ~ ~ ~ l~ ~` 0\ _ ~ ~ ~ ~ ~ , 3 eC - C~ O C-~ - c~ D V~ U~ ~ ~ ~ ~ ~ ~ ~ ~ - 1 1 - ~ o r~ 0 r~ ~ oo ~ ~ ~ ~ r~ X ~ 1- O - 1 1 o X ~ ', _ _ \o V) O g O \o ~ _ - -I ·~) 1 \0 ~ O O I I _ o 0 - ~ __~_______ O O O O ~ O O O 1 · · 1 · · O O - ~ O O r~ c~ O O O - - _ _ _ - r~ - O _ O O t- V) o ..... 1 O - - ~ O ~ oo ~ X _ ~ \o I_ r~ ~ r~ - . = ~ ~;2 _ C~ =, - - .- Cd ~ . ~1 O .0 .0 E ~ ~ o C. o ~ X ~ o ~ 3 ~ Ct ~ LL a~ c 4) o C~ .o - Ct o x x - x - . Ck ~3 3 C'S v, _ ~ ~ ~ ~, ~ ~D 3 ~ .~ ~ ~, a C O ._ . 00 ·~ ~ L~ 3 t0 ,~ .° <\, g ~ <: 0 - ~ _ E . ·~: ~ LL] D ~P. C 3 D e ct O ~ ~ E ~c~ E ^~e ed ~ O ~C~ >` C D ON r ~t ~c~ ~_ O _ u~ ~, ~O ~ ,~, E ~ ~, ~ 0 ~O c c ~ c.~ ~m O <: ~° ,, ~ x -.E ". j e Z C ~' e ~ ' ~ =' ~ ~' 3 3 ,#
100 DRINKING WATER AND HEALTH lion or removal, and hence could be useful in the control of waterborne . , . virus alsease. If a rapid serologic-identification test were available, it might be adaptable to detection of some indicator virus whose presence has correlated with that of other enteric viruses. The polioviruses (which are common in wastewater and easy to detect) have been proposed as indicators of other enteric viruses that are more difficult or impossible to detect by laboratory methods. A rapid immuno~hemical method for detecting polioviruses might be attainable, but this does not validate the use of polioviruses as indicators. Although polioviruses are the most common group of enteric viruses recovered from polluted waters, they are not always predominant or even present among the enteric viruses in field samples (Grinstein et al., 1970; Metcalf et al., 1974b). Because recovery methods are essentially the same, whether for polioviruses or for all the enteric viruses, there is little to be gained by adopting such an . .. . Indicator virus now. SENSITIVITY OF THE FLOWTHROUGH METHOD Laboratory-based trials of flowthrough equipment and methods, using at least 100 gal of tapwater experimentally contaminated with poliovirus, regularly yielded positive results from SOO-gal samples that contained 3-S PFU/100 gal, and sometimes only 1-2 PFU/100 gal (Hill et al., 1976~. Members of the reovirus, echovirus, and coxsackievirus groups A and 8, at inputs of 1 PFU/gal or less, were also recovered when 100 or SOO gal of water were sampled. Emphasis was placed on ability to recover small quantities of virus in large volumes of finished water. Performance at extremely low concentrations of virus was considered a better index of the merits of a method than high-percentage recoveries based on very large test doses. The success of the laboratory trials has led to inclusion of flowthrough procedures in the fourteenth edition of Standard Methods for the Examination of floater and Wastewater (1975) as a tentative method for finished water. PERSPECTIVES IN TESTING LARGE-VOLUME SAMPLES Some of the assets of the flowthrough method may also be viewed as liabilities. The portable concentrator permits the testing of large-volume samples without transporting great quantities of water to the laboratory; however, a probable consequence is that the laboratory virologist is not answerable for the validity of the field-sampling procedures. The method enables detection of as little as 1 PFU of virus in a 100-gal water sample,
Microbiology of Drinking Water 101 but the authenticity and significance of a positive test result is question- able when very few PFU are found. The method can detect only viruses for which susceptible tissue cultures or laboratory animals are available, so that the agents of hepatitis A and most viral gastroenteritides will be missed. The vaccine polioviruses have the highest overall probability of occurrence and detection, but they do not themselves constitute a threat to human health. The vaccine polioviruses, when compared to coliform organisms, may constitute a more sensitive or valid indicator of the threat of waterborne virus pathogens; however, this kind of comparison should be made only after more colifo~m tests have been performed on much larger samples than lOO ml. Both the flowthrough method of virus collection and larger-volume coliform tests may have a future as standard methods for discretionary, intensive testing of finished water; neither should be regarded as a routine monitoring method for on-line quality control. Health Effects of Viruses in Drinking Water Large numbers of enteric viruses are present in some human feces and, therefore, in wastewater. These viruses may not be completely removed or inactivated by wastewater treatment, so they may be discharged to surface water that serves as another community's raw-water source (Clarke et al., 1964~. If the viruses are not completely removed or inactivated by water treatment, they may be ingested. People infected by these ingested viruses do not always become ill, but disease is a possibility in persons infected with the agent of hepatitis A or any of the other enteric viruses. Several diseases involving the central nervous system, and more rarely the skin and heart, are caused by the better-characterized enteroviruses: polioviruses, coxsackieviruses, and echoviruses. Some reoviruses and parvoviruses have been implicated in nonbacterial gastroenteritides. More than 100 serotypes of enteric viruses are known and are recovered from wastewater from time to time (Davis et al., 1967~. Sporadic occurrences of echoviruses and coxsackieviruses in a susceptible population may result from transmission of the viruses through water, but this has yet to be proven. Poliovirus infection by the oral route has been studied to some extent, especially in connection with the development of attenuated poliomyelitis vaccines for orally administration. [he remainder of the enteric viruses have been used only to a small extent in research of this kind. In the few studies of other orally administered viruses, there was usually no attempt to determine how much virus was needed to produce infection, or even how much virus was being administered.
102 DRINKING WATER AND HEALTH Some of the particles in any collection of virus are perceptibly defective. The rest apparently possess all the constituents necessary to initiate infection of a cell and, by successive replication, of a host organism; such a particle might be described as "nondefective." A nondefective virus particle, inoculated into a culture of susceptible cells, may or may not cause an infection. Incipient infections may be aborted in one way or another, so the number of apparently nondefective particles greatly exceeds the number of tissue-culture infections they can produce. Schwerdt's (1957) ratios of particles to tissue-culture doses (the quantities of virus required to produce perceptible infections in a tissue- culture, whether the infections are perceived by cytopathic ejects, plaque formation, or some other manifestation) were about 36:1 to over 100:1, whereas Joklik and Darnell (1961) reported a ratio of approximately 250:1. The latter study showed that the particles that had not succeeded in initiating infections were not defective; but the ways in which incipient infections were seen to abort would not account for an infectivity ratio as high as 250:1. However, only one poliovirus particle ordinarily infects a cell. Thousands of nondefective, genetically identical progeny particles are produced in replication. Once a cell is infected, its neighbors in a culture are virtually certain to become involved. Nevertheless, if small numbers of nondefective particles have been inoculated at the outset, infection of a first cell is unlikely to occur. Poliovirus infection by the oral route had to be studied during development of the "live-virus" poliomyelitis vaccines. Sabin found that nonhuman primates could not be substituted for man in such investiga- tions, because of species differences in the relative susceptibility of the intestines to poliovirus infection. He reported that, if fewer than 105 tissue-culture doses of vaccine poliovirus were ingested by a human, the virus would bypass the pharynx, but would infect the intestines (1957~. Ratios of physical particles to oral, infectious doses for man, comparable with those for tissue-culture doses, have not been reported. Orally administered poliovirus doses have therefore been expressed in terms of tissue-culture doses, even though infections of these two host systems have never been shown to be analogous. Indeed, the disagree- ment (Bodian, 1957; Sabin, 1957) over which intestinal cells are responsible for replication of poliovirus still seems to be unresolved. Horstmann (1961) described 105 tissue-culture doses as "an average minimal reliable dose for most strains," and thought it remarkable that some strains would infect by the oral route when fewer than 104 tissue- culture doses had been ingested. The subjects in several of the early poliomyelitis-vaccine trials were
Microbiology of Drinking Water 103 TABLE III-8 Results of Feeding Various Quantities of Poliomyelitis Vaccine to Infants No. Infected/ Percent Dose No. Fed Infected References pFua 0.2 0/2 0 Koprowski, 1956 2 2/3 66 Koprowski, 1956 20 4/4 100 Koprowski, 1956 5.5 x 106 16/18 89 Holguin et al., 1962 TCD.;ob 03.5 28/97 29 Lepow et al., 1962 04.5 42/96 46 Lepow et al., 1962 05.5 48/84 57 Lepow et al., 1962 lo6.6 12/20 60 Krugman et al., 1961 low 15/20 75 Krugman et al., 1961 103.5 208/308 68 Warren et al., 1964 105~5 133/169 79 Warren et al., 1964 1 3/10 30 Katz and Plotkin, 1967 2.5 3/9 33 Katz and Plotkin, 1967 10 2/3 67 Katz and Plotkin, 1967 05 5 4/8 50 Gelfand et al., 1960 107 5 9/9 100 Gelfand et al., 1960 QPlaque-forming units. hTcD50 = Tissue culture dose Two. infants in their first few days or months of life. They included babies born to mothers who had various serum concentrations of antibody against polioviruses. The criterion of infection was the shedding of the virus in the infant's feces after 1 week. Though large amounts of virus were often administered, not all vaccinated infants became infected. The findings of several investigators are summarized in Table III-8. A study by Koprowski (1956) showed that two of three children, each fed 2 PFU of vaccine poliovirus in enteric-coated capsules, became infected. Plotkin and Katz (1967) presented data that they believe demonstrated a parity between the quantity of vaccine poliovirus needed to infect a tissue- culture and that which will infect a human subject by the oral route. Studies by others indicated a need for more than 104 tissue-culture doses of vaccine in infants (Gelfand et al., 1961~. The available data on polioviruses have important limitations; unfortunately, far less information seems to be available concerning infectivity, by the oral route, of other intestinal viruses. The dose of coxsackievirus B5 that would infect 50% of newborn mice to which it was
104 DRINKING WATER AND HEALTH administered by the oral route was between 102 and 103 tissue-culture doses in studies by Loria et al. (1974~. Their criterion of infection was disease, so they were measuring the pathogenic dose, rather than the infecting dose. No other reports of attempts to determine the minimal infecting dose of ingested human intestinal viruses seem to be available- whether for humans or other species in terms of tissue-culture doses. In sum, poliovirus infection at the cellular level appears to be initiated by a single nondefective particle. If the ratio of nondefective particles to infections is the same for ingested virus, this is fortuitous: the yet-to-be- identified cells in the intestines to which the particles must attach may be either more or less susceptible than tissue-culture cells, just as the susceptibility of different tissue-culture cells may vary. Counts of virus particles are not likely to be useful in this application, and some important intestinal viruses cannot be enumerated in tissue-cultures at all. There is no valid basis for establishing a no-effect concentration for viral contamination of finished drinking water. Virus Removal in Water Treatment When viral disease occurs in association with a drinking-water supply, one first has to consider the possibility that cross-contamination or back- siphonage may have admitted waste water into the distribution system. However, this source will not be discussed here; the integrity of water- distribution systems is a technological, rather than a virological, problem. If the integrity of a system is not maintained, the most immediate consequences are more likely to involve enteric bacteria than viruses. Alternatively, viruses may occur in a community's raw water supply as a result of intended or inadvertent "recycling" of wastewater. If viruses are not removed or inactivated (a virus is said to be inactivated when it loses its ability to produce infection) in the treatment of the water, disease may occur among those who drink the finished product. The required degree of virus removal or inactivation cannot be stated explicitly, because of lack of good information as to how much virus must be ingested to cause disease. When goals in virus removal are considered, it is also important to remember that less than 1% of water consumed by a community is used internally by humans; more than 40 gal of water per person per day are used for household purposes (Wits et al., 1975~. One would like to know what amounts of virus might occur in raw water at the point of intake. Most reports of virus in raw water are not quantitative: for example, Foliguet et al. (1966a) reported detecting enteroviruses in 13 of 162 raw-water samples tested in France during 1961- 1963, but one cannot determine the volume of water that each sample
Microbiology of Drinking Water 105 represented. The highest virus concentrations measured at or near water intakes by Berg (1973) were somewhat less than 0.2 PFU/liter of water. Available data (Chang 1968) indicate that 30 PFU/liter constitutes a "reasonable virus load in moderately polluted water." One would hope to remove or inactivate all viruses before the treated drinking water were considered "finished." Berg (1973) has offered to resolve this difficulty by requiring that reclaimed and other potable water be disinfected so as to destroy at least 12 logarithmic units of a reference virus at 5°C and that the finished water be tested frequently to ensure that there is not more than 1 PFU in a 100-gal (379-liter) sample of the final product. An apparatus that will accommodate such samples has been developed. Recently, 4 of 12 samples (100-105 gal, or 379-397 liters each, of finished water from a modern, well-operated plant serving 600,000 people in Fairfax County, Va. were shown to contain 1-4 PFU of poliovirus each (T.G. Metcalf, unpublished data). The public-health significance of such findings is uncertain. The infectious dose of poliovirus by the oral route is not yet known, and the detected virus was apparently of vaccine origin and therefore nonpathogenic. Any virus detected in finished drinking water may indicate that other viruses, not detectable by laboratory techniques, are also present; however, no outbreak of human illness has been associated with the water supply that was examined. The detection of very few tissue-culture doses of virus in extremely large samples does not yet support the adoption of a numerical specification for the volume of finished water from which detectable virus must be absent, or for the projected viricidal effectiveness of a disinfection process. Safety of drinking water, from the virus standpoint, cannot at present be defined in numerical terms. None of the unit operations and processes used today in the treatment of water supplies were devised originally to remove or destroy viruses. Still, almost all these treatments have some antiviral eject. The potential antiviral electiveness of disinfection (chlorination) probably exceeds that of all other available treatments combined. Any treatment for which there are substantial data to indicate ability to remove at least 50% of incident-viruses will be considered here. Many of the data have been obtained from bench-scale studies with experimentally inoculated model viruses. Accurate measurement of virus removal (by a treatment being evaluated) is possible only if the initial concentrations of viruses are very high. The required concentrations of viruses would not occur naturally in water destined for human consumption (except in reclaimed wastewater), and would be difficult to attain by adding laboratory viruses to pilot-scale batches of water. Such experiments are conducted under more closely
