Anstiss R, Ahmed M, Islam S, Khan AW, Arewgoda M. Int J Environ Health Res (2001) 11(3) p267-274

The town of Chapainawabganj in northwest Bangladesh was chosen as the location for a pilot arsenic mitigation project. Arsenic concentrations in and around the town have been found to be as high as 2.4 ppm with many samples above 0.05 ppm, the Bangladesh permissible limit. The results of the first 22 months of the pilot project are reported here.

The arsenic removal process was based on adsorption by ferric oxyhydroxide in a household water treatment system. Four families agreed to participate and use the systems for all drinking and cooking water. The untreated well water contained 0.932 ppm arsenic. During use the arsenic concentrations in the de-contaminated water increased as the ferric oxyhydroxide became more saturated until it was replaced in a 16-day cycle. Arsenic concentrations in treated water increased from 0.001 ppm on the first day of the cycle to 0.047 ppm on the 16th day.

There is anecdotal evidence of reduced gastrointestinal problems and skin effects in those consuming the de-contaminated water. The technical mechanism used here is flexible and can be scaled up or down for different water volumes, cycle lengths, and different tubewell arsenic concentrations and chemistries. An expansion of the project including a coordinated health monitoring program is planned.

Rosef O, Rettedal G, Lageide L. Intern J Environ Health Res (2001) 11(4) p321-327

Campylobacters are often found in water and water supplies and have been responsible for many outbreaks in countries including Norway. These organisms can remain dormant under unfavourable conditions and cannot be easily recovered on artificial media.

This study describes the occurrence of campylobacters in the Bø river in southeastern Norway which has a variety of environmental influences. Of the 60 samples taken, 32 were contaminated with campylobacters. 47 strains were isolated using an enrichment procedure and 28 were isolated by direct plating of a filter membrane. Of these 75 isolates, 33 belonged to Campylobacter coli, 26 to Campylobacter jejuni and 11 to Campylobacter lari; five strains were not culturable. In 19 of the positive samples more than one species was detected. All three species were detected in 3 samples. Isolation of Campylobacter did not correlate well with the occurrence of faecal coliforms. The high isolation rate of campylobacters from surface water and the fact that the virulence of strains isolated from the environment is not fully known indicates that precautions should be taken to avoid transmission of campylobacteriosis from untreated water sources.

Pizarro F, Olivares M, Araya M, Gidi V, Uauy R. Environ Health Perspect (2001) 109(9)n p949-52

Consuming drinking water or beverages with elevated copper concentrations can cause acute gastrointestinal symptoms such as epigastric pain, nausea, vomiting and diarrhoea. The aim of this study was to determine whether total copper or soluble copper concentration is associated with gastrointestinal signs and symptoms. Participants were women aged 18-55 years who worked at home and who were not pregnant or lactating.

Forty-five healthy women participated who lived in Santiago, Chile. All houses were similar and had copper piping systems. The study was conducted for 9 weeks and volunteers were randomised into three groups based on the sequence of soluble/insoluble copper they received. Subjects were blinded to copper concentrations they were receiving. The proportions of copper sulfate (soluble) to copper oxide (insoluble) subjects were given was 0:5, 1:4, 2:3, 3:2 and 5:0 (mg:mg/L). Subjects received flasks filled with a solution to add to their home drinking water each day, and recorded the amount of water ingested, and any symptoms experienced during the day. Once a week the copper concentration and pH of water prepared by the subjects in their homes was analysed for actual copper concentration. Blood samples were obtained 1 week before the beginning of the study and at the end of the protocol.

The copper content of tap water was <0.1 mg/L and therefore not considered a significant source of copper. Mean consumption of water was similar among the groups. Serum analysis showed copper levels, ceruloplasmin and activities of liver enzymes were within normal limits. There were 20 subjects who had gastrointestinal disturbances at least once during the study, 9 had diarrhoea and 11 reported abdominal pain, nausea or vomiting. There were no significant differences in the incidence of diarrhoea or other symptoms regardless of the ratio of copper sulfate to copper oxide.

