Health Stream Issue 20 Dec 2000

The Australian Commonwealth Government recently commenced a three year project on rural and remote potable water supplies with the aim of developing a risk assessment framework for prioritising water improvement needs. The project is being undertaken by the Environmental Health Section of the Commonwealth Department of Health and Aged Care, in collaboration with the enHealth Council, and the Bureau of Rural Sciences (BRS).

The project forms part of the Australian National Environmental Health Strategy (NEHS) which was launched in October 1999. The NEHS is based around a set of entitlements and responsibilities, one of the most fundamental being the entitlement to “access to safe and adequate supplies of water” as recognised in United Nations Declaration of Human Rights.

During the development of the Implementation Plan for the NEHS, potable water supplies in rural and remote communities were identified as a priority issue for attention. While it is known that many small community supplies have problems with both the quality and quantity of available potable water, a comprehensive information base is lacking, making it difficult to develop improvement plans and prioritise actions. This project seeks to develop such a database and collect sufficient information to allow relative health risks and community needs to be characterised on a national basis.

The steering committee for the project is composed of a core of key stakeholders from the enHealth Council, Commonwealth Health & Aged Care, the Aboriginal and Torres Strait Islanders Commission, the CRC for Water Quality and Treatment, and the National Health and Medical Research Council. Additional organisations are likely to be become involved in the later stages as the project progresses.

In the first phase of the project during 2001, the BRS will conduct a survey of water supplies which serve between 10,000 and 50 people, and fulfill the Australian Bureau of Statistics definition of "rural". Water authorities and others who manage community water supplies will be asked to give a basic description of their system, any current microbial and chemical monitoring programs, and problems with water quality or quantity. Water quality data will also be sought for inclusion into a national database.

The coverage of the compiled data will be assessed and a strategic sampling program will be developed and implemented during the second and third years of the project to fill identified data gaps. Information from previous studies such as the recently completed ATSIC Community Housing and Infrastructure Needs Survey, and data already held by BRS will also be incorporated.

Water quality data and other information relating to the water source, level of protection and type of water treatment will then be used to assess the comparative level of health risk for each water supply. Community consultation on needs and preferences will also be an integral part of the process. The outcomes of the project will serve as the basis for the development of ongoing responses at the federal, state and local level by a range of government departments and agencies.

The New Zealand Ministry of Health recently released the Drinking-Water Standards for New Zealand 2000 (DWSNZ 2000). The new Standards will take effect from I January 2001, replacing the previous 1995 version. Although the term standards is commonly used to mean compulsory regulations, the NZ Standards are in fact voluntary, making them similar in nature to the Australian Drinking Water Guidelines.

The DWSNZ 2000 incorporates a number of revisions, including changes in several aspects of microbiological monitoring. E.coli has been designated as the sole indicator organism, with the relatively new DST methodology (defined substrate test for the beta-glucuronidase enzyme system) as the referee method for laboratories. Changes have also been made in the statistical methods used to derive monitoring frequencies for supplies of different sizes, resulting in a lower frequency in many cases. Monitoring frequency may be further reduced if specified levels of free available chlorine leaving the treatment plant are continuously monitored. Safeguards against protozoal pathogens are based on documenting the security of groundwater sources, and treatment requirements for surface water and groundwater under the influence of surface water.

In November the Ministry of Health also released a series of draft Public Health Risk Management Plans for water supplies for public comment. The plans are intended to provide water suppliers with guidance on potential sources of public health risk in different components of water supply systems, and the appropriate preventive or remedial measures which should be undertaken to remove or reduce these risks.

Aspects of water supply systems covered by the PHRMPs include; raw water, source abstraction, pre-treatment processes, coagulation /flocculation processes, filtration, disinfection, aesthetic property adjustment, and the reticulation network. The PHRMPs also address the importance of staff training and system monitoring in ensuring safe water supplies.