106 DRINKING WATER AND HEALTH controlled conditions than may be attainable in practice, and the aggregation that is always present in virus suspensions may differ in degree between inoculated virus and naturally occurring virus. Results will not be reported here to a greater degree of precision than seems reasonable to expect in practice. COAGULATION AND SETTLING An early step in the treatment of raw water often includes coagulating suspended solids and allowing them to settle to the bottom of a tank. The supernatant water is then collected for further treatment. Substances added to induce coagulation may include primary flocculants (such as aluminum sulfate and ferric chloride,j, coagulant aids (polyelectrolytes), acids, alkalis, and even clay used for supplemental turbidity. Four recent, extensive studies indicated that virus removal from 90% to well above 99% can be achieved by coagulation and settling under carefully controlled conditions (Foliguet and Doncoeur, 1975; Manwaring et al., 1971; Schaub and Sagik, 1975; and Thorup et al., 1970~. Either aluminum sulfate or ferric chloride proved a good primary floccueant for virus removal. Polycationic coagulation aids were useful in removing virus, but anionic or nonionic polyelectrolytes probably would not be elective. Clays are sometimes added, to ensure formation of a visible floe to which virus may adsorb. Organic matter in the water interfered with virus removal by coagulation and settling; this eject may be prevented by prechlorination or preozonation. Finally, it should be noted that the virus removed had not been inactivated. Virus adsorbed to floe was quite capable of causing infection. FILTRATION AND SORPTION Virus particles are too small to be retained mechanically on sand, or on most alternate media used in water filtration. Virus retention by such media depends on association of the virus with suspended matter that is large enough to be trapped mechanically, or by direct sorption of free virus particles to the surface of the filter medium. When sand filtration follows coagulation and settling in the treatment scheme, viruses that sorbed onto fine floe particles can be retained efficiently by sand (Foliguet and Doncoeur, 1975~. Virus removal by the filter ranges from 90~O to more than 99%, with the very highest figures resulting when the virus was poorly removed by previous settling of the flee. Sand, by itself, is reputed to be a poor medium for removing viruses from water (Chang, 1968~. However, Lefler and Kott (1974) removed
Microbiology of Drinking Water 107 more than 99% of their model viruses by slow sand filtration when the suspending water contained calcium chloride and magnesium chloride at 0.01 N. Diatomaceous earth has been evaluated by Brown et al. (1974) as a filter medium for removing viruses from water. The medium was coated with ferric hydroxide or a synthetic polymer developed for water treatment. The model virus was removed from dechlorinated tapwater (pH 9.5) to the extent of approximately 98%, but the process was less elective at a lower pH (6.7-6.8), or in the absence of coating. It is not clear whether this filtration process was intended principally to follow or to supplant the coagulation and settling-steps in water treatment. Activated carbon is another possible sorbant medium for removing viruses. Cliver (1971) found that retention of virus on the medium could be 90% initially, but was apparently lower after a column had been operated for some time. Relatively little Resorption seemed to occur when water, rather than wastewater, was treated. The possibility of virus Resorption from a filter-medium, or of resuspension of floe to which virus has adsorbed, must always be considered in evaluating treatments such as those described above. None of these treatments appears to inactivate the virus that has been retained. Indeed, if any of these treatments proves less elective in practice than was indicated in bench-scale studies, it may well be because of Resorption or resuspension, rather than inadequate initial retention of virus. WATER-SOFTENING Softening of water at the treatment plant may inactivate virus, rather than simply remove it. Sprout (1971) summarized research in his laboratory that showed modest virus removal with straight-lime soften- ing, and more than 99% removal with sodium hydroxide precipitation of magnesium ions, with excess lime-soda ash softening. He used the term "inactivation" to describe this removal, but he compared data from his and other laboratories which indicated that viral infectivity is lost only in some cases at a pH near 11, which occurs in excess lime-soda ash softening. DISINFECTION If all of the treatments described above were applied sequentially to raw water, one could expect a total of 6 logarithmic units of virus removal or destruction, without recourse to chemical disinfection. However, chemi- cal disinfection seems to be more reliable than the other treatments. Most
108 DRINKING WATER AND HEALTH chemical disinfectants that are elective against viruses are strong oxidizing agents. The general principles of action of these substances on viruses have been discussed by various authors, not all of whom reached the same conclusions. Liu et al. (1971) emphasized differences in resistance among the groups of enteric viruses (as well as the importance of clumping of virus particles) as a source of confusion in interpreting inactivation data. Given the contact times likely to be used, these differences are probably not critical to the success of an efficient disinfectant. However, Kruse et al. (1971) emphasized that not all slowly effective forms of disinfectants can be made to accomplish their task by extending contact time. Treatment prior to terminal disinfection accom- plishes two purposes reducing the virus load and preparing water for more effective terminal disinfection, by removal of interfering substances. CHLORINATION Chlorine, applied in its elemental form or as hypochlorite, is the standard of disinfection against which others are compared. Depending on the pH of the water and on the presence of ammonia, the chlorine may take the form of HOC1, OC1-, C12, or chloramines. The viricidal electiveness of these forms apparently decreases in the order listed. Neefe et al. (1947) studied chlorine inactivation of the virus of infectious hepatitis (in drinking water that was given to human volun- teers). The results were not quantitative, but they showed that, if viruses were added (as a suspension of infectious feces) to water that was then coagulated and filtered, a chlorine dose sufficient to yield a free residue of 0.4 mg/1 after 30 min of contact would render the virus noninfectious for the volunteers. Nothing has been done to refute or augment these findings in the intervening period of almost 30 yr. Clarke and Kabler (1954) showed that coxsackievirus A2 was inactivat- ed more slowly by chlorine at a pH of 9 than at a pH of 7, and more slowly at 3-6°C than at 27-29°C. Not surprisingly, the inactivation rate also depended on the concentration of free chlorine (hypochlorous acid and hypochlorite ion). Kelly and Sanderson (1958, 1960) extended these results to other human enteroviruses and showed that the antiviral action of chloramines is a great deal slower than that of free chlorine. Clarke et al. (1956) found adenovirus 3 to be more susceptible to chlorine than the enteroviruses; destruction rate of the former was said to be comparable to that of Escherichia colt. Liu et al. (1971) showed that the reoviruses were even more chlorine-sensitive than the adenoviruses, and that some of the differences in chlorine sensitivity among enteroviruses might be related to the degree of aggregation of the viruses in suspension.
Microbiology of Drinking Water 109 The aggregation always seen in virus suspensions imposes significant limits on the predictability of virus inactivation (Sharp et al., 1975~. Hypochlorous acid, which predominates below a pH of 7.5, is more active against viruses than is hypochlorite ion (Chang, 1968~. When the water contains nitrogenous substances, chlorine should be added to the "breakpoint," beyond which any further added chlorine will be free to inactivate viruses (Kruse et al., 1971~. Chlorine dioxide (Cl02) may offer some advantages over other forms of chlorine because it is less reactive with ammonia, and less affected by temperature and pH (Chang, 1968~. BRO MI N ATI ON Bromine has been studied (to a limited extent) as an alternative to chlorine in inactivating waterborne virus (Taylor and Johnson, 1974~. It is clearly effective, and it may be somewhat less subject than chlorine to "demand" losses in use. Bromamine, for instance, is capable of significant virus inactivation. More study will be required before bromine can be considered likely to supplant chlorine as a general disinfectant. It does have advantages for some specialized uses. IODINATION Iodine is also capable of inactivating enteroviruses in water under controlled conditions (Berg et al., 1964~. Although relatively high doses may be required, Chang (1968) suggested that iodine may be useful in special situations. Studies by Olivieri et al. (1975) on the comparative modes of action of the halogens on a small bacterial virus (f2) indicated that iodine reacts with the coat protein, whereas chlorine probably inactivates the viral nucleic acid. OZONATION Ozone is finding increasing use as a water disinfectant and is proving effective against viruses. Although Majumdar et al. (1973) found that ozone was relatively ineffective against poliovirus when the disinfectant was present at less than 1 mg/liter, Katzenelson et al. (1974) found ozone to be faster and more elective than chlorine against poliovirus, even at concentrations as low as 0.3 mg/liter. Ozone does seem to over significant advantages as an antiviral disinfectant for water. Its most conspicuous liability is that its activity cannot be sustained in the water all the way to the consumer, so that it does not provide any protection against post-treatment contamination.
110 DRINKING WATER AND HEALTH Effectiveness of Water Treatments Virus may occur in some raw water; standard treatments, such as coagulation and settling followed by rapid sand filtration (if carefully performed under practical conditions) could remove a combined total of 6 logarithmic units of virus present. Viruses removed in these ways are not inactivated. A virus that is still present can be inactivated by chemical disinfection. Ozone and chlorine seem most clearly suited to community use for water disinfection. With raw water that sometimes contained virus, both ozone and chlorine were found to produce finished water in which viruses could not be detected by the methods used (Foliguet et al., 1966b). In general, it appears that current treatment technology, diligently applied, can consistently produce finished water in which viruses are not likely to pose a threat to public health (Clarke et al., 1974~. The theoretical model for virus removal or inactivation by each treatment assumes that a zero concentration of virus will never be attained. This implies that testing ever-larger samples of finished water will, in some instances, result in detection of a virus. If the presence of virus in finished water results from undertreatment of virus-contaminated raw water, treatment should be intensified (to include coagulation, settling, filtration, and chemical disinfection, such as breakpoint chlorin- ation, if these are not already being done), and efforts should be made to alleviate the contamination of raw water. Bacterial Indicators and Viruses in Drinking Water It should be clear from the preceding discussion that virologic tests are not "routine" in the sense that many bacteriological methods are routine. Can bacteriological methods be of any value in ensuring the virological safety of drinking water? Viruses diner fundamentally from bacteria in size and in biological properties. Decreases in concentrations of bacterial indicators during water treatment ought not to be expected to correlate directly with virus removal or inactivation. Competent execution of water-treatment meth- ods ultimately determines the safety of the finished product. In the event of a treatment break or a loss of integrity of the distribution system, the presence of excessive turbidity, coliforms, or standard plate counts will indicate the existence of an unsafe situation and a possible viral hazard. Even though these indicators are only incidentally correlated with the presence of virus in drinking water, they
Microbiology of Drinking Water 111 are important, because they will evoke remedial action. Detection of viruses themselves in drinking water will take at least 2 days, in the case of gross contamination, and 1-2 weeks under most circumstances. The results of virological tests on contaminated water are unlikely to be known before human illness occurs. Microbiological indicators may have a limited correlation with viral contamination, but they and chlorination afford more protection to public health than does any available alternative approach to routine monitoring of finished water. Conclusions 1. The presence of infective virus in drinking water is a potential hazard to the public health, and there is no valid basis on which a no- e~ect concentration of viral contamination in finished drinking water might be established. 2. Continued testing for viral contamination of potable water should be carried out with the facilities and skills of a wide variety of research establishments, both inside and outside the government, and methodolo- gy for virus testing should be improved. 3. The bacteriological monitoring methods currently prescribed or recommended in this report (coliform count and standard plate count) are the best indicators available today for routine use in evaluating the presence in water of intestinal pathogens including viruses. Research Recommendations 1. Improved methods should be developed for recovery, isolation, and enumeration of viruses from water supplies. 2. A laboratory method should be devised for detecting the virus of hepatitis A in potable water. 3. More should be learned about the etiology of viral gastroenteritis; special attention should be given to detection methods for gastroenteritis viruses that are transmissible through water. 4. The amount of virus that must be ingested in drinking water to produce infections and disease should be determined for several different enteric viruses. 5. Additional research should be conducted on the ability of various water-treatment methods to remove or inactivate viruses. Low-cost modifications to increase the reliability and electiveness of existing methods should be sought.
