Norovirus outbreaks from drinking water.

Manunula, L., Miettinen, I.T. and von Bonsdorff, C.-H. (2005) Emerging Infectious Diseases, 11 (11); 1716-21.

As a part of an improved and intensified outbreak surveillance system commencing in 1997 in Finland , testing for viruses was included in the investigation of suspected waterborne outbreaks. The importance of viruses in waterborne outbreaks has only been recognised relatively recently, after techniques were developed to detect Noroviruses (formally called Norwalk-like viruses). These viruses commonly cause gastroenteritis in all age groups in the community, and they may be transmitted by multiple routes. Noroviruses cannot be grown in culture, and it was only after molecular detection techniques were established in the 1990s that it became possible to detect them in water supplies. This article describes an improved procedure to identify waterborne viral outbreaks.

There were a total of 41 waterborne outbreaks registered in Finland from 1998 to 2003. Of these outbreaks, 28 were investigated for viruses. Patients' stool samples and water samples were available for analysis in 24 of these outbreaks. For 3 of the outbreaks only water was available and in 1 outbreak only a patient sample was available.

Patients' stool samples were analysed by electron microscopy and reverse transcription-polymerase chain reaction (RT-PCR) for noroviruses and astroviruses and were screened by electron microscopy for other enteric viruses. If the test results were positive for a virus, a water sample was analysed. The virus concentration method used positively charged filters from 1L samples. Water samples were mostly only analysed for noroviruses.

Noroviruses caused 18 of the outbreaks, including 3 very large outbreaks involving 5500, 2500 and 2000 cases respectively. One outbreak was caused by rotavirus and no viruses were found for 9 outbreaks. Most of the norovirus contaminants occurred in groundwater systems and of these 8 occurred in public communal systems and 7 in private ground water wells. Half of the waterborne epidemics occurred in summer and norovirus outbreaks were the most common in late winter to spring. Summer outbreaks were mainly caused by bacterial pathogens. Inadequate disinfection of surface water, or use of undisinfected groundwater were considered to be responsible for most viral outbreaks.

Further characterisation of noroviruses from 16 outbreaks was undertaken by sequence analysis of amplicons from which the genotype was also deduced. Most of the outbreaks were found to be caused by a single norovirus strain/genotype. In 10 of the outbreaks norovirus was also found in the water samples, and in all but 1 of these outbreaks the same norovirus genotype was found in the water sample and in the patient samples. Coliforms were detected in water in only 9 of these 16 outbreaks, and the authors suggest that this may be explained by the longer persistence of infectious viruses compared to enteric bacteria.

The finding that noroviruses are frequent causes of waterborne outbreaks in Finland has given authorities an increased knowledge of the viral risks. As a result of this, laboratory techniques have been enhanced and the capacity for analysing environmental samples, especially water samples had increased. It is suggested that legislative measures for viral monitoring should be considered as part of a microbial risk assessment in drinking water production.