For each aspect of water supply, several PHRMP modules have been prepared. The modules describe the nature of a particular risk, the reasons why it poses a threat to public health, links to other elements in the water supply system, and a comparative level of risk. In order to apply the plans, water utilities are first required to prepare a flow chart of their water supply detailing the various components. The relevant PHRMPs are then selected for each component of the supply, and the plans are tailored for the characteristics of the individual system. Once the overall structure has been determined, the water utility is to work through the plans to identify:

While the draft PHRMPs are intended to cover the major sources of risk, water utilities should also consider the possibility of events not listed in the plans, and the characteristics of their specific supply system. Existing management practices and operating procedures should then be compared to the theoretical plan derived from the PHRMPs and any missing preventive measures should be noted. The water utility should then develop a priority action list for achieving any missing preventive measures that may be required, taking into account the level of risk and the cost of undertaking prevention.

The Draft PHRMPs are open for public comment until 28th February 2001. When the plans have been finalised, they will be enacted as an amendment to the NZ Health Act. Water utilities will then be required to begin a phased implementation over 5 years, beginning with larger supply systems.

An advisory committee comprising US EPA and stakeholder representatives has reached an Agreement-in-Principle on the framework for the Stage 2 Disinfectants/Disinfection Byproducts Rule (D/DBP Rule) and the Long-term 2 Enhanced Surface Water Treatment Rule (LT2ESWTR). The agreement follows 18 months of negotiation between the EPA and 22 stakeholder organisations. The two rules are scheduled for promulgation in May 2002.

The D/DBP Rule is aimed at lowering levels of disinfection byproducts in drinking water in order to reduce the potential risk of cancers which may be associated with prolonged exposure to high levels of these chemicals. The LT2ESWTR aims at achieving better control of pathogen levels in drinking water and is particularly targeted at Cryptosporidium. While the two Rules have separate status, they were negotiated in tandem in recognition that complex risk trade-offs are required to simultaneously achieve both aims ⁽¹⁾.

The Agreement-in-Principle also includes two recommendations for EPA action in new areas; firstly the development of a program for the analysis and control of public health risks originating from distribution systems, and secondly the development of national water quality criteria for microbial pathogens for stream segments designated as drinking water sources.

D/DBP Rule This rule will require a reduction in peak DBP levels through changes in the way that compliance monitoring is carried out for trihalomethanes (THMs) and haloacetic acids (HAAs). Rather than the present arrangement where sampling locations are located at representative sites, the new Rule will target sampling at peak DBP occurrence sites. Requirements for the timing of quarterly samples will also be tightened to require at least one sample date in the peak historical month for DBP levels, and less variation in the number of days between samples.

Water utilities will first be required to carry out a 1year survey of their water supply system in order to identify sampling locations where peak levels of THMs or HAAs occur. The format of the Initial Distribution System Evaluation (IDSE) varies according to the size of the population served by the system, whether it supplies ground or surface water, and the type of disinfectant used (chlorine or chloramine). Systems with historically very low DBP levels (all samples in last 2 years less than 40 micrograms/L THMs and 30 micrograms /L for HAAs) will not be required to carry out the ISDE.

Once peak DBP occurrence sites have been identified, long term compliance will be based on a Locational Running Annual Average (LRAA) calculated for each site. For large surface water systems, quarterly sampling must be carried out at 4 distribution system sites comprising 1 representative average site from among current sampling locations, 1 site with the highest HAA level and 2 sites with the highest THM levels identified from the ISDE.

Requirements for compliance with the new Rule will be phased in gradually over several years following its promulgation, and during the phase in period, compliance with the prevailing Stage 1 D/DBP rule must also be maintained:

LT2ESWTR This rule will apply to all surface water supplies and ground water supplies under the direct influence of surface water, except those which are exempted from filtration. Water supplies will be initially classified into categories (termed "bins") in terms of their Cryptosporidium risk in source water. Treatment requirements for measures to remove of oocysts in addition to conventional treatment are then specified in terms of the number of logs removal. Conventional treatment is specified as coagulation, flocculation, sedimentation and granular media filtration. Conventional treatment plants which already operate in compliance with the Interim Enhanced Surface Water Treatment Rule are considered to achieve 3 logs removal of Cryptosporidium.