112 DRINKING WATER AND HEALTH PATHOGENIC PROTOZOA AND HELMINTE1S Eggs and cysts of parasitic protozoa and helminths are deposited in the environment with excrete, and may enter water supplies. Water- purification processes have not been developed on the basis of considera- tions related to these organisms; but, fortuitously, the introduction of flocculation, filtration, and chlorination of water have been successful in diminishing the extent to which water serves as a source of parasitic infection. Perhaps more than water treatment, sanitary sewerage systems have diminished the spread of various intestinal helminths. Eggs of these helminths are not easily killed in sewage-treatment processes (Newton et al., 1948; Greenberg and Dean, 1958; World Health Organization, 1964~. Since there is a tendency for helminth eggs to be concentrated in sewage sludges, there is some concern when sludges are used in agriculture (Greenberg and Dean, 1958; World Health Organization, 1964~. There are specific drinking-water problems with protozoan parasites, such as Entamoeba histolytica, the cause of amebic dysentery and amebic hepatitis, and Giardia lamblia, a flagellate about whose pathogenicity there is no longer any serious doubt (Moore et al., 1969; Wolfe, 1975; Wanner et al., 1963; Zamcheck et al., 1963; and, Brandborg et al., 1967~. The waterborne-disease outbreak with the largest number of cases reported to have occurred in the United States in 1974 was due to Giardia lamblia (Center for Disease Control, 1976a). Attention will be focussed here on the protozoan organisms that represent a threat to human health. The intestinal helminths will be considered, especially in terms of the characteristics that make then susceptible to elimination from sewage effluents and that also make them unlikely to be found in raw or finished water. Protozoa Parasitic protozoa replicate in the human host, and may be responsible for severe disease. Entamoeba histolytica (Craig, 1934; World Health Organization, 1969) has been found to be responsible for severe outbreaks of dysentery. It is also capable of setting up chronic infections in the human host, with the eventual development of abscesses of the liver and occasionally other organs. Giardia lamblia is not so severe a pathogen, but is responsible for gastrointestinal disturbances, flatulence, diarrhea, and discomfort. Both these organisms are able--on occasion to penetrate our sanitary barriers.
Microbiology of Drinking Water 113 AMEBIASIS During the period 1946-1970, there were five reported outbreaks of amebiasis due to Entamoeba histolytica transmitted by water. Four of these outbreaks were related to private distribution systems, and involved a total of 50 clinical cases. The well-studied South Bend outbreak in 1953, in which there were at least 750 infections and 30 clinical cases, with 4 fatalities, was due to sewage contamination of a private water supply to a factory (LeMaistre et al., 1956~. One outbreak involved a public system and accounted for 25 cases (Craun and McCabe, 1973~. Before 1946, there were two major amebiasis outbreaks due to contaminated water. The well-known Chicago epidemic of 1933 was traced to cross-connections between sewage and water lines in a hotel (Bundesen et al., 1936~. The Mantetsu-apartment-building outbreak in 1947 (Ritchie and Davis, 1948) in Tokyo was similarly traced to sewage contamination of water due to faulty distribution systems. FACULTATIVELY PARASITIC AMEBAE In recent years, a relatively large number of cases of meningoencephalitis have been reported as caused by free-living, facultatively parasitic amebae of the genera Naegleria, Hartmanella, and Acanthamoeba. Most of these cases have been related to swimming in fresh water ponds or swimming pools. However, more recently (Baylis et al., 1936; Robert and Rorke, 1973) Naegleria fowleri and Acanthamaeba sp. have been isolated from tapwater in association with cases of primary amebic meningoen- cephalitis. It is possible that the occurrence of these fresh water amebae will increase in surface water because of the leaching of fertilizers from agricultural lands, and the other factors that are contributing to accelerated eutrophication of ponds and lakes. Viable cysts, identified as Naegleria and Hartmanella species, have been found in 8 of 15 finished- water supplies in a survey of large-city supplies across the United States (Robert and Rorke, 1973~. The cyst densities were low, about 10-15/gal. Little is known about the characteristics of the cysts and trophozoites of these organisms. It seems likely that the cysts will be at least as resistant to chlorine as those of the parasitic amebae and flagellates. Therefore, considerable emphasis must be placed on flocculation and filtration processes, in order to to remove these organisms from the waters in which they may be proliferating.
114 DRINKING WATER AND HEALTH GIARDIASIS Giardiasis is emerging as a major waterborne disease. The largest outbreak of giardiasis in the United States occurred in Rome, N.Y. (Center for Disease Control, 19751. There were 395 cases of symptomatic giardiasis identified by stool examinations in the 7-month period from November 1, 1974, to June 7, 1975. A random household survey, conducted during one week in early May 1975, identified 150 stool- positive cases of giardiasis among 1,421 persons. On the basis of the survey data, the attack rate was 10.6~o, and more than 4,800 persons may have been ill; a larger number probably harbored asymptomatic infections. Ten samples of the raw water were collected with a large- volume water sampler and were fed to 10 pathogen-free beagle puppies, of which 2, fed from different samples, became infected. The Center for Disease Control's diagnostic parasitology laboratory identified one G. Iamblia cyst in one of the water samples. The Rome water-supply system did not maintain a free chlorine residual, but used only combined residual disinfection of surface water. Information on this and other Ciardia outbreaks is reported in Foodborne and Waterborne Disease Outbreaks, Annual Summary 1974, Center for Disease Control (1976a). Veazie (1969) noted the occurrence of some 50,000 cases of gastroenter- itis in Portland from October 1954 to March 1955. Some of these cases were attributed to infection with giardiae. A report of this outbreak appeared in the form of a letter to the editor in the New England Journal of Medicine, after publication of the study on the Aspen outbreak (Moore et al., 1969~. Waterborne outbreaks of giardiasis in the United States numbered 14 between 1969 and 1975 and involved at least 700 diagnosed cases. Of these, 12 outbreaks,occurred during the period 1971-1974 (Craun 1975) and involved over 300 people. This apparent increase in incidence may be due to greater awareness on the part of physicians; much publicity was accorded to outbreaks of Giardia infections among tourists in Leningrad, starting in 1970. According to Schulz (1975), a questionnaire survey of 1,419 travelers to the USSR showed that 324 (23%) had acquired giardiasis. The mean time between entry into the USSR and the appearance of symptoms was 14 days, and the illness lasted 6 weeks. Leningrad was identified as the site of infection, and attack rates were highest among travelers who drank tapwater. All the U.S. outbreaks involving municipal systems, except for a large outbreak at Aspen, Colorado, in 1965-1966, have been associated with surface-water sources where disinfection was the only treatment. At Aspen, it is apparent that wells that served as a source of water for part of
Microbiology of Drinking Water 115 the community were contaminated from old, broken sewer lines that passed close to them (Moore et al., 1969~. Helminths In the United States the most important intestinal worms that are transmitted in drinking water include Ascaris lumbricoides, the stomach worm; Trichuris trichiura, the whipworm; Ancylostow~a duodenale and Necator americanus, the hookworms; and Stron~loides stercoralis, the threadworm. All of these are nematodes, or roundworms. The stomach worm and whipwo'~ are transmitted directly from one host to another by their eggs, after the eggs have had an opportunity to develop outside the body; the development involves the formation of an infective larva from the embryo contained in the egg. The hookworms live in the soil, grow, and undergo two molts, until they reach a stage which is infectious to man. They usually gain entry into their new host by penetrating the skin and then wandering through the body until they mature in the intestinal tract. The threadworm, like the hookworm, produces an infective larva that invades through the skin. It has the option of passing through one or several free-living generations before producing infective larvae; therefore, it can expand the number of infective larvae derived from each egg. In addition to the nematodes described above, there is one cestode, or flatworm, that takes up residence in the human intestinal tract when it can. This worm is Hymenolepis nana, the dwarf tapeworm. It is the only one of the cestodes infecting humans that has a direct life cycle; i.e., the egg serves as the infecting agent by the oral route. The larva and adult develop from the egg. The adult produces a string of proglottides, up to 200; eggs are produced in each. The gravid proglottides often rupture in the intestinal tract of the host, setting the eggs free. The eggs are immediately infective when passed in the feces. All other important helminthic parasites of man require intermediate hosts for the development of larval stages infective to man. Except for the guinea worm, Dracunculus medinensis, which develops as an infective larva in the water flea, Cyclops, the ingestion of water is a minor factor in the spread of these parasites from host to host. D. medinensis is not endemic in the United States. The eggs of the stomach worm and the whipworm measure 60 x 45 ,um and 52 x 23 ~m, respectively. Hookworm and threadworm eggs are similar in size. The egg of the dwarf tapeworm is about 44 x 37 ,um. All of these eggs are denser than water. They are also of such a size that they can be entrapped in sewage-treatment plants and in the sand filters of
116 DRINKING WATER AND HEALTH water-purification plants. However, motile larvae of the hookworms and of the threadworm are capable of moving through sand filters in water- purification plants, as has been shown to be the case with a number of free-living nematodes. Like the free-living nematodes, and possibly to a greater extent because infective larvae are resistant to the environment and do not ingest food, the larvae of these parasitic worms can be expected to survive usual chlorination procedures. All the parasites considered above are soil-transmitted parasites. It is possible that surface runoff during heavy rains can bring eggs or larvae into raw-water sources. In treatment, however, their large size practically ensures that they will be entrapped in flocculation and filtration processes. Viable hookworm and threadworm larvae, because of their mobility, might be able to traverse sand filters. Because of an enormous dilution effect, and because even the larvae tend to settle out in water, one would be hard-put to attribute a human infection with one of these parasites to a water source, other than wells polluted from the surface. One other factor should be mentioned: the worm burden acquired by a host is directly related to the number of infective eggs or larvae to which the host is exposed; there is no replication of the worms in the definitive host. Therefore, the odds against the establishment of a serious worm infection in a human via a water-distribution system are extremely high. It is unlikely that the transmission of helminthic infections via water systems is significant. Free-living nematodes that ordinarily are found in soil are occasionally washed into river water by heavy rains, and can pass rapid sand-filtration barriers and survive chlorination. These nematodes belong to the genera Cheilobus, Diplogaster, Trilobus, Aphelenchus, Rhabditis, and others. They are not parasitic and pose no direct threat to man. However, they are reported to produce a gummy substance, small quantities of which confer an unpleasant taste to finished water. Besides having this characteristic, the nematodes may pose a problem in that they are resistant to chlorine; free residual chlorine at 2.5-3.0 ppm failed to immobilize them in 120 m (Chang et al., l961~. Because these nematodes ingest bacteria, it is conceivable that they could serve to protect pathogenic organisms from chlorine and therefore take them to the consumer. However, the chances of this appear small, because the nematodes pass the bacteria when they are active, and shed bacteria if they molt to a sheathed (resistant) stage. As examples of actively motile nematode larvae, such as infective hookworm and Strong~loides larvae, these nematodes indicate the ease with which such organisms can penetrate sand filters. Statistically, however, the chances are small that water-distribution systems contribute significantly to the spread of parasitic nematode infections, or that viable
