Program 1 Newsletter - Issue 18

An algal bloom in South Australia's Yorke Peninsula has led to the discovery a new cyanobacterial toxin. The bloom in the Paskeville reservoir led to a health alert being issued for about 15,000 residents in 35 towns. Many visitors were also affected by the alert in the popular tourist area during the Easter holiday period. The incident began on Friday 14 April with a number of complaints from consumers about an unpleasant taste and odour in tap water. Tests detected the presence of MIB (2-methyl isoborneol), a cyanobacterial metabolite which although not considered toxic is often responsible for "musty" taste and odour in water supplies.

Investigations by SA Water revealed the source to be one of two shallow earthern storage reservoirs in Paskeville where filtered water was stored prior to chlorination and distribution. A cyanobacterial "mat" growing on the floor of the reservoir had broken up and become detached, resulting in algal fragments contaminating the water intake. The reservoir was removed from service and examination of algal samples showed the bloom to be a Phormidium species. This filamentous cyanobacteria commonly grows on the surface of submerged rocks in shallow clear water bodies, and has not been reported to produce toxins. It is generally not regarded as a concern in drinking water supplies. The local media were notified of the bloom and advised there were no health concerns. The South Australian Department of Human Services was also notified.

Despite the absence of reported toxin production by Phormidium, algal samples were collected by SA Water for toxicity testing by intraperitoneal injection in mice. These tests showed the material to be toxic, and a health alert was issued on Tuesday 18 April. Due to uncertainty over the identity of the toxin the South Australian Department of Human Services advised consumers not to use tap water for drinking, making cold or hot beverages, washing or cooking foods, making ice or cleaning teeth. Bottled water was distributed free through supermarkets and local fire stations. Farmers were warned not to use tap water for livestock.

A program of mains flushing was commenced to clear the affected water from the distribution system. Due to the large area affected, it was several days before the all clear was issued to the last townships on Tuesday 25 April. According to local media, no instances of human illness or stock losses have been reported. Shortly after the incident the state government announced a $9 million project to cover and line five major water storages including the Paskeville reservoirs. The project is part of a $36 million Country Water Quality Improvement Program to be carried out in South Australia over the next 5 years.

Preliminary characterisation of the toxin has been carried out by the Australian Water Quality Centre and the University of Adelaide Department of Clinical Pharmacology (both participants in the CRCWQT). The toxin appears chemically distinct from the known major toxin classes (not a microcystin, nodularin, saxitoxin or cylindrospermopsin), and is somewhat less toxic in mouse tests than the most toxin microcystins. The effects are protracted, with injected mice generally surviving more than 6 hours and sometimes more than 24 hours before succumbing. Examination of internal organs shows no gross organ damage, but liver damage is evident on histopathological examination. It has also been established that the toxin is effectively inactivated by boiling and by water chlorination.

A similar protracted toxic effect from cyanobacteria has been reported previously for Aphanizomenon flos-aquae. In this case chemical characterisation of the toxin was not performed but it was reported that no gross organ damage was seen in mice, but liver and lung damage was evident on histopathological examination (1). Histopathological data from this toxin is now being compared with observations on the South Australian isolate.

The finding of a new toxin from an algal species hitherto regarded as innocuous raises concerns for water and health authorities. Phormidium growth on the bottom of shallow reservoirs is not unusual, and this species has been regarded as a nuisance organism only on the rare occasions when algal growth becomes detached and enters the water column. Now it appears that water authorities may need to consider testing for toxicity whenever Phormidium is identified, rather than assuming there are no health concerns.

A major waterborne outbreak of E. coli O157 has struck the small town of Walkerton, Ontario, about 180km north of Toronto. As of 7 June, seven fatalities had been attributed to the outbreak, with four more deaths under investigation, and several people still in serious condition. The deaths have occurred mainly in the elderly, but also included a 30 year old man and a 2 year old child. More than 1,000 people are believed to have been ill, with over 400 attending the emergency department of the local hospital with severe gastroenteritis symptoms. Schools, childcare centres and many businesses in the town have been closed for the duration of the outbreak.