For large systems, the bin classification will be based on the results of a 24 month monitoring program for Cryptosporidium using 10 litre samples tested by the EPA Method 1622/23 technique ⁽²⁾. Systems which already have equivalent monitoring data using this technique will not be required to carry out a new monitoring program. Those which already provide 2.5 logs of removal in addition to conventional treatment will be exempt from the monitoring requirement. For small systems a limited program of Cryptosporidium monitoring triggered by high E.coli levels has been proposed, but has not yet been finalised.

Bin Requirements Table

Bin No.	Average *Cryptosporidium* Concentration (a)	Additional treatment requirements (b)
1	< 0.075/L	No action
2	0.075/L to < 1.0/L	1.0 log treatment
3	1.0/L to < 3.0/L	2.0 log treatment (c)
4	>=3.0/L	2.5 log treatment (c)

(c) systems must achieve at least 1-log of the required removal treatment using ozone, chlorine dioxide, UV, membranes, bag/cartridge filters, or in-bank filtration. (1 log = 10-fold, 2 logs = 100-fold etc.)

In recognition of the widely varying characteristics of water supplies and the diversity of methods that may be employed to reduce Cryptosporidium numbers in finished water, a high degree of flexibility is permitted in achieving the required level of removal. The Agreement describes a "Microbial Toolbox" of preventive and remedial measures under the headings of Watershed Control, Alternative Water Sources, Pretreatment, Improved Treatment, and Improved Disinfection. Under each heading, specific measures are listed together with their associated "potential log credit". By selecting the most appropriate and practical combination of measures for each water supply, a water utility can aggregate credits to provide the required level of additional treatment.

In several areas, water utilities have the option of gaining "extra" credit for a given control measure by carrying out system specific studies to demonstrate that its effectiveness is greater than the standard assumed value. In addition, there is provision for utilities to demonstrate the effectiveness of alternative technologies not listed in the Microbial Toolbox. Enhanced performance verified by peer review programs may also be recognised as evidence of effectiveness (eg the Partnership for Safe Water Phase IV).

Following promulgation of the new Rule, and the classification of a water supply system into a "Bin", the utility will have up to 3 years to achieve compliance. An additional 2 years may be granted where capital expenditure is required. The Agreement-in-Principle also recognises that the rate of implementation of the Rule may be limited by the availability of accredited laboratories to carry out Cryptosporidium monitoring by the approved method.

It is anticipated that many utilities will choose to add UV treatment to their conventional treatment plants as this technology alone is considered to provide an additional 2.5 logs of removal. However it is acknowledged that while small scale experiments have indicated that UV is highly effective against Cryptosporidium, information is presently lacking on its suitability for full scale plants, and on many practical engineering issues. In the interval prior to promulgation of the Rule, the EPA has undertaken to develop standards for compliance of UV treatment systems, dosage tables and appropriate guidance manuals. The EPA will also carry out an estimate of public health effects, and a health risk reduction and cost analysis. The Agreement also includes an undertaking that the EPA Science Advisory Board will be asked to comment on the proposed Rule.

The EPA proposal to reduce the Maximum Contaminant Level (MCL) for arsenic from 50 ppb to 5 ppb (ppb=micrograms/litre) has drawn criticism from many sources including the American Water Works Association (representing US and international water supply professionals) and the Association of State Drinking Water Administrators (representing US drinking water program regulators).

After the public issue of the proposed Rule, the AWWA found that the cost calculation formulae made available by the EPA did not match those described in the Regulatory Impact Assessment. A revised set of formulae were released by the EPA in a Notice of Data Availability, but again according to the AWWA, these did not match the unit cost curves in the RIA. Thus it not possible for other parties to carry out independent verification of the EPA cost calculations.