Microbiology of Drinking Water 117 free-living nematodes will carry any significant numbers of bacterial pathogens past the disinfection systems. Water-Treatment Practices and Parasite Removal Cysts of Entamoeba histolytica are resistant to high concentrations of chlorine (Brady et al., 1943; Stringer et al., 19751. Concentrations of chlorine required to kill cysts in raw water in 20-30 m can exceed 9 ppm, depending on pH and other conditions; certainly, residual-chlorine concentrations in water-distribution systems are not adequate. Giardia cysts are also not reportedly destroyed by chlorination at conventional doses and contact times. It has been stated that they will survive 0.5 ppm for 30 m. Ozone, applied at 0.5 ppm, can kill 97% to more than 99% of cysts of E. histolytica suspended in tapwater (Newton and Jones, 1949~. Bromine is a more effective cysticide than chlorine or iodine. At a pH of 4, a bromine residue at 1.5 ppm was needed to produce.99.9% cyst mortality in 10 m. The same degree of cyst destruction was achieved by iodine residue at 5 ppm and chlorine residue at 2 ppm, respectively. High pH's decrease the cysticide activity of all three halogens (Stringer et al., 19751; however, the cysts of both these protozoans are of such size that flocculation and filtration can be expected to remove them from finished water (Anderson et al., 19731. The cyst of E. histolytica measures 10-15 Am in diameter, and that of C. Iamblia measures 9-12 x 6-9 inn. When outbreaks of G. Iamblia have been traced to municipal water, the system has relied on disinfection, rather than including flocculation and filtration. This suggests that mechanical means of clarifying water are important as antiparasitic measures. Because the protozoan cysts are resistant to usual residual-chlorine concentrations maintained in the distribution system, the need for filtration is emphasized. It is also important to emphasize the danger of faulty distribution systems, even when relatively high residual-chlorine concentrations are maintained; cross-connection control is essential. Conclusions The principal pathogenic parasites that may escape our sanitary barriers in public water supplies are the protozoa Entamoeba histolytica and Giardia lamblia. The cysts of these organisms are not completely destroyed by the usual chlorination. Most cysts, however, will be removed by sedimentation and filtration through sand. Filters should be of adequate depth, and the rates of filtration and back-flushing should be adjusted, to ensure entrapment of cysts and to prevent turnover of the
118 DRINKING WATER AND HEALTH filter or channeling. Proper maintenance is essential. The greatest risks will occur in water systems that use only disinfection, and in contamina- tion of water in the distribution system by sewage from broken lines or from cross-connections. Normal residual chlorine in such circumstances cannot kill cysts. More study is needed to define the conditions required for destruction of Giardia lamblia cysts; little is known of their survival. Occasionally, facultatively parasitic amebae can pass through the treatment processes and appear in finished water. Although these amebae can produce fatal encephalitides in man, the usual route of entry is the nasal passages, while a person swims in a pool or a pond of untreated or inadequately treated fresh water. The numbers of free-living ameba cysts found in finished water are small. Entrapment of pathogenic bacteria by parasites and other metazoa, such as free-living nematodes, may lead to protection of the bacteria from disinfection. However, it appears unlikely on the basis of the statistics of the occurrence of such organisms in finished water, the condition of the organisms seen, and the fact that viable nematodes are continually passing bacteria through their gut that a significant number of pathogenic organisms would survive in finished water with an adequate residual disinfectant. SUMMARY MICROBIOLOGY OF DRINKING WATER The incidence of enteric disease in the United States has been reduced by elective water treatment systems, but in 1974 the Center for Disease Control reported 28 waterborne disease outbreaks and 8,413 cases. In 1975 there were 24 outbreaks and 10,879 cases. Very few of these outbreaks were caused by chemical poisoning. Most cases (6,832 or 81~o) in 1974 were caused by microorganisms. The category designated "acute gastrointestinal illness" accounted for the largest number of cases (9,760) in 1975. These cases were characterized by symtoms for which no etiologic agent was identified. They account for approximately 90% of the total cases reported by the Center for Disease Control in 1975. Microorganisms account for most of the remaining cases. Improved detection and reporting systems are needed to determine more accurately the incidence of waterborne diseases nationwide. Outbreaks occur that are not reported. The etiologic agent was not identified in 15% of the cases reported in 1974, and 90% of those reported in 1975. Improved alarm systems are needed. In 1971-1974, deficiencies in treatment, such as inadequate or
Microbiology of Drinking Water 119 interrupted chlorination and contamination of groundwater, were responsible for a majority (kilo) of the waterborne-disease outbreaks. In 1975, treatment deficiencies were responsible for most outbreaks; however, deficiencies in the distribution systems were responsible for the highest number of cases. Bacteria The control of waterborne epidemics still depends largely upon the control of infectious enteric diseases. Much of the success in this regard can be attributed directly to the use of chlorine as a disinfectant. Although the excessive use of chlorine in water treatment may result in the formation of several compounds that are known carcinogens for animals, and suspected carcinogens for humans, the advantages far outweigh this limitation. Several substitutes for chlorine (e.g. ozone, chlorine dioxide, bromine, and iodine) that are also powerful oxidants and disinfectants have been suggested, but much more research is required before any of them can be recommended as a sole substitute for chlorine in water treatment. Questions concerning disinfection effectiveness, toxicity of by-products, and residual in the distribution system must be answered for proposed substitutes as well as for chlorine. It may be possible to reduce the concentrations of undesirable organic by-products of chlorination without compromising disinfection by changing the sequence or rate of chlorine addition in relation to other steps in water treatment. Partial use of other oxidizing agents, before chlorination, may also help to modify organic matter before significant amounts of chlorinated derivatives can be formed. Good engineering and public health practices emphasize the need for using raw water of the highest possible quality. Bacteriological testing, or the imposition of bacteriological standards, are adjuncts to, not substi- tutes for, good~uality raw water, proper water treatment, and integrity of the distribution system. Application of the present coliform standards appears adequate to protect public health when raw water is obtained from a protected source, is appropriately treated, and is distributed in a contamination-free system. Current coliform standards are not satisfactory for water reclaimed directly from wastewater. The meeting of current coliform standards is insufficient to protect public health for water reclaimed directly from waste water, or for water containing several percent of fresh sewage effluent. For such raw-water supplies, standards for viable virus content,
120 DRINKING WATER AND HEALTH Clostridium content, or total bacterial count should be developed and applied as a supplement to coliform standards. The standard plate count is not a substitute for total coliform measurements of the sanitary quality of potable water; it is, however, a valuable procedure for assessing the bacterial quality of drinking water. Ideally, standard plate counts (SPC) should be performed on samples taken throughout the systems. The SPC has major health significance for surface-water systems that do not use flocculation, sedimentation, filtration, and chlorination, and for those groundwater systems that do not include chlorination. Viruses The bacteriological monitoring methods currently prescribed (coliform count, standard plate count) are the best indicators available today for routine use in determining the probable levels of viruses in drinking water. The best available water treatment technology, if diligently applied, should provide a high degree of assurance that viruses injurious to human health are absent from finished drinking water. However, because knowledge of the scale of potential viral contamination is scanty, and because there is no rigorous basis for establishing a harmless level of viral concentration in water, research on the problems of viral contami- nation should be strongly supported. In particular, the following subjects deserve special attention: 1. Methods for testing of potable water for viral contamination. 2. Methods for recovery, isolation, and evaluation of viruses (especial- ly hepatitis A). 3. Specific etiology of viral gastroenteritis. 4. Methods for evaluating and improving electiveness of present water treatments to remove or inactivate viruses. 5. The amount of virus that must be ingested to produce infections and disease should be determined for several different enteric viruses. Parasites The most important waterborne parasitic diseases in the United States are amoebiasis and giardiasis. Outbreaks of amoebiasis appear to have resulted from sewage contamination in the distribution systems. Giardia- sis, which in recent years has become a major problem in some areas, appears to be associated with inadequately treated surface waters. The cysts of both of these parasites are more resistant to chlorine than
Microbiology of Drinking Water 121 are bacteria, but flocculation and filtration can remove them. Neverthe- less, knowledge of the vulnerability of all of these organisms to disinfection is incomplete, and, in particular, the conditions necessary for destruction of Giardia cysts require further study. The same considera- tions apply to a few other parasitic protozoa that, although rare, have been identified in public water systems. Metazoan parasites ~elm~nths, nematodes) that can be present in raw water will be controlled in public water supplies by well-regulated flocculation, filtration, and disinfection. Testing One of the greatest deficiencies of customary methods for evaluating the bacteriological quality of water is that results from tests are unknown until after the sampled water has already entered the distribution system and been used. Successful regulation of the microbiological quality of drinking water therefore depends on the use of raw water supplies of relatively invariant high quality. Sudden invasions of contamination are unlikely to be detected promptly enough to prevent exposure and, in addition, may overwhelm the corrective treatments. Nevertheless, it is essential that present methods of microbiological testing be continued, In order to validate the electiveness of disinfection, and for detecting defects within the system. REFERENCES Ahmed, Z., I.A. Poshini, and M.A. Siddiqui. 1967. Bacteriological examination of drinking water of Karachi and isolation of enteric pathogens. Pakistan J. Sci. Ind. Res. 7:103-110. Allen, M.J., and E.E. Geldreich. 1975. Bacteriological criteria for groundwater quality. Ground Water 13:45-52. Anderson, K., A. Jamieson, J.B. Jadin, and E. Willaert. 1973. Primary amoebic meningecen- cephalitis. Lancet 1:672. Anonymous. 1957. Infectious hepatitis in Delhi (1955-56): A critical study. Ind. J. Med. Res. 45(Suppl.~. 155 pp. Baylis, J.R., O. Gullans, and B.K. Spector. 1936. The efficiency of rapid sand filters in removing cysts of amoebic dysentery organisms from water. Public Health Rep. 51:1567- 1575. Bell, J.B., and J.M. Vanderpost. 1973. Comparion of membrane filtration methods in the isolation of "coliforms." Proc. 16th Conf. Great Lakes Res., pp. 15-20. International Association for Great Lakes Research. Braun-Bromf~eld, Ann Arbor. Bellelli, E., and G. Leogrande. 1967. Ricerche batteriologiche e virologiche sui mitili. Ann. Sclavo 9: 82~828. Bendinelli, M., and A. Ruschi. 1969. Isolation of human enterovirus from mussels. Appl. Microbial. 18:531-532.