In contrast to the harmless E. coli strains normally found in the intestine of humans and warm blooded animals, the O157 strain produces a potent toxin (verotoxin) which causes cell damage, leading to the development of bloody diarrhoea (haemorrhagic colitis). The infection may also result in severe dehydration and kidney damage requiring dialysis treatment (Haemolytic Uremic Syndrome, HUS). Fatalities may occur particularly among young children and the elderly, and some survivors suffer permanent kidney damage which may require life-long dialysis or a transplant.

Toxigenic E. coli (including O157 and other related strains) are carried by 10-15% of healthy ruminants (including cattle, sheep, goats and deer). The bacteria may be transmitted to humans by consumption of raw or undercooked meats, or by contamination of foodstuffs or water supplies with faeces from infected humans or animals. Outbreaks have also been associated with recreational water bodies, and direct contact with animals. The infectious dose may be as low as 5 to 10 organisms, and the incubation period ranges from 2 to 8 days. The bacteria are readily killed by chlorination of drinking water.

No specific treatment is yet available to combat the infection, and recent research suggests that antibiotics and anti-diarrhoeal medication may increase the risk of HUS developing. The Canadian government has granted special permission for victims of this outbreak to be treated with an experimental drug called Synsorb Pk developed by a Canadian biotechnology company. The drug binds the verotoxin produced by E.coli O157 infection in the gut, preventing it from reaching the bloodstream and thus reducing the risk of organ damage. The drug is presently completing the final stages of a 4 year clinical trial and is scheduled for release later this year.

The Walkerton community of 4,800 residents is served by a groundwater supply drawn from several wells. It is suspected that one of the wells became contaminated with animal waste after heavy rainfall in the area on May 12. The chlorination system for the water supply had reportedly been performing unreliably for some time and new equipment had been ordered.

Water samples taken on Monday 15 May tested positive for E. coli and the results were reported to the local water utility on Thursday 18 May. A program of flushing and chlorination was commenced to clear the contamination. The first signs of the outbreak occurred on 19 May when a local paediatrician reported two cases of bloody diarrhoea to the public health authority. By Sunday 21 May cultures of faecal specimens had confirmed E. coli O157 as the cause of the illness. Meanwhile more cases of severe gastroenteritis were reported by doctors and the local hospital, and on Sunday 21 May a boil water notice was issued by health authorities.

According to reports in Canadian newspapers, a police investigation is now underway into the actions of the Walkerton Public Utilities Commission which supplies water and electricity to the town. The utility has only nine employees, with two individuals primarily responsible for managing the water supply.

Allegations have been made that the utility did not report either the E. coli findings or the problems with chlorination equipment to public health officials even when directly questioned about the safety of the water supply. The Health Officer for the region has been quoted as saying he contacted the utility when the first cases of illness were reported on Friday 19 May and was assured that the water supply was safe and secure. These assurances were repeated when the utility was again contacted the following day.

Health investigators initially concentrated their investigation on the possibility of foodborne transmission, but by Sunday 21 May they had become convinced that the municipal water supply was the most likely source. A boil water order was issued and the Public Health Unit collected water samples for analysis. These samples were reported positive for E. coli on Tuesday 23 May. It is reported that only then did the Public Utilities Commission personnel inform health officials of the previous E. coli test results and the chlorination problems. Control of the town's water supply has now been handed over to the Ontario Clean Water Agency for a six month period while the investigation continues.

The outbreak is the first waterborne E. coli outbreak from a municipal supply reported in Canada, and has attracted intense national media attention. A number of legal actions on behalf of victims have been lodged with the courts, citing negligence on the part of various parties. The outbreak has also triggered alarm in other towns and among homeowners with private wells, and testing laboratories have reportedly been overwhelmed with requests for water tests.

Some commentators have charged that the Ontario governments cutbacks in funding for environment protection and the devolution of responsibilities to regional and municipal authorities led to a failure of public health protection. The Ministry of the Environment (MOE) budget has been reduced by 40% since 1995, and it has been alleged the ministry has adopted a “soft” approach to enforcement and prosecution. The closure of government owned water testing laboratories in 1996 has also drawn criticism, after it emerged that the private testing laboratory had no legal obligation to report E. coli detections to government officials.

The MOE has launched an internal investigation after admitting it failed to follow protocols to notify the regional Medical Officer of Health of several recent detections of total coliforms or E. coli in the Walkerton supply. These test results were reported voluntarily to the MOE by a different private laboratory which was performing routine tests for the Walkerton water utility. This laboratory ceased working for the utility only weeks before the current outbreak, and testing was taken over by a different private contractor.

