Cover-up over North West's crypto outbreaks

Cryptosporidiosis is a severe gastrointestinal illness which can be fatal to people with compromised immune systems. The protozoan parasite responsible, cryptosporidium, is commonly spread in water supplies and is resistant to chlorine disinfection. To protect public health, water companies must maintain efficient filtration of supplies at all times.

Details of an increased number of cases of the disease in the North West emerged in a Channel 4 Dispatches documentary in November. The programme drew a link between clusters of cases in Lancashire and Greater Manchester in June 1997 and water from the Thirlmere aqueduct - a 100 mile-long channel bringing water south from the Lake District. Dr Ed Wilkins of the North Manchester Hospital told Dispatches of his concern over the increase in reports of cryptosporidiosis in the area from about 100 two years ago to around 400 in 1997.

The clusters were investigated by an outbreak control team composed of staff from local health authorities and North West Water. The team's chairman, Professor Qutub Syed, regional epidemiologist at the Communicable Disease Surveillance Centre, refused ENDS' request for their report, claiming it was "not a public document" and that there was "no requirement to publish it or make it available to the public."

A copy obtained by ENDS elsewhere shows that some 370 cases of cryptosporidiosis were notified to health authorities in May and June 1997 compared with only 127 in the previous year. Epidemiological investigations found the cases to be localised in north-west Lancashire - mainly Preston and Blackpool - and parts of Greater Manchester and south Lancashire, including Salford, Trafford and Wigan.

The studies showed two waves of infection, one in north-west Lancashire in late May and a second in Greater Manchester in early June. Cases were found to span a wide age range which is consistent with a waterborne source of infection.

A case-control study of the whole area showed no significant association with the consumption of unboiled tap water. But a second study limited to Greater Manchester did find an association. There was a significant difference in the amount of raw tap water drunk by cases compared to the controls.

The distribution of water supplies in the area is highly complex, but water from the Thirlmere aqueduct could be supplied to all of the areas affected in quantities varying according to the local supply and demand situation. Thirlmere water is only treated by microstraining and chlorination - both ineffective against cryptosporidium - and is supplemented by untreated groundwaters from surrounding agricultural land.

The report also reveals that heavy rain in May was followed by high turbidity readings at the input to the aqueduct. Turbidity rose from below one unit to over two units, and supplies to Manchester at the far end of the aqueduct showed a similar deterioration in quality. High turbidity is often associated with episodes of cryptosporidium contamination and the timing of the peak is consistent with the supply causing widespread cases of the illness.

The report concludes that the cryptosporidiosis clusters identified "did not amount to an outbreak" and were only "possibly associated" with drinking water supplies - the weakest possible link to water under guidelines issued by the Public Health Laboratory Service (PHLS). The guidelines say that this conclusion is appropriate either where there is evidence of a water quality failure or where descriptive epidemiology suggests the outbreak is water-related and excludes other obvious causes.

However, on the basis of the evidence for Greater Manchester, the report has greatly underplayed the link with water. The combination of a deterioration in turbidity and the positive link to water in the case-control study is sufficient to justify a much more powerful conclusion - that the incident was "strongly associated" with the water supply under the PHLS definitions.

ENDS put these concerns to Professor Syed, who said that at the time the "possibly associated" verdict was reached the results of the Greater Manchester study were not available. He concurred that the Manchester incident should indeed be considered to be "strongly associated" with the supply.

Despite this strong link, North West Water never formally reported the cryptosporidiosis clusters to the Drinking Water Inspectorate (DWI) as an "incident" under the Water Undertakers (Information) Direction 1992. A company spokesman contended that "there was no water quality incident" and that the investigation had concluded "there was no outbreak" of the disease.

The company appears to have taken the team's reluctance to describe the clusters as an "outbreak" as a clearance of its role in the affair. But Professor Syed told ENDS that he attached no significance to the terminology used: "Cluster or outbreak, it is all the same. There is no real distinction," he said.

North West Water says that the Thirlmere supply is a "fairly pristine source", and regular monitoring "only occasionally showed a single oocyst" of cryptosporidium. The company does not consider the source to be a high risk for cryptosporidium, but admitted that its turbidity measurements were "noisy" - with many peaks of little apparent significance.

At the time of the outbreak, the supply was being monitored weekly, and this was increased to daily after the alarm was raised on 5 June. All the results proved negative, but monitoring for cryptosporidium is notoriously unreliable and the organism is most likely to have been present before the testing frequency was stepped up. Despite the company's failure to report the incident, the DWI was involved from an early stage and is now conducting its own investigation.

The North West appears to be something of a national blackspot for cryptosporidiosis. In addition to the figures Dr Wilkins gave to Dispatches, a 1994 study by the Centre for Research into Environment and Health noted that the Blackpool area had an above average incidence of the disease but was unable to find any link to water. North West Water has also reported four cryptosporidium incidents to the DWI since 1990 (ENDS Report 261, pp 31-32 ).

Despite the flaws in the outbreak control team's report, it concluded that any improvements to the water supply would be "welcomed". In particular, it recommends reducing the ingress of untreated water into aqueducts and increasing filtration - the only effective barrier against cryptosporidium.

North West Water uses microstrainers on at least two major water treatment works - Dunmail Raise, supplying the Thirlmere aqueduct, and Whitebull, supplying the Preston area. Following the recent cryptosporidiosis clusters, the company will be considering whether full filtration is needed at these plants. For Thirlmere, a more appropriate solution might be to use the aqueduct solely to supply untreated water - thus overcoming the problem of groundwater ingress.

A driver for improved treatment will be Government proposals announced in May to impose a new limit of one cryptosporidium oocyst in ten litres of drinking water (ENDS Report 280, p 39 ). It will be a criminal offence to supply water breaching the limit.

A spokesman for North West Water confirmed that measures to counter cryptosporidium were a "significant part" of its submissions to Ofwat on investment needs for 2000-05. Whether funding is provided will depend on the DWI and Ofwat's perceptions of the urgency of the company's proposals. However, new guidance from a Government expert committee has recommended a means of putting such judgements on a firmer footing.

The committee's third report on cryptosporidium, published in November, recommends a formal risk assessment procedure for all supplies. ¹ The report was commissioned following a surprising outbreak of the disease in 1997 which was linked to an apparently low-risk groundwater supply in Hertfordshire (ENDS Report 266, pp 6-7 ). Many of the group's recommendations concern groundwater sources which, it advises, may be affected by rapid ingress of surface waters.

The report also follows its predecessors in emphasising the importance of efficient filtration as a barrier to the organism, and regards turbidity monitoring as a "vital element" in ensuring the process is operating effectively. However, its new recommendations include advice that immunocompromised people should always be advised to boil tap water before consumption, and that cryptosporidiosis be made a reportable disease in England and Wales to achieve greater consistency in reporting.

The report will substantially influence the Government's proposals for new controls on cryptosporidium. Environment Minister Michael Meacher said in November that continuous monitoring for the organism will be required at plants where assessments show a "significant risk". This is likely to involve more trouble and expense for companies than the proposal made in May only to require monitoring at surface water supplies drawn directly from rivers.

The original proposals were costed at £7.9 million, but this is now likely to increase. There will also be a shift in costs towards companies with more high-risk water supplies, possibly including North West Water.