122 DRINKING WATER AND HEALTH Berg, G. 1973. Reassessment of the virus problem in sewage and in surface and renovated waters. In S.H. Jenkins, ed. Progress in Water Technology, Vol. 3, pp 87-94. Pergamon, Oxford. Berg, G., H.L. Bodily, E.H. Lennette, J.L. Melnick, and T.G. Metcalf. 1976. Viruses in Water. Proceedings of the International Conference on Viruses in Water, Mexico City, 1974. American Public Health Association, Washington, D.C. Berg, G., S.L. Chang, and E.K. Harris. 1964. Devitalization of microorganisms by iodine. 1. Dynamics of the Revitalization of enteroviruses by elemental iodine. Virology 22:469- 481. Blacklow, N.R., R. Dolin, D.S. Fedson, H. Dupont, R.S. Northrup, R.B. Hornick, and R.M. Chanock. 1972. Acute infectious nonbacterial gastroenteritis: etiology and pathogenesis. Ann. Intern. Med. 76:993-1008. Bodian, D. 1957. Mechanisms of infection with polioviruses. In Cellular biology, nucleic acids and viruses, Vol. 5, pp. 57-72. Special Publications of the New York Academy of Sciences. Boring, J.R., III, W.T. Martin, and L.M. Elliott. 1971. Isolation of Salmonella typhimurium from municipal water, Riverside, Calif., 1965. Am. J. Epidemiol. 93:49-54. Brady, F.J., M.F. Jones, and W.L. Newton. 1943. The effect of chlorination of water on viability of cysts of Entamoeba histolytica. War Med. 3 :409-419. Brandborg, L.L., C. B. Tankersley, S. Gottlieb, M . Barancik, and V. E. Sartor. 1967. Histological demonstration of mucosal invasion by Giardia lamblia in man. Gastroenter- ology 52:143-150. Brown, T.S., J.F. Malina, Jr., B.D. Moore, and B.P. Sagik. 1974. Virus removal by diatomaceous earth filtration. In J.F. Malina, Jr., and B.P. Sagik, eds. Virus Survival in Water and Wastewater Systems, pp. 129-144. Center for Research in Water Resources, University of Texas at Austin. Buchanan, R.E., and N.E. Gibbons, eds. 1974. Bergey's Manual of Determinative Bacteriology, 8th ed. Williams & Wilkins, Inc., Baltimore. Bundesen, H.N., J.I. Connolly, I.D. Rawlings, A.E. Gorman, G.W. McCoy, and A.V. Hardy. 1936. Epidemic amebic dysentery. Nat. Inst. Health Bull. no. 166.187 pp. Butterfield, C.T., E. Wattie, S. Megregian, and C.W. Chambers. 1943. Influence of pH and temyrature on the survival of coliforms and enteric pathogens when exposed to free chlorine. Public Health Rep. 58:1837-186~. Butterfield, C.T., and E. Wattie. 1946. Influence of pH and temyrature on the survival of coliforms and enteric pathogens when exposed to chloramine. Public Health Rep. 61: 157-192. Cady, P., and S.W. Dufour. 1974. Automated defection of microorganism metabolism end growth by impedance measurements. Abstracts, 74th Meeting of the American Society for Microbiology, Chicago, Ill. No. E. 43. Carlson, G.F., Jr., F.E. Woodard, D.F. Wentworth, and O.J. Sproul. 1968. Virus inactivation on clay particles in natural waters. J. Water Pollut. Control Fed. 40:R89- 106. Cash, R.A., S.I. Music, J.P. Libonati, M.J. Snyder, R.P. Wenzel, and R.B. Hornick. 1974. Response of man to infection with Vibrio cholerae. I. Clinical, serolog~c and bacterio- log~c responses to a known inoculum. J. Infect. Dis. 129:45-52. Caverly, C.S. 1896. Notes of an epidemic of acute anterior poliomyelitis. J. Am. Med. Assoc. 26: 1-5. Center for Disease Control. 1973. Morbidity and Mortality Weekly Report 22:217. U.S. Department of Health, Education and Welfare, Atlanta, Ga.
Microbiology of Drinking Water 123 Center for Disease Control. 1974. Morbidity, and Mortality in the United States. Vol. 23 No. 53. Year ending Dec. 28. U.S. Department of Health, Education and Welfare, Atlanta, Ga. Center for Disease Control. 1975. Giardiasis-In residents of Rome, New York and in U.S. travelers to the Soviet Union. Morbidity and Mortality Weekly Report 24(43):366, 371. U.S. Department of Health, Education and Welfare, Atlanta, Ga. Center for Disease Control. 1976a. Foodborne and Waterborne Disease Outbreaks, Annual Summary 1974. Publication no. (CDCj 76-8185. U.S. Department of Health, Education and Welfare. Atlanta, Ga. Center for Disease Control. 1976b. Foodborne and Waterborne Disease Outbreaks, Annual Summary 1975. Publication no. (CDC) 76-8185. U.S. Department of Health, Education and Welfare. Atlanta, Ga. Chang, P.W., O.C. Liu, L.T. Miller, and S.M. Li. 1971. Multiplication of human enteroviruses in northern quahogs. Proc. Soc. Exp. Biol. Med. 136:138~1384. Chang, S.L. 1961. Viruses, amebas and nematodes and public water supplies. J. Am. Water Works Assoc. 53:288-296. Chang, S.L. 1968. Water borne viral infections and their prevention. Bull. WHO 38:401-414. Chang, S.L., G. Berg, N.A. Clarke, and P.W. Kabler. 1960. Survival and protection against chlorination of human enteric pathogens in free-living nematodes isolated from water supplies. Am. J. Trop. Med. Hyg. 9:136-142. Chang, S.L., R.L. Woodward, and P.W. Kabler. 1960.Surveyoffree-living nematodes and amebas in municipal supplies. J. Am. Water Works Assn. 52:613-168. Clarke, N.A. 1976. Personal Communication with T.G. Metcalf. Clarke, N.A. and P.W. Kabler. 1954. Inactivation of purified Coxsackie virus in water by chlorine. Am. J. Hyg. 59: 119-127. Clarke, N.A., R.E. Stevenson, and P.W. Kabler. 1956. The inactivation of purified type 3 adenov~rus in water by chlorine. Am. J. Hyg. 64:314-319. Clarke, N.A., E.W. Akin, O.C. Liu, J.C. Hoff, W.F. Hill, Jr., D.A. Brashear, and W. Jakubowski. 1975. Virus study for d~nking-water supplies. J. Am. Water Works Assoc. 67: 192-197. Clarke, N.A., G. Berg, P.K. Kabler, and S.L. Chang. 1964. Human enteric viruses in water: Source, survival and removability. Proceedings of the International Conference, London, 1962. Adv. Water Pollut. Res. 2:523-536. Clarke, N.A., W.F. Hill, Jr., and W. Jakubowski. 1974. Detection of viruses in water: Tentative standard method. Proceedings, Second Annual Water Technology Conference, American Water Works Association, December 14, 1974, Dallas, Texas. Cliver, D.O. 1967. Enterovirus detection by membrane chromatography. In G. Berg, ed. Transmission of Viruses by the Water Route, pp. 139-141. Interscience, New York. Cliver, D.O. 1971. Viruses in water and wastewater: Effects ofsome treatment methods.In V. Snoeyink and V. Griffin, eds. Virus and Water Quality: Occurrence and Control, Proceedings of the Thirteenth Water Quality Conference, vol. 69, no. 1, pp. 149-157. University of Illinois Bulletin, Urbana. Colwell, R. R., and F. M. Hetrick. 1976. Annual Report, Office of Naval Research. In A.L.H. Gameson, ed. Discharge of Sewage from Sea Outfalls. Proc. Symp. August 27 - Sept. 2,1974. Pergamon Press. 455 pp. Committee on the ECHO Viruses. 1955. Enteric cytopathogenic human orphan (ECHO) viruses. Science 122:1187-1188. Committee on the Enteroviruses. 1957. The enteroviruses. Am. J. Public Health 47:1556- 1566.
124 DRINKING WATER AND HEALTH Committee on the Enteroviruses. 1962. Classification of human enteroviruses. Virology 16:501-504. Cowan, S.T. 1974. Enterobacteriaceae. In R.E. Buchanan and N.E. Gibbons, eds. Bergey's Manual of Determinative Bacteriology, 8th ea., pp. 290 293. Williams & Willcins, Baltimore. Craig, C.F. 1934. Amebiasis and Amebic Dysentery. Charles C Thomas, Springfield, Ill. 315 PP Craun, G.F. 1975. Microbiology-Waterborne outbreaks. J. Water Poll. Control Fed. 47:1566-1581. Craun, G.F., and L.J. McCabe. 1973. Review of the causes of waterborne-disease outbreaks. J. Am. Water Works Assoc. 65:7484. Craun, G.F., L.J. McCabe and J.M. Hughes. 1976. Waterborne Disease Outbreaks in the U.S. 1971-1974. J. Am. Water Works Assoc. 68:420-424. Culp, G.L., R.L. Culp, and C.L. Hamann. 1973. Water resource preservation by planned recycling of treated wastewater. J. Am. Water Works Assoc. 65:641-647. Dahling, D.R., G. Berg, and D. Berman. 1974. BGM, a continuous cell line more sensitive than primary rhesus and African green kidney cells for the recovery of viruses from water. Health Lab. Sci. 11:275-282. Dalldorf, G., and G.M. Sickles. 1948. Unidentified, filterable agent isolated from the feces of children with paralysis. Science 108:61-62. Danielsson, D. 1965. A membrane filter method for the demonstration of bacteria by fluorescent antibody technique. 1. A methodological study. Acta Pathol. Microbiol. Scand. 63:597-603. Davis, B.D., R. Dulbecco, H.N. Eisen, H.S. Ginsberg, and W.B. Wood. 1967. Microbiology. Harper and Row, New York. 1464 pp. Denis, F. 1974. Les virus pathogenes pour l'homme dans les eaux de mer et dans les mollusques: Survie-Recherche-Bilan. Med. Mall Infect. 4-6 bis:325-334. DeBlanc, H.J., Jr., F. DeLand, and H.N. Wagner, Jr. 1971. Automated radiometric detection of bacteria in 2967 blood cultures. Appl. Microbiol. 22:846-849. Di Girolamo, R., J. Liston, and J.R. Matches. 1970. Survival of virus in chilled, frozen and processed oysters. Appl. Microbiol. 20:58-63. Dismukes, W.E., A.L. Bisno, S. Katz, and R.F. Johnson. 1969. An outbreak of gastroenteri- tis and infectious hepatitis attributed to raw clams. Am. J. Epidemiol. 89:555-561. Dolivo-Dobrovolskiy, L.B., and V.S. Rossovskaya. 1956. The problem of the survival of dysentery bacteria in reservoir water. Gig. Sanit. 1956(6): 52-55. Cited in Biol. Abstr. 34: 13898 (1958). Dougherty, W.J., and R. Altman. 1962. Viral hepatitis in New Jersey 1961) 1961. Am. J. Med. 32:704716. ~ Duff, M.F. 1967. The uptake of enteroviruses by the New Zealand marine blue mussel Mytilus edulis aoteanus. Am. J. Epidemiol. 85:486-493. Dulbecco, R. 1952. Production of plaques in monolayer tissue cultures by single particles of an animal virus. Proc. Nat. Acad. Sci. USA 38:747-752. Dulbeeco, R., and M. Vogt. 1954. Plaque formation, and isolation of pure lines with poliomyelitis viruses. J. Exp. Med. 99: 167-182. Dunlop, S.G., R.M. Twedt, and W.L.L. Wang. 1952. Quantitative estimation of Salmonella in irrigation water. Sewage Ind. Wastes 24:1015-1020. Dutka, B.J. 1973. Coliforms are an inadequate index of water quality. J. Environ. Health 36:39-46.