In the face of mounting public pressure, Ontario’s Premier announced a full inquiry into the outbreak on 31 April. Opposition parties are calling for broad ranging terms of inquiry to cover environmental regulation and enforcement in relation to water quality protection, as well as an investigation of the Walkerton outbreak.

In 1998 Australian laboratories had been testing for the presence of Cryptosporidium oocysts and Giardia cyst for some years, and a number of tests procedures/techniques were developed for detection of these organisms in water. The National Association of Testing Authorities, Australia (NATA) was however not prepared to offer accreditation at that stage due to perceived lack of consensus on the technical validity and comparability of these methods.

The need for development of principles of accreditation for potentially parasitic protozoa was precipitated when high levels of these organisms were detected in the Sydney water supply in 1998, which in turn led to an inquiry by Peter McClellan QC (1). The final report of the Sydney Water Inquiry acknowledged the still developing state of science relating to Cryptosporidium and Giardia, especially in relation to water supplies, but it also stressed the need to restore public confidence in the water supply and recommended greater transparency in all stages of water quality monitoring.

One of the recommendations of the final report of the McClellan inquiry was that: “Laboratories providing parasitic analysis should be accredited by the National Association of Testing, Australia (NATA) utilising an approach similar to that implemented by the US EPA”.

The rather challenging task of developing guidelines for accreditation was overcome with the help of dedicated and enthusiastic professionals who provided NATA with their expertise and time. The working group of NATA’s Water Biology Technical Group includes :

The members of the working group have visited 7 laboratories to discuss the principles of proposed criteria and to gather more specific information. On the basis of these visits, the working group developed a checklist that covers relevant technologies and quality control requirements. The checklist was circulated to laboratories and placed on NATA’s website in November 1999. Accreditation will require total compliance with relevant sections of the checklist.

As agreed, the working group has also developed and implemented a plan for the provision of a regular proficiency program. Two preliminary rounds of this program were offered to laboratories in August 1999 and March 2000. Results of these rounds provided the basis for defining the performance acceptable for accreditation.

The acceptable range for percentage recoveries, in the first instance was set at 10% - 110%. This range has the possibility of changing once the confidence levels based on a history of proficiency testing data has been established.

As usual, individual laboratory testing performance will be monitored for each round of this program. Laboratories that have reported an extreme result (false positive/false negative) or reported low recoveries, will be required to instigate investigative action to identify the cause. Details of this investigative action, and any associated corrective action, will be required to be reported in writing to NATA by a stated date. Laboratories’ investigations will be reviewed and technical comments will be provided if required.

NATA’s Water Technical Group will continue its input into the review of criteria to ensure the program’s currency in light of ongoing technical developments and the performance of laboratories in the proficiency program.

One laboratory that voluntarily offered to be the first to be assessed against the checklist is now accredited. The rest of the visits are planned for July-September 2000. NATA is now planning its next year of accreditation and proficiency testing, including financial and program management cooperation with the Water Services Association Australia.

This 2 1/2 day conference sponsored by Health Canada and the US EPA brought together over 70 participants from a wide range of fields including water analysis, toxicology, water system management, risk assessment, public health and epidemiology. The aims of the workshop were:

The workshop consisted of 4 half-day sessions of presentations with a questonand answer session after each, plus a final morning of discussion led by a panel of experts. The main themes of discussion and outcomes are summarised here.

Speakers in this session emphasised the complexity of disinfection byproduct (DBP) formation and occurrence in water supply systems. The types and concentrations of DBPs formed depends on raw water composition (amount and nature of organic matter, bromide concentration), temperature, pH and the disinfectant used (chlorine, chloramine, ozone). The reactions leading to DBP formation continue as water moves through the distribution system, and some DBPs may also decompose due to chemical and/or microbial degradation. Even at a fixed point in a distribution system, DBP concentrations may vary markedly with time and water temperature.

More than 200 halogenated and 400 non-halogenated DBPs are known to be formed by chlorination and new compounds continue to be identified. For many compounds, standard assay methods are not available and little or nothing is known of their biological effects. Information on the occurrence of DBPs is also very limited except for those where monitoring has been a regulatory requirement.