Carmichael WW, Azevedo SMFO, An JS, Molica RJR, Jochimsen EM, Lau S, Rinehart KL, Shaw GR, Eaglesham GK. Environ Health Perspect (2001) 109(7): p663-8

This study reports further information on the first documented cases of human fatalities due to cyanobacterial hepatotoxins occurring in a dialysis clinic in Caruaru, Brazil in 1996 (see Health Stream Issue 10). Up to 76 deaths were attributed to severe neurological symptoms or liver failure, after patients were exposed to contaminated water.

Two groups of cyanobacterial toxins were identified, the hepatotoxic cyclic peptide microcystins and the hepatotoxic alkaloid cylindrospermopsin. It was concluded that intravenous exposure to microcystins, in particular microcystin-YR, -LR and -AR was the major cause of the fatalities. It was estimated that 19.5 micrograms/L microcystin was in the water used in dialysis, this is 19.5 times the guideline level of 1 microgram/L proposed by the World Health Organisation for safe drinking water. The tragedy was attributed to inadequate maintenance of filtration equipment at the clinic, coupled with use of heavily toxin-contaminated water. The clinic used trucked water from the cityās water treatment plant that had received only alum flocculation rather than full conventional treatment, and thus probably contained more algal cells and toxins than the tap water supply.

Falconer IR. Phycologia (2001) 40(3) p228-233

This paper reviews the major species of toxic cyanobacteria found in Australian waters and discusses the potential health impacts of exposure to these organisms and their toxins. The author also describes the basis of the World Health Organisation Guideline Values for safe cyanobacterial exposure levels for drinking water and recreational water.

Human deaths from oral consumption of cyanobacteria-contaminated water have not been unequivocally documented, however it is suspected that an unexplained outbreak of severe gastroenteritis in Brazil including a number of deaths may have been caused by cyanobacterial poisoning. Cyanobacterial toxins have been implicated in carcinogenesis and in tumour promotion although further evidence is needed to clarify these associations. In Australia there have been two reported instances of human injury resulting from toxic cyanobacterial contamination of drinking water reservoirs after treatment with copper sulphate which lyses the bloom but also liberates the toxins into water. Recreational exposure to cyanobacteria toxins is quite common with skin rashes, eye irritations and increases in gastrointestinal effects being reported.

Lahti K, Rapala J, Kivimaki AL, Kukkonen J, Niemela M, Sivonen K. Wat Sci Technol (2001) 43(12) p225-228

The aim of this study was to determine the occurrence of microcystins in raw water sources and in treated drinking water of a number of Finnish surface water and bank filtration plants. Water samples were taken for microcystin and phytoplankton analysis from four bank filtration plants and from nine surface waterworks using different processes for water treatment. Microcystins were analysed by immunoassay and phytoplankton was identified and quantified.

The concentrations of microcystins in all raw water samples from bank filtration plants was reasonably low (maximum value 1.9 microgram/L as MC-LR equivalents). Microcystins were found in treated water less frequently and in low concentrations (the highest concentration was 0.1 microgram/L). Cyanobacteria were absent from bank filtered drinking water in most cases.

Microcystins were present in most of the surface water sources that were monitored in the summer of 1999. In treated drinking water, microcystins were only detected in 3 of 52 samples and concentrations were below 1 microgram/L. Cyanobacterial cells were rarely detected in treated waters. Bank filtration removed cyanobacteria and microcystins satisfactorily however the efficiency varied depending on the plant and circumstances. The surface waterworks efficiently removed microcystins.