Microbiology of Drinking Water 125 DuPont, H.L., S.B. Formal, R.B. Hornick, M.J. Snyder., J.P. Libonati, D.G. Sheahad, E.H. LaBrec, and J.P. Kalas. 1971. Pathogenesis of Escherichia cold diarrhea. New Engl. J. Med. 285: 1-9. DuPont, H.L., R.B. Hornick, A.T. Dawkins, M.J. Snyder, and S.B. Fornlal. 1969. The response of man to virulent Shigella flexneri 2a. J. Infect. Dis. 119:296-299. DuPont, H.L., R.B. Hornick, M.J. Snyder, J.P. Libonati, S.B. Formal, and E.J. Gangarosa. 1972a. Immunity to shigellosis. I. Response of man to attenuated strains of Shigella. J. Infect. Dis. 125:5-11. DuPont, H.L., R.B. Hornick, M.J. Snyder, J.P. Libonati, S.B. Formal, and E.J. Gangarosa. 1972b. Immunity in shigellosis. II. Protection induced by oral live vaccine or primary infection. J. Infect. Dis. 125: 12-16. Edwards, P.R., and W.H. Ewing. 1972. Identification of Enterobacteriaceae, 3rd ed. Burgess Publishing Co., Minneapolis. Enders, J.F., T.H. Weller, and F.C. Robbins. 1949. Cultivation of Lansing strain of poliomyelitis virus in cultures of various human embryonic tissue. Science 109:85-87. England, B., R.E. Leach, B. Adame, and R. Shiosaki. 1967. Virologic assessment of sewage treatment at Santee, California. In G. Berg, ed. Transmission of Viruses by the Water Route, pp. 401-417. Interscience, New York. Ewing, W.H., and W.J. Martin. 1974. Enterobacteriaceae. In E.H. Lennette, E.H. Spaulding, and J.P. Truant, eds. Manual of clinical microbiology, pp. 189-221. American Society for Microbiology, Washington, D.C. Ewing, W.H., R. Hugh, and J.G. Johnson. 1961. Studies on the Aeromonas group. U.S Department of Health, Education, and Welfare Public Health Service. Communicable Disease Center, Atlanta, Ga. Fattal, B., E. Katzenelson, M. Nevo, and H.I. Shuval. 1973. Evaluation of difl~erent methods for the detection and concentration of small quantities of viruses in water. Proc. Fourth Scientific Conference, Israel Ecological Soc., Tel-Aviv, April 8-9, pp. A49-A69. Favero, M.S. and C.H. Drake. 1964. Comparative study of microbial flora of iodinated and chlorinated pools. Public Health Rep. 79:251-257. Fenner, F., B.R. McAuslan, C.A. Mims, J. Sambrook, and D.O. White. 1974. The Biology of Animal Viruses, 2nd ed. Academic Press, New York. Fields, H.A., and T.G. Metcalf. 1975. Concentration of adenovirus from seawater. Water Res. 9:357-364. Foliguet, J.M., and F. Doncoeur. 1975. Elimination des enterovirus au cours du traitement des eaux d'alimentation par coagulation-flocculation-filtration. Water Res. 9:953-961. Foliguet, J.M., J. Lavillaureix, and L. Schwartzbrod. 1973. Viruses and water. II. General review of the methods available to detect viruses in water. Rev. Epidemiol. Med. Soc. Sante Publ. 21: 185-259. Foliguet, J.M., L. Schwartzbrod, and O.G. Gaudin. 1966a. La recherche des virus dans les eaux d'egout, de surface et d'alimentation en Meurthe-et-Moselle. II. Resultats quantitatifs et qualificatifs definitifs. Rev. Hyg. Med. Soc. 14:411-432. Foliguet, J.M., L. Schwartzbrod, and O.G. Gaudin. 1966b. La pollution virale des eaux usees, de surface et d'alimentation; etude effectuee dans le departement francais de Meurthe-et-Moselle. Bull. WHO 35:737-749. Gallagher, T.P., and D.F. Spino. 1968. The significance of numbers ofcoliform bacteria es an indicator of enteric pathogens. Water Res. 2: 169-175. Gard, S. 1957. General discussion. In F.W. Hartman, G.A. LoGrippo, J.G. Mateer, and J. Barron, eds. Hepatitis Frontiers, pp. 241-243. Little, Brown & Co. Boston.
126 DRINKING WATER AND HEALTH Gartner, H. 1967. Retention and recovery of polioviruses on a soluble ultrafilter. In G. Berg, ed. Transmission of Viruses by the Water Route, pp. 121-127. Interscience Publishers, New York. Geldreich, E.E. 1968. Literature review. Microbiology. J. Water Pollut. Control Fed. 40: 1052-1068. Geldreich, E.E. 1969. Literature review. Bacterial pathogen detection in water. J. Water Pollut. Control Fed. 41: 1054. Geldreich, E.E. 1970a. Applying bacteriological parameters to recreational water quality. J. Am. Water Works Assoc. 62: 113-120. Geldreich, E.E. 1970b. Literature review. Bacterial pathogen detection in water. J. Water Pollut. Control Fed. 42: 1060-1062. Geldreich, E.E. 1973. Is the total count necessary? Proc. Am. Water Works Assoc. Water Quality Technology Conference, pp. VII-l-VII-l l, Cincinnati (Dec.34, 1973). Geldreich, E.E. 1975. Handbook for evaluating water bacteriology laboratories. Water Supply Research Laboratory, Cincinnati, Ohio. U.S. Environmental Protection Agency EPA-670/9-75-006: 135-158. Geldreich, E.E., H.D. Nash, D.J. Reasoner, and R.H. Taylor. 1972. The necessity of controlling bacterial populations in potable waters: Communitv water sunolv. J. Am. Water Works Assoc. 64:596-602. ~ I, Gelfand, H.M., D.R. LeBlanc, A.H. Holguin, and J.P. Fox. 1960. Preliminary report on susceptibility of newborn infants to infection with poliovirus strains of attenuated virus vaccine. In Second International Conference on Live Poliovirus Vaccines, pp. 308-314. Scientific Publication no. 50. Pan American Health Organization, Washington, D.C. Gelfand, H.M., D.R. LeBlanc, J.P. Fox, and A. H. Holguin. 1961. The susceptibility of infants to infection with natural ("wild") and attenuated (vaccine) strains of polioviruses. In Poliomyelitis: Papers and Discussions Presented at the Fifth International Poliomyeli- tis Conference, Copenhagen, Denmark, July 26-28, 1960, pp. 285-289. Lippincott, Philadelphia. Ginsburg, W. 1973. Improved total count techniques. In Proc. Am. Water Works Assoc. Water Quality Technology Conference, pp. VIII-I-VIII-8, Cincinnati (Dec. 3-4, 1973). Ginsburg, W., B. Crawford, and J.J. Knipper. 1972. Filter-fluorescent antibody technique for rapid screening of indicator organisms. J. Am. Water Works Assoc. 64:499-505. Greenberg, A.E. and B. H. Dean. 1958. The beef tapeworm, measly beef and sewage-A review. Sewage Ind. Wastes 30(3):262-269. Grindrod, J. and D.O. Cliver. 1970. A polymer two-phase system adapted to virus detection. Archiv Ges. Virusforsch. 31 :365-372. Grinstein, S., J.L. Melnick, and C. Wallis. 1970. Virus isolations from sewage and from a stream receiving effluents of sewage treatment plants. Bull. WHO 42:291-296. Harmon. P.H. 1937. The use of chemicals as nasal soravs in the orophYlaxis of poliomyelitis ''~''''~ ~ -r~-J~ r--r--J-~ r in man. J. Am. Med. Assoc. 109: 1061. Hatch, M.H., and G.E. Marchetti. 1975. A comparative study of agar overlay and standard tissue culture methods for isolation of enteroviruses. Public Health Rep. 90:29-33. Hedstrom, C.E., and E. Lycke. 1964. An experimental study on oysters as ViIUS carriers. Am. J. Hyg. 79:134142. Hill, W.F., Jr., E.W. Akin, and W.H. Benton. 1971. Detection of Viruses in Water: A review of Methods and Application. In V. Snoeyink and V Griffln, ed. Virus and Water Quality: Occurrence and Control. Proceedings of the Thirteenth Water Quality Conference, vol. 69, no. 1, pp. 17-46. University of Illinois Bulletin, Urbana.
Microbiology of Drinking Water 127 Hill, W.F., Jr., E.W. Akin, W.H. Benton, and T. G. Metcalf. 1972. Virus in water. II. Evaluation of membrane cartridge filters for recovering low multiplicities of poliovirus from water. Appl. Microbiol. 23:880-888. Hill, W.F., Jr., W. Jakubowski, W.E. Akin, and N.A. Clarke. 1976. Detection of virus in water: Sensitivity of the tentative standard method for drinking water. Appl. Environ. Microbiol. 31 :254261. Hoff, J.C., and R.C. Becker. 1969. The accumulation and elimination of crude and clarified poliovirus suspensions by shellfish. Am. J. Epidemiol. 90:53-61. Holguin, A.H., J.S. Reeves, and H.M. Gelfand. 1962. Immunization of infants with the Sabin oral poliovirus vaccine. Am. J. Public Health 52:600 610. Hornick, R.B., S.E. Greisman, T.E. Woodward, H.L. DuPont, A.T. Dawkins, and M.J. Snyder. 1970. Typhoid fever: pathogenesis and immunologic control. New Engl. J. Med. 283:686-691. Horsfall, F.L., and I. Tamm. 1965. Viral and Rickettsial Infections of Man. Lippincott, Philadelphia. pp. 574. Horstmann, D.M. 1961. Factors affecting optimum dosage levels of live poliov~rus vaccines. In Poliomyelitis: Papers and Discussions Presented at the Fifth International Poliomyeli- tis Conference, Copenhagen, Denmark, July 26-28, 1960, pp. 304310. Lippincott, Philadelphia. Joklik, W.K., and J.E. Darnell, Jr. 1961. The adsorption and early fate of purified poliovirus in HeLa cells. Virology 13:439447. Jordan, H.E. 1937. Editorial statement-The coliform group of bacteria. J. Am. Water Works Assoc. 29: 1999-2000. Kapikian, A.Z., R.G. Wyatt, R. Dolin, T.S. Thornhill, A.R. Kalica, and R.M. Chanock. 1972. Visualization by immune electron microscopy of a 27-nm particle associated witb acute infectious nonbacterial gastroenteritis. J. Virol. 10: 1075-1081. Katz, M., and S.A. Plotkin. 1967. Minimal infective dose of attenuated poliovirus for man Am. J. Public Health 57: 1837-1840. Katzenelson, E. 1976. Virologic and engineering problems in monitoring viruses in water. I, Viruses in Water. Proceedings of the International conference on Viruses in Water, pp 152-164. Mexico City, 1974, American Public Health Association, Washington, D.C Katzenelson, E., B. Kletter, H. Schechter, and H.I. Shoval. 1974. Inactivation of viruses and bacteria by ozone. In A.J. Rubin, ed. Chemistry of Water Supply, Treatment and Distribution, pp. 409-421. Ann Arbor Science Publishers, Ann Arbor, Mich. Kehr, R.W., and C.T. Butterfield. 1943. Notes on the relation between coliforms and enteric pathogens. Public Health Rep. 58:589-607. Kelly, S., and W.W. Sanderson. 1958. The effect of chlorine in water on enteric viruses. Am J. Public Health 48: 1323-1334. Kelly, S.M., and W.W. Sanderson. 1960. The effect of chlorine in water on enteric viruses II. The effect of combined chlorine on poliomyelitis and Coxsackie viruses. Am. J. Publi~ Health 50: 14-20. Klein, D.A., and S. Wu. 1974. Stress: A factor to be considered in heterotrophi~ microorganism enumeration from aquatic environments. Appl. Microbiol. 27:429-431 Klein, M. 1966. Adenovirus. In J.E. Prier, ed. Basic Medical Virology, pp.385 402. William~ ~ WiLicins Co., Baltimore. Kling, C. 1929. Recherches sur l'epidemiologie de la poliomyelitis. Svenska Laksallsk Handl. 55:2348. Koff, R.S. and H.S. Sear. 1967. Internal temperature of steamed clams. New Engl. J. Med 276:737-739.