During discussions it was agreed that DBP sampling programs designed to fulfill regulatory requirements or operational management purposes are inappropriate for short term health studies (ie reproductive outcome studies). These programs may fail to cover the relevant exposure window, and do not adequately capture the variability in exposure levels. For cancer studies where long term exposures are estimated, there may be significant inaccuracies due to variations in sampling and assay methods between different utilities and different time periods. Modelling of DBP concentrations in distribution systems is very complex, and site specific models must be used which incorporate relevant local parameters.

This session examined the exposure measures used to date in epidemiological studies of cancer and reproductive effects, and discussed possible options for improvements in accuracy. Early studies considered only drinking water exposure but it is now recognised that inhalation and dermal routes of exposure may be equally significant or more significant than ingestion for many DBPs.

Fourteen studies of DBPs and reproductive effects have been published, with a range of pregnancy outcomes examined by different researchers. Exposure has been assessed in terms of water source (surface vs groundwater) or disinfection type (none, chlorine, chlorine dioxide, chloramine), or DBP measurement (usually only total THMs). Only a few studies have examined water consumption or water use behaviours in individual women, and most have used retrospective assessment of exposures and pregnancy outcomes.

Reproductive study designs are limited by the frequency of the outcome under study - for example stillbirths occur at a rate of 5 in 1,000 births, therefore it is not feasible to conduct a prospective study as this would require enrolling many thousands of pregnant women to achieve adequate statistical power. Only relatively more common outcomes such as spontaneous abortion, premature birth or low birth weight are feasible to study in an entirely prospective manner.

Exposure assessment in cancer studies requires long term measures, which should include a full residential and occupational history, but it is uncertain whether questionnaires about water consumption or showering behaviour 30 to 40 years in the past produce accurate answers. Given present knowledge about variation in DBP profiles in different water sources, exposure classifications in past studies should be reexamined. For example, reclassification of exposure on the basis of brominated THMs rather than total THMs may yield different risk estimates.

A number of potential exposure biomarkers for DBPs are under investigation. Attention has focused on detection of THMs and haloacetic acids in blood, urine or exhaled breath as these are the most abundant classes in drinking water, and also the most widely monitored by water utilities. The relative importance of the different exposure routes (ingestion, dermal, inhalation) depends on the chemical nature of the DBP (volatile /non-volatile, lipophilic /non-lipophilic).

Some substances formed during drinking water disinfection may also occur in foods or beverages, and some are produced as normal body metabolites. Such compounds are unsuitable as biomarkers as they are not specific for drinking water exposure. Compounds which are primarily absorbed by ingestion may be difficult to detect in biological fluids if they are rapidly metabolised on the first pass through the liver. It is feasible that lipophilic DBPs and their metabolites may be stored in adipose tissue, but it is difficult to verify this experimentally in human subjects. During discussions it was also noted that DBPs which are grouped together in chemical classes (eg trichloroacetic acids) may have very heterogenous biological properties in terms of absorption, metabolism and excretion.

Biomarkers may eventually offer a means to accurately estimate exposure and reduce misclassification in epidemiological studies, however at present considerably more work is required to characterise and validate these markers under well controlled conditions before they can be routinely used. New developments in analytical techniques are likely to improve the speed and sensitivity of DBP assays both in water and in biological fluids.

This session examined sources of error in exposure assessment, their potential effect on risk estimates, and modelling approaches to the problem. In epidemiological studies, non-differential exposure misclassification is generally assumed to reduce risk estimates towards the null value (ie assumed to underestimate the degree of risk), however reanalysis of published studies on reproductive outcomes has demonstrated that under some circumstances exposure misclassification can produce a spuriously high risk estimate.

A number of research groups are developing models of in-home exposure to DBPs using information such as room sizes, ventilation, hot water temperature, showering times, water flow rates etc. Such models are designed to simulate the microenvironment inside the home in order to predict DBP exposures by various routes. Exposure estimates from these models are being compared to reported behaviour as recorded in diaries, blood samples from participants, and data logging of water use in individual homes. Sensitivity analysis will indicate which are most important parameters in the models.