Zhao HX, Mold MD, Stenhouse EA, Bird SC, Wright DE, Demaine AG, Millward BA. Diabetic Med (2001) 18(9) p709-717

It has been suggested that nitrate, mercury and arsenic in potable water may increase the risk of developing childhood-onset Type 1 diabetes mellitus and zinc may have a protective effect, however study results have been inconsistent. This study examined the relationship between zinc, magnesium and nitrate in drinking water and childhood-onset Type 1 diabetes mellitus using the Cornwall and Plymouth Childrenās Diabetes Register (CPCDR) and local water quality data.

The CPCDR includes all children aged 0-15 years who were diagnosed with Type 1 diabetes between 1 January 1975 and 31 December 1996. There were 517 children from the register for whom water quality data was known. The water company provided water quality data for nitrate, nitrite, copper, magnesium, zinc, iron, aluminium and calcium concentrations in the 40 water supply zones covered by the study. Standardised incidence ratios (SIR) of Type 1 diabetes were calculated for each water supply zone using the UK 1991 census data. Analysis on thirds of the data set with SIR by x² tests for trend and Poisson regression analysis was undertaken.

The total child population in the study area was 157,000. In the whole study area between 1993-97, Ca and Fe showed an increasing trend and Al and Mn showed a decreasing trend. Mean copper concentrations showed a significant decrease with SIR (x² tests for trend = 6.58, d.f = 1, P =0.01), and a significantly decreased risk in the higher concentrations (middle and top thirds). Significant decreases were also found with an increase in magnesium concentration (x² tests for trend = 6.39, d.f = 1, P =0.01), and a significantly decreased risk in the top third.

The mean nitrate levels showed borderline significance of a decreased trend (x² tests for trend = 3.899 d.f = 1, P =0.01). Other variables examined showed no significant relation with the SIR of the disease. The data suggests a protective effect of increasing levels of zinc or magnesium, although dose-response trends were not consistent. A reduction in risk associated with high zinc levels has been previously reported. The study did not find an association between nitrate in drinking water and risk of diabetes, as reported in some previous studies.

Graves CG, Matanoski GM, Tardiff RG. Regul Toxicol Pharmacol (2001) 34(2) p103-124

This paper reviews 13 studies on human reproductive and development effects of DBPs published since 1989, and a larger number of animal studies.

The review process indicated that the overall weight of evidence does not support an association with DBP exposure for the following reproductive or developmental outcomes: low birth weight, very low birth weight, preterm delivery, cesarean delivery, congenital anomalies by severity, spina bifida, cleft lip and palate, cardiac anomalies, gastrointestinal anomalies, genital anomalies, integument anomalies, musculoskeletal anomalies, chromosomal abnormalities and neonatal death.

Outcomes for which the weight of evidence was mixed, inconsistent or weak were the following: neonatal jaundice, all congenital anomalies/birth defects, all central nervous system anomalies, neural tube defects, respiratory anomalies, spontaneous abortion/miscarriages and stillbirth/fetal death.

Outcomes for which there was a suggestive positive association were: growth retardation including term low birth weight, intrauterine growth retardation or small for gestational age, small body length, cranial circumference and urinary tract defects.

The major limitation of the exposure assessment in the published epidemiological studies has been that exposure is based on residence rather than on individual exposure to DBPs at the appropriate time in pregnancy. Accurate measures of individual exposures are needed, taking account of variations in the water supply system, water consumption and exposure through dermal contact and inhalation.

Sohn W, Heller KE, Burt BA. J Public Health Dent (2001) 61(2) p99-106

The recommended fluoride concentration in US public water systems ranges between 0.7-1.9 ppm and is based on the assumption that water intake is higher in warmer climates. This study assessed fluid consumption among US children relative to the local climate. The study population included 3,869 children aged 1-10 years who completed a 24-hour dietary interview during the National Health and Nutrition Examination Survey (NHANES III).

Multiple regression analysis found that age, socioeconomic status, sex, race and ethnicity were significantly related to fluid intake. There was no significant association between amount of total fluid or plain water intake and mean daily maximum temperature, even after controlling for the other factors. This study suggests that the temperature related guidelines for fluoride in drinking water set in 1962 might need to be re-evaluated.