128 DRINKING WATER AND HEALTH Koff, R.S., G.F. Grady, T.C. Chalmers, J.W. Mosley, B.L. Swartz, and the Boston Inter- hospital Liver Group. 1967. Viral hepatitis in a group of Boston hospitals. III. Importance of exposure to shellfish in a non-epidemic period. New Engl. J. Med. 276:703-710. Koprowski, H. 1956. Imm~ni~tion against poliomyelitis with living attenuated virus. Am. J. Trop. Med. Hyg.5:440 452. Krugman, S., J. Warren, M.S. Eiger, P.H. Berman, R.M. Michaels, and A.B. Sabin. 1961. Immunization with live attenuated poliovirus vaccine. Am. J. Dis. Child. 101:23-29. Kruse, C.W., V.P. Olivieri, and K. Kawata. 1971.The enhancement ofviralinactivation of halogens. In V. Snoeyink and V. Griffin, eds. Virus and Water Quality: Occurrence and Control. Proceedings of the Thirteenth Water Quality Conference, vol. 69, no. 1, pp 197- 209. Univasity of Illinois Bulletin, Urbana. Kurstak, E., and R. Morisset. eds. 1974. Viral Immunodiagnosis. Academic Press, New York. Lal, S.M., and E. Lund. 1976. Recovery of virus by chemical precipitation followed by elusion. In Proceedings of the Seventh International Conference on Water Pollution Research, Paris, 1974, vol. 1, pp. 687-693. (Technical Papers, S.H. Jenkins, ea.) Pergamon, New York. Lee, L.H., C.A. Phillips, M.A. South, J.L. Melaick, and M.D. Yowl 1965. Enteric virus isolation in different cell cultures. Bull. WHO 32:657-663. Lefler, E., and Y. Kott. 1974. Virus retention and survival in sand. In J.F. Malina, Jr. and B.P. Sagik, eds. Virus Survival in Water and Wastewater Systems, pp. 8491. Center for Research in Water Resources, University of Texas at Austin. LeMaistre, C.A., R.W. Sappenfield, C. Culbertson, F.R.N. Carter, A.C. Offutt, H. Black, and M.M. Brooke. 1956. Studies of a waterborne outbreak of amebiasis, South Bend, Indiana. I. Epidemiological aspects. Am. J. Hyg. 64:30-45. Lepow, M.L., R.J. Warren, V.G. Ingram, S.C. Dougherty, and F.C. Robbins. 1962. Sabin type I oral poliomyelitis vaccine. Effect of dose upon response of newborn infants. Am. J. Dis. Child. 104:67-71. Levin, G.V., V.L. Stauss, end W.C. Hess. 1961. Rapid coliform organism determination with C'4. J. Water PoDut. Control Fed.33: 1021-1037. Levin, G.V., E. Usdin, and A.R. Slonim. 1968. Rapid detection of microorganisms in aerospace water systems. Aerospace Med.39: 14-16. Levine, Max. 1953. Should bacteriological standards for drinking water be reevaluated? J. Am. Water Works Assoc. 45 :927-932. Levine, M.M., H.L. DuPont, S.B. Formal, R.B. Hornick, A. Takeuchi, E.J. Gangarosa, M.J. Snyder, and J.P. Libonati. 1973. Pathogenesis of Shigella dysenteriae 1 (Shiga) dysentery. J. Infect. Dis. 127. 261-270. Liu, O.C., H.R. Seraichekas, and B.L. Murphy. 1966a. Fate of poliovirus in northern quahaugs. Proc. Soc. Exp. Biol. Med. 121:601 607. Liu, O.C., H.R. Seraichekas, and B.L. Murphy. 1966b. Viral pollution ofshellfish. 1. Some basic facts of uptake. Proc. Soc. Exp. Biol. Med. 123 :481487. Liu, O.C., H.R. Seraichekas, and B.L. Murphy. 1967. Viral depuration of the northern quahaug. Appl. Microbiol. 15:307-315. Liu, O.C., H.R. Seraichekas, E.W. Akin, D.A. Brashear, E.L. Katz, and W.J. Hill, Jr. 1971. Relative resistance of twenty human enteric viruses to free chlorine in Potomac water. In V. Snoeyink and V. Griflin, eds. Virus and Water Quality: Occurrence and Control, Proceedings of the Thirteenth Water Quality Conference, vol. 69, no.l, pp. 171-195. University of Illinois Bulletin, Urbana.
Microbiology of Drinking Water 129 Loria, R.M., S. Kibrick, and W.A. Broitman. 1974. Peroral infection with group B coxsackievirus in the newborn mouse: a model for human neonatal infection. J. Infect. Dis. 130:225-230. Lund, E., and C.E. Hedstrom. 1966. The use of an aqueous polymer phase system for enterovirus isolations from sewage. Am. J. Epidemiol. 84:287-291. Lurman, A. 1885. Eine Icterusepidemie. Berlin Klin. Wochschr. 22: 2~23. Mack, W.N., L. Yue-Shoung, and D.B. Cochon. 1972. Isolation of poliomyelitis virus from a contaminated well. Health Serv. Rep. 87:271-274. MacKenzie, C.R., and D.J. Livingstone. 1968. Salmonellae in fish and food. S. Afr. Med. J. 42:999-1003. Majumdar, S.B., W.H. Ceckler and O.J. Sprout. 1973. Inactivation of poliovirus in water by ozonation. J. Water Pollut. Control Fed. 45:2433-2443. Manwaring, J.F., M. Chaudhuri, and R.S. Engelbrecht. 1971. Removal of viruses by coagulation and flocculation. J. Am. Water Works Assoc. 63:298-300. Mason, J.O., and W.R. McLean. 1962. Infectious hepatitis traced to the consumption of raw oysters. An epidemiologic study. Am. J. Hyg. 75:9~111. McCabe, L.J., and G.T. Craun. 1975. Status of waterborne diseases in the U.S. and Canada: Committee report. J. Am. Water Works Assoc. 67:95-98. McCabe, L.J., J.M. Symons, R.D. Lee, and G.G. Robeck. 1970. Survey of community water supply systems. J. Am. Water Works Assoc. 62: 67~687. McCullough, N.B., and C.W. Eisele. 195 la. Experimental human salmonellosis. I. Pathogenicity of strains of Salmonella meleagridis and Salmonella anatum obtained from spray-dried whole egg. J. Infect. Dis.88:278-289. McCullough, N.B., and C.W. Eisele. 1951b. Experimental human salmonellosis. II. Immunity studies following experimental illness with Salmonella meleagridis and Salmonella anatum. J. Immunol. 66:595-608. McCullough, N.B., and C.W. Eisele. 1951c. Experimental human salmonellosis. III. Pathogenicity of strains of Salmonella newport, Salmonella derby and Salmonella bareilly obtained from spray-dried whole egg. J. Infect. Dis. 89:209-213. McCullough, N.B., and C.W. Eisele. l951d. Experimental human salmonellosis. IV. Pathogenicity of strains of Salmonella pullorum obtained from spray-dried whole egg. J. Infect. Dis. 89:259-265. McFeters, G.A., G.K. Bissonnette, J.J. Jezeski, C.A. Thomson, and D.G. Stuart. 1974. Comparative survival of indicator bacteria and enteric pathogens in well water. Appl. Microbiol. 27:823-829. McKee, J.E. and H.W. Wolf, eds. 1963. Water Quality Criteria, 2nd ed. The Resources Agency of California State Water Resources Control Board, Publication no. 3-A (Reprint December, 1971). Sacramento. Melaick, J.L. 1957. A waterborne urban epidemic of hepatitis. In F.W. Hartman, ed. Hepatitis Frontiers, pp.211-225. Little Brown & Co., Boston. Melaick, J.L., V. Rennick, B. Hampil, N. Schmidt, and H.H. Ho. 1973. Lyophilized combination pools of enterovirus equine antisera: Preparation and test procedures for the identification of field strains of 42 enteroviruses. Bull. WHO 48:263-268. Merson, M.H., W.H. Barker, Jr., G.F. Craun, and L.J. McCabe. 1974. Outbreaks of waterborne disease in the United States 1971-1972. J. Infect. Dis. 129:614615. Metcalf, T.G., and W.C. Stiles. 1965. The accumulation of enteric viruses by the oyster, Crassostrea virginica. J. Infect. Dis. 115 :68-76. Metcalf, T.G., and W.C. Stiles. 1968. Viral pollution of shellfish in estuary waters. Am. Soc. Civil Eng. Sanit. Eng. Div. J. SA4:595-609.
130 DRINKING WATER AND HEALTH Metcalf, T.G., C. Wallis, and J.L. Melnick. 1974a. Environmental factors influencing isolation of enteroviruses from polluted surface waters. Appl. Microbiol. 27:921-926. Metcalf, T.G., C. Wallis, and J.L. Melnick. 1974b. Virus enumeration and public health assessments in polluted surface water contributing to transmission of virus in nature. In J.F. Malina, Jr. and B.P. Sagik, eds. Virus Survival in Water and Wastewater Systems, pp. 57-70. Center for Research in Water Resources, University of Texas at Austin. Meynell, G.G. 1961. Phenotypic variation and bacterial infection. Symp. Soc. Gen. Microbiol. 1 1:174195. Meynell, G.G., and E. Meynell. 1965. Theory and Practice in Experimental Bacteriology, pp. 199-203. Cambridge University Press. Miller, R.S. 1975. Check sampling for monitoring bacteriological quality. J. Am. Water Works Assoc. 67:28-32. Mitchell, J.R., M.W. Presnell, E.W. Akin, J.M. Cummins, and O.C. Liu. 1966. Accumula- tion and elimination of poliovirus by the,eastern oyster. Am. J. Epidemiol. 84:40-50. Moore, B.E.D., L. Funderburg, and B. P. Sagik. 1974. Application of viral concentration techniques to field sampling. In J.F. Malina and B.P. Sagik, eds. Virus Survival in Water and Wastewater Systems, pp. 3-15. Center for Research in Water Resources, University of Texas at Austin. Moore, G.T., W.M. Cross, D. McGuire, C.S. Mollohan, N.N. Gleason, G.R. Healy, and L.H. Newton. 1969. Epidemic giardiasis at a ski resort. New Engl. J. Med. 281:402 407. Morse, R.B., and A. Wolman. 1918. The practicability ofadopting standards ofquality for water supplies. J. Am. Water Works Assoc. 5:198-228. Mosley, J.W. 1967. Transmission of viral diseases by drinking water. In G. Berg, ed. Transmission of Viruses by the Water Route, pp. 5-23. Interscience Publishers, New York. Neefe, J.R., and J.S. Stokes. 1945. An epidemic of infectious hepatitis apparently due to a waterborne agent. J. Am. Med. Assoc. 128:1063-1075. Neefe, J.R., J.B. Baty, J.B. Reinhold, and J. Stokes, Jr. 1947. Inactivation of virus of infectious hepatitis in drinking water. Am. J. Public Health 37:365-372. Newman, J.S., and R.T. O'Brien. 1975. Gas chromatographic presumptive test for coliform bacteria in water. Appl. Microbiol. 30:584588. Newton, W.L., and M.F. Jones. 1949. Effect of ozone in water on cysts of Endamoeba histolytica. Am. J. Trop. Med. 29:669-681. Newton, W.L., W.B. Figgat, and S.R. Weibel. 1948. The e~ects of sewage-treatment processes upon ova and miracidia of Schistosoma japonicum. Sewage Works J. 20:657- 664. Nupen, E.M. 1970. Virus studies on the Windhoek waste water reclamation plant (southwest Africa). Water Res. 4:661-672. Olivieri, V.P., C.W. Kruse, Y.C. Hsu, A.C. Griffiths, and K. Kawata. 1975. The comparative mode of action of chlorine, bromine, and iodine on f2 bacterial virus. In J.D. Johnson, ed. Disinfection-Water and Wastewater, pp. 145-162. Ann Arbor Science Publishers, Inc., Ann Arbor, Mich. Parr, L.W. 1939. Coliform bactria. Bacteriol. Rev. 3: 148. Pfeiffer, K.R. 1973. The Homestead typhoid outbreak. J. Am. Water Works Assoc. 65:803- 805. Plotkin, S.A., and M. Katz. 1967. Minimal infective doses of viruses for man by the oral route. In G. Berg, ed. Transmission of Viruses by the Water Route, pp. 151-166. Interscience, New York. Prescott, S.C., C.E.A. Winslow, and M.H. McCrady. 1946. Water Bacteriology with Special Reference to Sanitary Water Analysis. 6th ed. John Wiley ~ Sons, New York.