An overview of epidemiological issues emphasised the need to recognise real world limitations. While the aim must be to develop more accurate exposure assessments, we must be careful not to confuse detail with accuracy. This is particularly relevant for cancer studies, where the long retrospective time frame imposes limitations on data collection. For example, a detailed water consumption /water use questionnaire covering 30 years may not necessarily produce accurate answers due to unreliable recall of the people involved. Perhaps the effort spent in such assessments is not justified by the resultant improvement in accuracy over "ecological" measures such as residential history - if an improvement is gained at all. The difficulty in separating DBP exposure from other potential risks is also important. Long term exposure to chlorinated water supplies correlates strongly with residence in urban areas, which in turn entails exposure to many urban pollutants which may impact on cancer risks.

Studies of reproductive effects have the opportunity to collect much more accurate exposure data, but again there is a need for some caution. It has been demonstrated that people show a distinct "digit preference" when recording showering times - if the showering time exceeds 5 minutes there is a tendency to round up times in 5 minute increments. Thus reported exposure times will tend to cluster in a way that does not reflect actual behaviour. When considering “exposure windows” for reproductive effects we perhaps need to consider the impact of differential survival of foetuses on recorded pregnancy outcomes as well as the time period when a particular type of defect may have arisen.

The priority for improvements in epidemiological studies must be to identify measures which can be reliably reported or assayed, and which enable us to distinguish between people with different exposure levels.

Routine sampling programs conducted by water utilities for regulatory purposes are unsatisfactory for epidemiological studies. Future studies should incorporate water sampling programs to better assess DBP profiles in different areas of the distribution system. Epidemiological researchers should involve water utilities in the design of such programs to ensure an optimal sampling strategy within the financial and other constraints of the study. Industry bodies (AWWA, CWA) could assist by developing guidelines to facilitate contact between researchers and utilities.

There is insufficient knowledge to identify a likely causative agent or agents for adverse health effects among the wide range of known and unknown DBPs, therefore a range of water quality parameters should be tested in addition to a select range of DBPs. These measurements will allow the estimation of levels of any particular DBP that is identified as being of concern in the future. Participants with experience in water analysis strongly recommended that the following parameters be measured:

It would also be preferable to utilise standard assay techniques for water quality parameters to ensure comparability of results. It was noted that commonly used kits for measuring free and total chlorine were not very accurate.

Current hydraulic models were developed for operational purposes (eg managing disinfectant residuals, maintaining pressure), and are not good for modelling DBP formation and decay. Better models with more accurate information on water age and travel times are needed.

Brominated compounds - there was considerable discussion during the workshop of emerging evidence that brominated DBPs may be of more concern on the basis of animal toxicology and human reproductive studies than their non-brominated counterparts. However concerms were also expressed that we should not focus too heavily on these compounds to the extent that we neglect others.

There is still insufficient evidence to ascribe a "causative" role to any DBP or class of DBPs, and indeed the existence of adverse health effects from chlorination DBPs is still unproven. We must exercise care in communications with the public on this issue. There is already an erroneous perception that other means of water disinfection (eg ozonation) produce no DBPs. The risks from bottled water are also unknown.

As more data accumulates on DBP occurrence it may be possible to define a subset of compounds that will provide adequate surrogates for the presence of others. Consideration also needs to be given to the biological characteristics of DBP when choosing what to measure – similarity in chemical structures does not necessarily imply similarity in metabolic properties.

Exposure assessment - we need to resolve how much detail and what degree of accuracy is needed for exposure studies. It is wasted effort to ask detailed questions about water use behaviours (eg drinking, showering etc) unless:

Data is lacking on home microenvironmental effects - for example overnight storage of water in the domestic hot water tank will significantly alter the DBP profile compared to the cold tap water supply. This information is needed for incorporation into models. Research to date has focused mainly on modelling THMs - work is needed on other DBPs also.

Biological samples - it would be desirable to develop a concensus protocol for collection and storage of biological samples (eg blood, buccal epithelial cells, bladder epithelial cells) for epidemiological studies in order to maximise the potential for future DBP or biomarker analysis. Some research has been done on potential gene-environment interactions and on potential pathways of effect for DBPs. Future work would be aided by the collection of biological samples.

There is a need to design studies to maximise differences in exposure - when choosing locations for health studies it may bepreferable to first examine the range of DBP exposures in several locations then site the study to maximise contrasts. Other factors that might be considered are the quality of existing distribution system models and the extent of DBP monitoring. However when comparing different locations, influences such as socio-economic differences may affect the health outcomes being measured.