McConnell S, Horrocks M, Sinclair MI, Fairley CK. Intern J Environ Health Res (2001) 11(4) p299-303

There is debate over whether endemic waterborne disease exists in the developed world. This study examined gastroenteritis rates before and after the introduction of water treatment in rural communities in Victoria and South Australia. Disease incidence was measured by the number of requests for analysis of gastroenteritis related faecal specimens. Communities were identified with populations of 1000 or more that had upgraded their surface water supply between December 1992 and December 1996. To control for changes in the provision of pathology services, data on common laboratory test (mid stream urine samples (MSU)) that would not be influenced by rates of gastroenteritis was collected. Three water quality parameters were assessed: the percentage of water samples with coliforms or E. coli detected, and turbidity. Seventeen systems were included in the study; ten systems had filtration added to existing disinfection, four had both disinfection and filtration added to no existing treatment, and three had disinfection alone added to no existing treatment.

There were no consistent trends among groups of communities with the same type of change in treatment. Changes in water quality parameters did not correlate with changes in gastroenteritis related faecal requests. The results suggest that any reduction in gastrointestinal disease from the introduction of water treatment was small as a percentage of all gastroenteritis. Water treatment still remains however an essential measure to protect the public from the risk of outbreaks of waterborne disease.

Leclerc H, Mossel DAA, Edberg SC, Struijk CB. Ann Rev Microbiol (2001) 55 p201-234

This paper reviews the historical development of coliform organisms as indicators of water contamination, the evolving definition of this group of bacteria as detection methods have changed over the year, and the characteristics of different subgroups. The suitability of different coliform groups as indicators of water contamination is critically examined and the authors discuss the distinction between "index organisms" (signifying the potential presence of pathogens) and indicator organisms (signifying process failure). They conclude that among the coliforms, E. coli holds a unique place as being almost completely limited to the intestinal tract of humans and warm blooded mammals. However this organism, and other coliforms can not serve as reliable markers for all types of enteric pathogen, and more work is needed to assess the suitability of additional markers for the effective monitoring of drinking water safety.

Whan LB, Grant IR, Ball HJ, Scott R, Rowe MT. Letters Appl Microbiol (2001) 33(3) p227-231

The bacterium Mycobacterium paratuberculosis is the cause of Johneās disease in cattle and other ruminants and may be associated with Crohnās disease in humans. Potential routes of transmission to humans include recreational water, drinking water and milk. The aim of this study was to determine whether this organism can survive standard water treatment processes. Two strains of M. paratuberculosis were used, a bovine strain and a human strain. These were subjected to chlorine concentrations of 0, 0.5, 1.0 and 2.0 micrograms ml^-1 for 15 and 30 minutes.

The chlorine concentrations used did not entirely kill either strain of M .paratuberculosis, Log₁₀ reductions in the range 1.32-2.82 were found. The largest Log₁₀ reduction with the bovine strain was 2.83 and with the human strain it was 2.35. The combination of the highest chlorine concentration (2.0 microgram ml^-1) and the longest contact time (30 min) resulted in the greatest reduction in culturability.

Brown CM, Cann JW, Simons G, Fankhauser RL, Thomas W, Parashar UD, Lewis MJ. Epidemiol Infect (2001) 126 (3) p425-432

An outbreak of gastroenteritis occurred at a large resort hotel in Bermuda in February 1998. The investigation included assessment of possible food and waterborne sources. Water from the hotelās potable water supply was tested for coliform bacteria and residual chlorine, but no food samples were available for analysis. Stool specimens from sick people were analysed for a range of bacteria, parasites and viruses. There were 448 people identified with at least one gastrointestinal symptom. Those who ate and/or drank at the hotel were significantly more likely to report gastroenteritis than those who didnāt (OR=6.0, 95% CI = 2.4-15.1). The laboratory analysis found genogroup-II Norwalk-like virus (NLV) in 18 of 19 tested stool samples. NLV of the same genotype and DNA sequence was found in a 3L water sample. There were no bacterial or parasitic pathogens identified in faecal specimens.