Microbiology of Drinking Water 131 Provost, P.J., B.S. Wolanski, W.J. Miller, O.L. Ittensohn, W.J. McAleer, and M.R. Hilleman. 1975. Biophysical and biochemical properties of CR326 human hepatitis A virus. Am. J. Med. Sci. 270:87-92. Ptak, D.J., W. Ginsburg, and B.F. Willey. 1973. Identification and incidence of Klebsiella in chlorinated water supplies. J. Am. Water Works Assoc. 65:604 608. Putnam, J.J., and E.W. Taylor. 1893. Is acute poliomyelitis unusually prevalent this season? Boston Med. Surg. J. 129:509-510. Rao, N.U., and N.A. Labzoffsky. 1969. A simple method for the detection of low concentrations of viruses in large volumes of water by the membrane filter technique. Can. J. Microbiol. 15:399403. Rao, V.C., R. Sullivan, R.B. Read, and N.A. Clarke. 1968. A simple method for concentrating and detecting viruses in water. J. Am. Water Works Assoc. 60:1288-1294. Reasoner, D.J. 1972. Literature review. Microbiological pathogens in animals and their detection. J. Water Pollut. Control Fed. 44:1182-1193. Reasoner, D.J. 1973. Literature review. Microbiology-Detection of bacterial pathogens and their occurrence. J. Water Pollut. Control Fed. 45:1278-1289. Reasoner, D.J. 1974. Microbiology-Detection of bacterial pathogens and their occurrence. J. Water Pollut. Control Fed. 46:1395-1408. Reasoner, D.J. 1975. Microbiology-Detection of bacterial pathogens and their occurrence. J. Water Pollut. Control Fed. 47:1581-1587. Reasoner, D.J., and E.E. Geldreich. 1974. Rapid bacteriological methods. Proc. AWWA Water Quality Technology Conference, Dallas, Texas, pp. IX-1-IX-15. Ritchie, L.S. and C. Davis. 1948. Parasitological f~ndings and epidemiological aspects of epidemic amebiasis occurring in occupants of the Mantetsu apartment building, Tokyo, Japan. Am. J. Trop. Med. 28:803-816. Robbins, F.C., J.F. Enders, T.H. Weller, and G.L. Florentino. 1951. Studies on the cultivation of poliomyelitis viruses in tissue culture. V. The direct isolation and serologic identification of virus strains in tissue culture from patients with nonparalytic and paralytic poliomyelitis. Am. J. Hyg. 54:286-293. Robert, V.B., and L.B. Rorke. 1973. Pr~mary amebic encephalitis, probably from Acantha- moeba. Ann. Int. Med. 79:174-179. Roos, B. 1956. Hepatitepidemi spridd genom ostron. Svenska Lakartidn. 53:989-1003. Rosen, L. 1965. Reovirus group. In F.L. Horsfall and I. Tamm, eds. Viral and Rickettsial Infections of Man, 4th ed. pp. 569-579. Lippincott, Philadelphia. Rowe, W.P., R.J. Huebner, L.K. Gilmore, R.H. Parrott, and T.G. Ward. 1953. Isolation of a cytopathogenic agent from human adenoids undergoing spontaneous degeneration in tissue culture. Proc. Soc. Exp. Biol. Med. 84:570-573. Ruddy, S.J., R.F. Johnson, J.W. Mosley, J.B. Atwater, M.A. Rossetti, and J.C. Hart. 1969. An epidemic of clam-associated hepatitis. J. Am. Med. Assoc. 208:649-655. Sabin, A.B. 1957. Properties of attenuated polioviruses and their behavior in human beings. In Cellular Biology Nucleic Acids and Viruses. vol. 5. on. 113-127. Soecial Publications of the New York Academy of Sciences. - ~ rr Sabin, A.B. 1959. Reoviruses, a new group of respiratory and enteric v~ruses formerly classified as ECHO type 10 is described. Science 130:1387-1389. Scarpino, P.V. 1971. Bacterial and viral analysis of water and waste water. In L.L. Ciacco, ed. Water and Water Pollution Handbook, vol. 2, pp. 639-761. Marcel Dekker, New York. Scarpino, P.V., M. Lucas, D.R. Dahling, G. Berg, and S.L. Chang. 1974. Effectiveness of hypochlorous acid and hypochlorite ion in destruction of v~ruses and bacteria. In A.J.
132 DRINKING WATER AND HEALTH Rubin, ed. Chemistry of Water Supply, Treatment and Distribution, pp. 359-368. Ann Arbor Science Publishers, Ann Arbor, Mich. Schafer, E. 1971. Experimentelle Untersuchungen zur quantitativen Erfassung van Polio- Impfvirus Type II in Oberflachenwasser. GWF-Wasser/Abwasser 112:109-113. Schaub, S.A., and B.P. Sagik. 1975. Association of enteroviruses with natural and artifically introduced colloidal solids in water and infectivity of solids-associated virions. Appl. Microbiol. 30:212-222. Schmidt, N.J., H.H. Ho, and E.H. Lennette. 1975. Propagation and isolation of group A coxsackieviruses in RD cells. J. Clin. Microbiol. 2: 183-185. Schulz, M.G. 1975. Giardiasis. J. Am. Med. Assoc. 233:1383-1384. Schwerdt, C.E. 1957. Physical and chemical characteristics of purified poliomyelitis virus. In Cellular Biology, Nucleic Acids and Viruses. vol. 5. Do. 157-166. Special Publications of the New York Academy of Sciences. - ~ or Scott, R.M., D. Seiz, and H.J. Shaughnessy. 1964. I. Rapid Cl4 test for coliform bacteria in water. II. Rapid Cl4 test for sewage bacteria. Am. J. Public Health 54:827-844. Seligmann, R. and R. Reitler. 1965. Enteropathogens in water with low Esch. cold titer. J. Am. Water Works Assoc. 57:1572-1574. Seraichekas, H.R., D.A. Brashear, J.A. Barnick, P.F. Carey, and O.C. Liu. 1968. Viral deputation by assaying individual shellfish. Appl. Microbiol. 16:1865-1871. Sharp, D.G., R. Floyd, and J.D. Johnson. 1975. Nature of the surviving plaque-forming units of reovirus in water containing bromine. Appl. Microbiol. 29:94101. Shaughnessy, H.J., R.C. Olsson, K. Bass, F. Friewer, and S.D. Levinson. 1946. Experi- mental human bacillary dysentery. J. Am. Med. Assoc. 132:362-368. Shuval, ELI., B. Fattal, S. Cymbalista, and N. Goldblum. 1969. The phase-separation method for the concentration and detection of viruses in water. Water Res. 3:225-240. Shuval, H.I., S. Cymbalista, B. Fattal, and N. Goldblum. 1967. Concentration of enteric viruses in water by hydro-extraction and two-phase separation. In G. Berg, ed. Transmission of Viruses by the Water Route, pp. 45-55. Interscience, New York. Smith, R.J., and R.M. Twedt. 1971. Natural relationships of indicator and pathogenic bacteria in stream waters. J. Water Pollut. Control Fed. 43:2200 2209. Sobsey, M.D. 1976. Methods for detecting enteric viruses in water and wastewater, In G. Berg, H.L. Bodily, E.H. Lennette, J.L. Melaick, and T.G. Metcalf, eds. Viruses in Water, Proceedings of the International Conference on Viruses in Water, Mexico City, 1974. American Public Health Association, Washington, D.C. Sobsey, M.D., C. Wallis, M. Henderson, and J.L. Melnick. 1973. Concentration of enteroviruses from large volumes of water. Appl. Microbiol. 26:529-534. Sorber, C.A., B.P. Sagik, and J.F. Malina, Jr. 1972. Monitoring of low-level virus in natural waters. Appl. Microbiol. 22:334338. Sprout, O.J. 1971. Recent research results on virus inactivation by water treatment processes. In V. Snoeyink and V. Griffin, eds. Virus and Water Quality: Occurrence and Control, vol. 69, no. 1, pp. 159-169. Proceedings of the Thirteenth Water Quality Conference, University of Illinois Bulletin, Urbana. Standard Methods for the Examination of Water and Wastewater. 1975. 14th Ed. American Public Health Association. Washington, D.C. Sternberg, G.M. 1892. Manual of Bacteriology. William Wood and Co., New York. Stevenson, A.H. 1953. Studies of bathing water quality and health. Am. J. Public Health 43:529-538. Stringer R.P., W.N. Cramer, and C.W. Kruse. 1975. Comparison of bromine, chlorine and iodine as disinfectants for ameobic cysts. In J.D. Johnson, ed. Disinfection-Water and
Microbiology of Drinking Water 133 Wastewater, Chapter 10, pp. 193-209. Ann Arbor Science Publishers, Inc., Ann Arbor, Mich. Sweet, B.H., R.D. Ellender, and J.K.L. Leong. 1974. Recovery and removal of viruses from water utilizing membrane techniques. In Developments in Industrial Microbiology, vol. 15: Symposium: Detection of Viruses in Waste and Other Waters, August 1973, pp. 143- 159. Proc. 30th General Meeting of the Society for Industrial Microbiology. Taylor, A., Jr., G.F. Craun, G.A. Faich, L.J. McCabe, and E.J. Gangarosa. 1972. Outbreaks of waterborne diseases in the United States, 1961-1970. J. Infect. Dis. 125:329-331. Taylor, D.G. and J.D. Johnson. 1974. Kinetics of viral inactivation by bromine. In A.J. Rubin, ed. Symposium on the Chemistry of Water Supply, Treatment and Distribution, Dallas, 1973, pp. 369408. Ann Arbor Science Publishers, Ann Arbor, Mich. Thomas, H.A., Jr., and R.L. Woodward. 1955. Estimation of coliform density by the membrane filter and the fermentation tube methods. Am. J. Public Health 45:1431-1437. Thorup, R.T., F.P. Nixon, D.F. Wentworth, and O.J. Sprout. 1970. Virus removal by coagulation with polyelectrolytes. J. Am. Water Works Assoc. 62:97-101. U.S. Environmental Protection Agency. 1975. NERC annual report. National Environmen- tal Research Center, United States Environmental Protection Agency, Cincinnati, Ohio. EPA 670/9-75-002. U.S. Treasury Department. 1914. Bacteriological standard for drinking water. Public Health Rep. 29:2959-2966. U.S. Public Health Service. 1962. Drinking water standards. Public Health Service Publ. no. 956. U.S. Department of Health, Education and Welfare, Washington, D.C. Van Donsel, D.J. 1971. Literature review. Microbiology-waterborne outbreaks. J. Water Pollut. Control Fed. 43:1221-1236. Veazie, L. 1969. Epidemic giardiasis. N. Engl. J. Med. 281:853. Wallis, C., and J.L. Melnick. 1967a. Concentration of viruses from sewage on Millipore membranes. Bull. WHO 36:219-225. Wallis, C., and J.L. Melnick. 1967b. Concentration of viruses on aluminum and calcium salts. Am. J. Epidemiol. 85:459-468. Wallis, C., A. Homma, and J.L. Meloick. 1972. Apparatus for concentrating viruses from large volumes. J. Am. Water Works Assoc. 64:189-196. Wallis, C., J.L. Melnick, and J.E. Fields. 1970. Detection of viruses in large volumes of natural waters by concentration on insoluble polyelectrolytes. Water Res. 4:787-796. Wang, W.L.L., S.G. Dunlop, and R.G. DeBoer. 1956.ThesurvivalofShigellain sewage. I. An effect of sewage and fecal suspensions on Shigella flexneri. Appl. Microbiol. 4:3438. Wanner, R.G., F.O. Atchley, and M.A. Wasley. 1963. Association of diarrhea with Giardia lamblia in families observed weekly for occurrence of enteric infections. Am. J. Trop. Med. Hyg. 12:851-853. Warren, R.J., M.L. Lepow, G.E. Bartsch, and F. C. Robbins. 1964. The relationship of maternal antibody, breast feeding and age to the susceptibility of newborn infants to infection with attenuated poliovirus. Pediatrics 34:4-13. Wattie, E., and C.W. Chambers. 1943. Relative resistance of coliform organisms and certain enteric pathogens to excess lime treatment. J. Am. Water Works Assoc. 35:709-720. Wellings, F.M., C.W. Mountain, A.L. Lewis, J.L. Nitzlcin, M.S. Saslaw, and R.A. Graves. 1976. Isolation of an enterovirus from chlorinated tap water. J. Infect. Dis. (In press.) White, G.C. 1975. Disinfection: The last line of defense for potable water. J. Am. Water Works Assoc. 67:410-413. Witt, M., R. Siegrist and W.C. Boyle. 1975. Rural household wastewater characterization. In Proceedings of the National Home Sewage Disposal Symposium, Dec. 9-10, 1974, pp. 79- 88. American Society of Agricultural Engineers, St. Joseph, Mich.
134 DRINKING WATER AND HEALTH Wolf, H.W. 1972. The coliform count as a measure of water quality In R. Mitchell, ed. Water Pollution Microbiology. Wiley Interscience, New York. Wolfe, M.S. 1975. Giardiasis. J. Am. Med. Assoc. 233:1362-1365. World Health Organization. 1970. European Standards for ~inking-Water, 2nd ed. Geneva. World Health Organization. 1971. International Standards for Drinking-Water, 3rd ed. Geneva. World Health Organization. 1964. Soil-transmitted helminths: Report of a WHO Expert Committee on helminthiasis. WHO Tech. Report Series no. 277. Geneva. World Health Organization. 1969. Amoebiasis: Report of a WHO Expert Committee. WHO Tech. Report Series no. 421. Geneva. Zamcheck, N., L.C. Hoskins, S.J. Winawer, S.A. Broitman, and L.S. Gottlieb. 1963. Histology and ultrastructure of parasite and intestinal mucosa in Human giardiasis: Effect of Atabrine therapy. Gastroenterology 44:860. Zdra~lek, J. 1974. Sensitivity of human diploid embryonic lung cells for isolation of Echoviruses from sewage. J. Hyg. Epidemiol. Microbiol. Immunol. 18:2-8.