Numerous aspects of the cost-benefit analysis have also been questioned by the Drinking Water Committee of the EPA's own Scientific Advisory Board (DWC/SAB). The DWC/SAB conducted a 3 day meeting to examine the issues and produced a report to the EPA in September this year. The committee commented that its deliberations were hampered by a lack of detail on the operation of the "decision tree" developed by the EPA to estimate the likely costs of different treatment options for water supplies. Based on the information available to the committee the model did not appear to account for the costs of several important components (eg land acquisition, training and certification of operators, chemical costs), EPA assumptions about the efficiency of arsenic removal treatments in full scale plants are generally unproven, and the presentation of single cost figures does not indicate the degree of uncertainty in the underlying assumptions.

The committee also concurred with submissions from the water industry that disposal of arsenic-containing waste to sewers or streams would often not be possible either due to toxicity or other characteristics such as TDS load. Disposal mechanisms will therefore add significantly to the costs of compliance.

The DWC/SAB expressed the opinion that the EPA had misinterpreted some of the conclusions of the National Research Council's (NRC) arsenic report. In this report the NRC carried out an assessment to determine whether the available human data were adequate to provide the basis for risk modelling. The EPA has accepted this exercise as an actual risk assessment, rather than exploring alternative methods of assessing risk. Overall the DWC/SAB suggested that in the light of large uncertainties both in costs and health benefits, a phased reduction in the arsenic MCL under the Rule would be preferable. This would provide the opportunity to verify cost assumptions and improve estimates of the health benefits. The EPA is required to issue the final arsenic regulation by June 22, 2001.

The boil water advisory for the Canadian town of Walkerton was lifted on 5 December, more than 6 months after it was imposed because of a waterborne disease outbreak. The outbreak, predominantly caused by E.coli O157 and Campylobacter, resulted in 7 deaths and widespread illness, and led to the establishment of a judicial inquiry by the Ontario government ⁽¹⁾.

An epidemiological report, released by the Bruce-Grey-Owen Sound Health Unit on 10 October, confirmed the most likely cause of the outbreak as contamination of one of three wells serving the town by runoff from a cattle farm following heavy rainfall and flooding in mid-May.

Testing of livestock at 13 farms within a 4km radius of the wells showed the presence of Campylobacter at nine farms, and both E.coli and Campylobacter at two farms. Molecular finger-printing of the bacteria showed those at the farm closest to Well No. 5 were identical to the isolates in the majority of human infections during the outbreak. This well was closed when the outbreak was detected. The flooding in May is believed to have overwhelmed the water disinfection system with an increased pathogen load and high turbidity, however low level contamination probably occurred at least one month earlier as some people who became ill in April had the same E.coli genotype.

Investigation of the hydrological characteristics of the wells supplying the town have shown that all three may be subject to surface water influence. According to a report by consultants engaged by the Ontario Ministry of the Environment, pools of surface water can be observed to infiltrate into the ground and a nearby natural spring reverses its flow whenever the pump at Well No. 5 is activated. These clear indications of surface water infiltration and a long history of turbidity increases and faecal indicator bacteria in the well resulted in the recommendation that the well be plugged and abandoned, as it cannot be made secure from contamination. Several other wells which have not been used for some years have also been plugged to improve the security of the groundwater supply.

Wells No. 6 and 7 were found to have a lesser degree of surface water influence, and have been retained as the town's water supply. However a hollow-fibre ultrafiltration system and improved chlorination equipment have been installed to ensure that water treatment is adequate to deal with potential contamination events. A control system of on-line monitoring, alarms and automatic shutdown has been installed, and a wellhead protection program will also be undertaken to provide a higher degree of security. In the longer term, the feasibility of providing an alternative water supply (by pipeline from another source, or from a new well field), or alternative treatment for the current supply will be explored.