The hotelās water supply was not chlorinated and the underground storage tank was vulnerable to surface runoff and overflow from nearby toilets. There were no service logs or engineering records for the water supply, and the tank had not been cleaned for at least 5 years. The authors note that regions where tourism is important should consider some food and water safety initiatives such as the regular inspection and monitoring of drinking water supplies and waste water systems, as well as active surveillance so outbreaks are detected and controlled early.

Craun GF, Calderon RL. J AWWA (2001) 93(9) p64-75

In the United States from 1971 to 1998, there were 619 waterborne disease outbreaks reported in community water systems (CWSs) and noncommunity water systems (NCWSs), from both chemical and microbial causes. Of these, 113 outbreaks were caused by distribution system contamination, resulting in more than 21,000 cases of illness. Most of these cases of illness (81.7%) and almost all of the 498 hospitalisations occurred in CWSs, with 13 deaths recorded. The largest distribution system outbreak caused 5,000 illnesses. The causes of outbreaks included cross-connections, backsiphonage, contamination of storage tanks, mains breaks/repairs, and corrosion of household or commercial plumbing.

Since 1995, distribution system deficiencies were the cause of 45% of outbreaks in CWSs. Inadequate treatment of groundwater and surface water caused 20% and 15% of outbreaks respectively. In noncommunity water systems (NCWSs) outbreaks were mainly caused because of inadequate or no treatment of groundwater, although 14% of outbreaks reported since 1995 were caused by distribution system deficiencies. For both CWS and NCWSs surface water systems, significantly higher distribution system outbreak rates were reported compared with groundwater systems.

The authors note that not all waterborne disease outbreaks are recognised or investigated. It is estimated that only 10-30% of US waterborne disease outbreaks are reported. Outbreaks are more likely to be recognised and investigated in large CWSs and when they involve serious illness.

Craun GF, Frost FJ, Calderon RL, Hilborn ED, Fox KR, Reasoner DJ, Poole CL, Rexing DJ, Hubbs SA, Dufour AP. Int J Environ Health Res (2001) 11(3) p229-43.

This article summarises the outcomes of a workshop held in December 1998. A significant percent of waterborne outbreaks do not have the etiological agent identified, and/or the water system deficiencies and sources of contamination are not determined. If public health authorities and water utilities are to prevent waterborne outbreaks, then this type of information is needed. A number of recommendations were made by the workshop participants to improve the recognition and investigation of waterborne outbreaks, including enhanced surveillance, pre-planning of investigation methodology, use of standardised methods and questionnaires, and better training of personnel.

Hunter P. Commun Dis Public Health (2001) 4(2) p136-8

From 1991 to 1999 there were 22 outbreaks of Cryptosporidium infection in England and Wales associated with mains drinking water. From 1997-99 there were three outbreaks of cryptosporidiosis in the North West region of England. These outbreaks were linked to a chlorinated unfiltered supply drawn from a surface water reservoir in the English Lake district.

A retrospective analysis of laboratory reports was conducted in this region. Data from the PHLS Communicable Disease Surveillance Centre from laboratories in the North West Region of England was used in the analysis. Time series of weekly reported cases for the six implicated authorities involved in the three outbreaks and for the remaining unaffected authorities were analysed to estimate crude rates of reporting.

The average crude rate of case reporting for the six authorities was 2.8 times higher than for the other health authorities in the region ie 3.97 cases/week per 1 million people compared to 1.41 (p<0.0001). In a single week the maximum rate of reporting for the six authorities was 35.2 compared to 7.01 for the others. The six authorities experienced major peaks in Cryptosporidium infection that were not observed in the rest of the region. These peaks may represent unrecognised outbreaks due to the unfiltered surface water source that was implicated in the three reported outbreaks.