Natural experiments - consideration should be given to designing "before and after" studies in locations where a significant shift in DBP exposure is occurring due to changes in water treatment practices.

Cancer studies vs reproductive studies - different exposure assessment tools are needed for the two types of studies, and these two issues must be independently assessed. Evidence supporting one type of adverse effect cannot be inferred to support the other.

Information is needed on DBPs from other exposure sources (for example food, beverages, occupational exposures), and those which are natural body metabolites. We need to identify which DBPs occur solely or primarily in drinking water in order to concentrate our research focus.

More work is needed to characterise large molecular weight DBPs. During discussions, it was noted that some toxicologists believe that high MW compounds are unlikely to be of concern in terms of health effects as they may not be readily absorbed or metabolised. However there is no concensus on what molecular weight cutoff might be used to define such substances, with estimates ranging from 800 to 2,000 or perhaps 5,000 Daltons. Additionally, substances which are present at levels of 10 microgram/L or less should not represent a risk unless they are markedly more toxic than any known substance.

The debate was organised by Victorian Branch of the Australian Water Association in Melbourne on 9th May 2000. The meeting was chaired by Dr John Langford of the Water Services Association of Australia, and consisted of six presentations followed by a panel discussion and debate on the value of total coliforms as a water quality indicator. The proceedings began with Dr Melita Stevens, Principal Microbiologist for Melbourne Water Corporation who provided a brief history and background to coliforms for the benefit of the non-microbiologists.

The next speaker was Dr Michael Taylor, a senior adviser with the New Zealand Ministry of Health who gave an overview of the New Zealand approach to drinking water practice. E.coli is used as an indicator of faecal contamination in NZ, while total coliforms and heterotrophic plate counts are used to assess the effectiveness disinfection. Turbidity is viewed as the best indicator of protozoal risks. Dr Taylor acknowledged that there is no direct relationship between indicator organisms and disease organisms, and no technology for direct monitoring of disease organisms at present. In his view E.coli is the best solution at the moment.

Mr Jeff Wright, CEO of the Sydney Catchment Authority questioned the importance of the total coliform test and expressed the view that it was time for a change. In many Australian water supplies, particularly in rural communities, total coliforms occur more frequently than specified in the relevant guidelines (WHO 1984 or NHMRC 1987). Reduction of coliforms could be achieved by raising chlorination levels, but rural communities often dislike high chlorination levels, and excess DBP formation may be a concern even though Australian guidelines for THMs are not as strict as European or US regulations. The use of alternate disinfection techniques such as UV or ozone in order to reduce total coliforms to near zero would require significant increases in capital, operating and mains cleaning costs especially for small communities.

Mr Martyn Kirk from the Disease Control section of the Victorian Department of Human Services noted that Victoria is the only Australian state with legal Regulations regarding water quality. These specify sampling, reporting and notification of diseases, but do not prescribe standards, although compliance with guidelines may be required in licensing agreements. The DHS require monitoring of total coliforms, and if an alert level of 50 cfu/ 100 ml is reached then re-sampling must occur and if this is confirmed the authority must take action. This action involves: identifying the organisms, speciation if possible, investigating the operations and rectifying any problems, then resampling and documentation for future reference. The DHS will continue to monitor total coliforms, particularly since it is easy and can be done at the same time as E. coli and this measure has international support.

Dr David Cunliffe of the South Australian Department of Human Services and the NHMRC Coordinating Group for Revision of the Australian Drinking Water Guidelines (ADWG) mentioned that the drinking water guidelines have both strengths and weaknesses and they require both the supplier and regulator to think, interpret and apply. South Australia has a reporting structure for the detection of total coliforms. If for example, a system shows total coliforms in absence of E coli then additional sampling is required which includes Enterococci. The results are interpreted as follows:

Modifications to Fact Sheets in the ADWG are being proposed to incorporate the new defined substrate technology for B-galactosidase detection (eg the Colilert test), and to recognise that total coliforms may not signify faecal contamination in the absence of faecal /thermotolerant species.

The final speaker was Associate Professor Nick Ashbolt, from the University of NSW and a member of the WHO Guideline Expert Group. He advocated a whole system approach to water quality management using a number of parameters to monitor for source water contamination, efficiency of treatment and distribution system integrity. While total coliforms have a role to play, they neither indicate faecal contamination or pathogen presence, and the time lag to obtain a result makes them an unsuitable trigger for rapid operational response. Physico-chemical parameters should form the basis for real-time monitoring of critical control points in system operation. For assessing health risks the focus should be on E. coli or the thermotolerant species.