Prior to the lifting of the boil water notice, a extensive decontamination program was carried out on the Walkerton water supply system at an estimated cost of CA$11 million. This included:

More than 5,000 water samples were tested to confirm the effectiveness of the decontamination program. The program was audited by the Ontario Ministry of the Environment which concluded that restoration work had brought the Walkerton supply into compliance with the Ontario Drinking Water Standards. On receiving this advice and the results of independent monitoring tests, the Bruce-Grey-Owen Sound Health Unit recommended the lifting of the boil water notice, and this was supported by the Chief Medical Officer of Health for Ontario.

The Walkerton Inquiry ⁽²⁾ began hearing evidence on 16 October and is expected to continue for several months. A number of witnesses have testified that the town's chlorination equipment was unreliable or improperly operated, and that records were falsified prior to inspections by Ministry of the Environment officials. The Walkerton Public Utilities General Manager, who reportedly denied problems with the water supply when directly questioned by local health officials at the time of the outbreak, is expected to give evidence in late December.

A team of researchers from the University of York recently delivered a major report on the human health effects of drinking water fluoridation to the British government. The report reviewed evidence on five questions:

A total of 3231 references and submissions were considered and assessed against predetermined relevance criteria. The criteria were fulfilled by 732 references. These references were then examined to determine whether they fitted inclusion criteria relevant to positive effects (dental caries prevention) or negative effects (dental fluorosis, osteoporosis, cancer etc). The quality and validity of each of the 214 included studies was then rated on a points scale for specific design features and a three level scale for susceptibility to bias. The review found that research in this area was only of low to moderate quality, with few studies achieving high scores for validity and freedom from potential bias.

Objective 1: (26 studies) The evidence suggests that fluoridation reduces caries prevalence whether this is measured by the proportion of children who are caries-free or by changes in the average number of decayed /missing/ filled teeth (DMFT). Due to the limited quality of the studies it is difficult to ascertain the exact degree of benefit, however it was estimated that for 1 additional person to remain caries-free, six people would needed to receive fluoridated water.

Objective 3: (15 studies) Research was limited and of particularly poor quality, but there was some evidence that fluoridation reduced health inequalities in relation to DMFT in 5-12 year olds. However no difference was seen in caries-free prevalence in 5 year olds. The reviewers urged caution in interpreting these results.

Dental Fluorosis (88 studies) This was the most widely studied negative effect of fluoridation, but studies were mainly of low quality and susceptible to bias. Nevertheless a dose-response relationship with water fluoridation was evident. It was estimated that at 1.0 ppm about 12.5% of people may have fluorosis of aesthetic concern.

Bone fracture and bone development problems (29 studies) This was the next most frequently studied adverse effect after dental fluorosis. For hip fractures there was sufficient evidence to carry out a meta-analysis which showed no association of fractures with fluoridation. For other fractures the evidence was less abundant but again no effect was evident.

Cancer (26 studies) There was no association between fluoridation and the incidence of all cancers, nor with osteosarcoma and other bone/joint cancers. No association was seen in 2 studies of thyroid cancer.

Other possible negative effects (33 studies) a diverse range of potential adverse effects has been studied, but most research has been of poor quality. There is insufficient evidence to draw conclusions.

Objective 5: Very few studies have compared the effects naturally and artificially fluoridated waters. From the little evidence available, no major differences in effect were apparent.

In its discussion the review team commented on the methodological limitations and poor quality of much of the research on the effects of water fluoridation, and the consequent difficulties in interpretation of the results.

Vancouver Turbidity Study

A Canadian research group has released a report describing a relationship between drinking water quality and measures of community gastro-enteritis in Vancouver. The report has not been subject to formal independent peer review as is the normal procedure for publication in high quality scientific journals, although the authors circulated draft versions for comments by other scientists. Not all reviewers supported the methodology nor the interpretation of the results.