The reasons these outbreaks were not recognised at the time is not obvious. The overall lower numbers of positive laboratory reports in the initial years of the study may be an explanation, or it may be that consultants and environmental health officers in affected areas accepted that a rise in incidence in spring was normal and therefore did not warrant investigation. The authors conclude that excess rates of cryptosporidiosis in the affected health authorities suggest that waterborne disease in the North west region may be more common than earlier thought.

Robertson LJ, Gjerde B. Scand J Public Health (2001) 29(3) p200-207

A total of 408 raw water samples were collected between June 1998 and November 1999 from 147 sites across Norway and analysed for Cryptosporidium and Giardia. Parasites were detected in 103 samples from 47 sites. Cryptosporidium oocysts were found in more sites and in more samples than Giardia cysts. Maximum concentrations of Cryptosporidium oocysts were 3.75/10 litres and Giardia cysts 2/10 litres.

There were no statistically significant associations between seasons and detection of protozoa in the water samples. Samples with turbidity of 2.0 NTU or more were significantly more likely to contain parasites, and there was also a correlation with animal numbers in the catchment.

Conroy RM, Meegan ME, Joyce T, McGuigan K, Barnes J. Arch Dis Childhood (2001) 85 (4) p293-295

A cholera outbreak occurred in an area of Kenya between November 1997 and January 1998 where a field trial of solar disinfection of drinking water had recently been completed. The trial had involved all families with children under 6 years of age living in the Kajiado District. In the field trial, families were randomised into either solar disinfection or control conditions. In the solar disinfection group, drinking water for children under 6 was kept in plastic bottles on the roof. In the control group, childrenās drinking water was kept in plastic bottles indoors.

Households were visited within six weeks of the cholera outbreak and interviewed to determine whether illnesses meeting specific criteria to be diagnosed as cholera had occurred. Cases of cholera were found in 31 of 131 households. In the 67 households using solar disinfection there were 155 children under 6 years and there were 144 in the control group. Three cases of cholera were recorded in children in the solar disinfected group and 20 cases in the children in the control group, odds ratio 0.12 (95% CI 0.02 to 0.65, p=0.014). For children aged 6-15 and for adults there were no significant differences in cholera rates between the two groups.

These observations therefore support the value of solar disinfection, which is a low to no cost intervention method that can be utilised quickly by the community as a first line of defence.

Rijal GK, Fujioka RS. Wat Sci Technol (2001) 43 (12) p155-162

This study evaluated the efficiency of two solar systems to disinfect drinking water. The Family Sol*Saver (FSP) system using a non-UV transmittable cover sheet so sunlight heats the water (System A) was compared with the modified FSP systems with a UV transmittable cover sheet where heat and direct solar radiation are used to treat the water (System B). The two systems were set up on the roof of a building at the University of Hawaii. Stream and diluted sewage water samples were tested in each system under sunny and partly sunny conditions. Disinfection efficiency of both systems was assessed by reduction of the natural populations of faecal coliform, E. coli, enterococci, C. perfringens, total heterotrophic bacteria, hydrogen sulphide producing bacteria and FRNA virus.

Under sunny and partly sunny conditions water was heated to the critical 60 degrees C in both systems and concentrations of faecal coliforms, E.coli and enterococci were inactivated to undetectable levels of <1 CFU/100mL within 3-5 hours of exposure to sunlight. System A showed a faster temperature increase but a lower inactivation rate than system B. A synergy between solar radiation and heat was suggested in system B. This synergistic effect was observed in water temperatures as low as 41-50 degrees C.

Under sunny conditions both systems were able to disinfect C. perfringens spores and FRNA virus. Inactivation of C. perfringens was enhanced by the synergistic effect of system B. When conditions were cloudy neither system could reduce levels of faecal indicator bacteria to <1 CFU/100mL reliably. Generally, the gram-negative enteric bacteria were reduced more rapidly than gram-positive enterococci.