The panel discussion and debate followed. The panel consisted of the invited speakers plus Prof Nancy Millis, Ms Jo O’Toole and Dr Daniel Deere. The chairman posed a number of questions to the panel which were representative of the major issues associated with the use of total coliforms as indicators for Australian Water supplies.

This conference is sponsored by the CRC for Water Quality and Treatment, the Water Services Association of Australia, and the Australian Society for Microbiology. Further details will be published in Health Stream at a later date. The conference organisers may be contacted at:

The US Court of Appeals for the District of Columbia has ruled against the EPA in a case concerning the MCLG (minimum contamination level goal) for chloroform. The case was brought by the Chlorine Chemistry Council on the basis that the EPA had violated the Safe Drinking Water Act by failing to use the "best available peer-reviewed scientific evidence" in setting the MCLG for chloroform.

In 1998 the EPA had proposed an MCLG of 0.3 mg /L for chloroform on the basis that the scientific evidence supported a safe threshold level of exposure for this compound. However the MCLG was later revised to zero in the face of public and political criticism. The MCLG is not an enforceable standard but rather provides a target for water utilities to aim for. MCLGs are defined under the US Safe Drinking Water Act as levels "at which no known or anticipated adverse health effects occur, allowing for an adequate margin of safety". The regulatory level or MCL (maximum contaminant level) is set equal to or above the MCLG taking into account practical considerations of cost and available technology.

In a decison issued on March 31 the Court found the adoption of the zero MCLG for chloroform by the EPA to be "arbitrary and capricious and in excess of statutory authority" and therefore vacated the rule.

The Australian Bureau of Statistics has released a report detailing water supply, use and consumption by various sectors of the Australian economy from 1994 to 1997. Despite being the driest inhabited continent, Australia ranks second in the world in terms of water use per capita with 1 million litres of fresh water used per person. Agriculture, including forestry and fishing, accounted for over 70% of consumption, while household water use was responsible for 8%. About 52% of water is supplied via mains infrastructure while the rest is extracted directly from the environment. The report can be obtained from the ABS (Catalog No. 4610.0).

The UK Environment Agency is assessing the need for new regulations covering pharmaceutical compounds in sewage effluent, following an expert review of the issue. Evidence of drug residues and their metabolites in effluent and drinking water sources has been presented at recent conferences in Brighton and San Francisco, and there are concerns over the impact of such chemicals on aquatic ecosystems. Some scientists have speculated that residues from pharmaceuticals and personal care products may be responsible for mass die-offs of aquatic organisms, or interference with breeding cycles.

An outbreak of Legionnaire's disease in Melbourne Australia has affected at least 100 people and caused two confirmed deaths, with a third fatality also under investigation. The source of the outbreak appears to have been the cooling system at a recently opened aquarium which attracted up to 4,500 local and international visitors each day. All cases had visited the area between April 11 and 25. Some infections occurred in people who had been near but not inside the building, suggesting that external drifts of water droplets from cooling towers may have been responsible.

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.

Circulation Report – Issue 18 June 2000 Circulation for this issue of Health Stream is 3393 copies, with readers in 50 countries.
Australia Argentina Austria Belgium Brazil Cameroon Canada Chile China Czech Republic Denmark Finland France	2844 1 4 2 2 5 23 1 6 1 1 2 26	Germany Greece Hong Kong India Indonesia Iran Israel Ivory Coast Japan Jordan Korea Lesotho Luxembourg	29 4 23 11 7 1 9 1 78 2 1 1 1	Malaysia Mexico Morocco Myanmar Netherlands New Caledonia New Zealand Norway Palestine Papua New Guinea Philippines Singapore	50 1 3 1 9 2 24 4 2 4 16 4	Slovak Republic South Africa Sri Lanka Sweden Switzerland Taiwan Thailand Togo UK USA Vietnam Yugoslavia	3 8 2 2 4 18 7 1 53 86 2 1

Editor - Martha Sinclair	email martha.sinclair@med.monash.edu.au
Assistant Editor - Pam Lightbody	email pam.lightbody@med.monash.edu.au