The Greater Vancouver Regional District has a population of about two million, and is served by an unfiltered chlorinated water supply. This supply is drawn from three catchments which are closed to public access and farming. Although it is not mentioned in the report, substantial logging activity has been permitted in the catchments, producing deforested areas and sediment runoff during heavy rainfall. Turbidity peaks up to 19 NTU were recorded during the study period.

The study used a complex statistical modelling analysis to look for correlations between water quality parameters (mainly turbidity) and three measures of gastroenteritis (hospital admission data, physician billing data, and paediatric hospital emergency room visit data) from 1992 to 1998. The analysis was conducted separately for the three water supply areas and zones with mixed supply were excluded.

It was estimated that variations in drinking water turbidity were associated with 1.6% of all gastroenteritis-related physician visits, 0.6% of gastroenteritis-related hospital admissions and 1.6% of gastroenteritis-related paediatric hospital emergency room visits over the 6 year period.

Comment The methodology is similar to that of previous studies in the Philadelphia water supply, however here the measured changes in turbidity are much greater. In the previous studies serious doubts about the data quality and methodology were raised by an EPA peer review (See Health Stream Issues 8 and 9).

The New York health department has been criticised for delaying notification to the Mayor's Office of several cases of bottled water contamination. In early September several people reported illness after drinking two different brands of bottled water. The health department began an investigation and notified local police and the FBI, but did not contact city officials who were unaware of the incidents until questioned by the media several days later. It was reported that several of the cases involved ammonia contamination.

Two brands of British mineral water were recalled on 22 November when tests showed the presence of faecal indicator organisms in samples taken in October and November. The recall affected supplies for office water coolers as well as bottled water in a range of containers. The natural mineral water comes from bores in a chalk aquifer in the Chiltern Hills, which is believed to have become contaminated after heavy rainfall.

A suspected waterborne outbreak of Campylo-bacter has occurred in a housing estate in Wales. The estate is situated on elevated ground and supplied from a gravity feed tank fed by a pump from the main municipal supply. Fifteen people became ill between 17 and 24 September, with Campylobacter infection confirmed in all but one case. There was no common exposures from food sources, however routine water samples taken from the water storage tank on 18 September were positive for total and faecal coliforms. Samples taken on and prior to 15 September were clear. The water supply system was chlorinated and flushed, and tests will be carried out to check the physical integrity of the tank.

Researcher Dr Thaddeus Graczyk has demonstrated transmission of Cryptosporidium parvum by flies feeding on calf faeces. The experiment involved trapping wild filth flies from a barn containing Cryptosporidium infected calves, and a control study with uninfected calves. Oocysts on the surface of the flies were removed and used to infect mice. The results suggested that flies could carry oocysts for at least three weeks, and infectivity was retained. Similar results were reported with flies raised on cattle faeces in the laboratory. Dr Graczyk has previously shown that C. parvum oocysts retain infectivity during passage through the digestive tract of birds (although they cannot grow in these hosts) and can be accumulated by oysters.

Comment This finding is not unexpected - indeed the "5 Fs" (food, fingers, faeces, flies and fomites) have long been recognised as the main vectors of infectious gastrointestinal disease.

The US Marine Corps is to undertake a study of an estimated 16,500 military families who may have been exposed to chemically contaminated drinking water at Camp Lejeune in North Carolina. The chemicals tetrachloroethylene and trichloroethylene were first discovered in the groundwater supply to the camp in 1982, and the wells were capped in 1985. The contamination is believed to have come from a dry cleaning business. It is suspected that the chemicals may be associated with increased risks of childhood cancers and birth defects.

Contact Information

The printed version of Health Stream is available free of charge - to be added to our mailing list please contact Pam Lightbody (email above or fax + 61 3 9903 0576). Past issues can be found under Publications, and there is a searchable archive of articles, news items and literature summaries.