Mintz E, Bartram J, Lochery P, Wegelin M. Am J Public Health (2001) 91(10) p1565-1570

There are approximately 1.1 billion people in the world who do not have access to safe water sources. This paper reviews two low-cost technologies, which can be used to improve drinking water quality in developing countries.

Boiling of water inactivates viral, parasitic and bacterial pathogens but is not sustainable economically and environmentally and there is the risk of scalding. A variety of point-of-use chemical agents for water treatment have been reviewed. Sodium hypochlorite seems to be the safest, most effective and least expensive, and has been demonstrated to reduce diarrhoeal illness by up to 85%. The limitations of this disinfectant are it is ineffective against parasites and viruses, and when water with large amounts of organic material is treated the efficiency is reduced and a disagreeable taste and odour may be produced.

Solar disinfection by means of ultraviolet radiation and/or increased temperature can be used to inactivate pathogens in water, using clear plastic soda bottles or bags made of polyethylene terephthalate which transmit ultraviolet A. These are inexpensive and widely available. Field trials in Kenya have shown that this method is an acceptable and effective way of to improve water quality with significant reductions in the incidence of diarrhoea in children. The limitations are that sufficient solar radiation and reasonably clear water are required, and treating large volumes is also a problem.

Water that has been disinfected can be contaminated during collection, transport and storage. By replacing unsafe water storage vessels with safer ones with tight-fitting lids and narrow mouths, rates of disease can be lowered. The introduction of a new vessel or disinfection processes must be accompanied by changes in behaviours. Improvements in the quality of drinking water provide far more benefit when coupled with improvements in hygiene and sanitation. Self-sustaining and decentralised approaches to providing safe drinking water reviewed here, target those most affected directly, enhance health and contribute to productivity and development.

Position of the American Dietetic Association: The impact of fluoride on health.
Palmer CA, Anderson JJB. J Am Dietetic Assoc (2001) 10(1) p126-132.

Aluminium as a risk factor in Alzheimer's disease, with emphasis on drinking water.
Flaten TP. Brain Res Bull (2001) 55(2) p187-96.

Benefit-cost estimation for alternative drinking water maximum contaminant levels.
Gurian PL, Small L, Lockwood JR, Schervish MJ. Wat Resources Res (2001) 37(8) p2213-26.

Is the water safe for my baby?
Balbus JM, Lang ME. Pediatr Clin North Am (2001) 48(5) p1129-52.

The role of natural organic matter during formation of chlorination by-products: A review.
Nikolaou AD, Lekkas TD. Acta Hydrochim Hydrobiol (2001) 29(2-3) p63-77.

Hemolytic-uremic syndrome and Escherichia coli O121 at a Lake in Connecticut 1999.
McCarthy, T. A., N. L. Barrett, et al. (2001). Pediatrics 108(4): 991-2.

HACCP (Hazard Analysis and Critical Control Points) to guarantee safe water reuse and drinking water production - a case study.
Dewettinck T, Van Houtte E, Geenens D, Van Hege K, Verstraete W. Wat Sci Tech (2001) 43(12) p31-8.

An examination of risk factors associated with beef cattle shedding pathogens of potential zoonotic concern.
Hoar BR, Atwill ER, Elmi C, Farver TB. Epidemiol Infect (2001) 127 (1) p147-155.

Participant Blinding and Gastrointestinal Illness in a Randomized Controlled Trial of a In-Home Drinking Water Intervention.
Colford, J. M., J. R. Rees, et al. (2002). Emerg Infect Dis 8(1).

Genotyping Cyptosporidium parvum by single-strand conformation polymorphism analysis of ribosomal and heat shock gene regions.
Gasser, R. B., X. Zhu, et al. (2001). Electrophoresis 22(3): 229-43.