Health Stream Literature Summary - Issue 55 - September 2009

Rainfall and outbreaks of drinking water related disease and in England and Wales.
Nichols, G., Lane, C., Asgari, N., Verlander, N.Q. and Charlett, A. (2009) Journal of Water and Health, 7(1); 1-8.

Studies in the US and Canada have indicated a relationship between waterborne outbreaks and heavy rainfall in the preceding period. This current study examined outbreaks in England and Wales that were caused by public or private drinking water supplies to investigate the relationship between rainfall and outbreaks of drinking water related disease.

A case-crossover design was used with the 90 days before the outbreak representing “cases” and the same 90 day period in the previous five years representing “controls”. Cases were waterborne outbreaks reported during the years 1910-1999 where the location and time of the outbreak was known. Information came from Medline, Communicable Disease Reports, unpublished reports held by the Health Protection Agency Centre for Infections and published papers. Data collected for each outbreak included: geographical location, average incubation period of the pathogen, season, water supply (private water supply, mains), water source (surface, ground), rainfall implicated (yes, no) and whether the exact date of first symptoms in infected individuals was known or estimated.

Daily precipitation data were collected on-line from the British Atmospheric Data Centre. Each outbreak location was cross-referenced against a list of available rainfall stations for the United Kingdom. There were two approaches used to examine the relationship between outbreaks and rainfall. The first approach focused on total rainfall in a particular period and the second on excessive rainfall events. The first approach determined the cumulative rainfall in four time periods prior to each outbreak (1-7, 8-14, 15-21, 22-28 days prior to the outbreak) for the outbreak year and the self-control period by averaging over the five non-outbreak years. Rainfall was categorised into four groups (0 to 10 mm, greater than 10 to less than 20 mm, greater than 20 to less than 40 mm and greater than 40 mm). The second approach was used where an excessive rainfall event was defined as rainfall on a day preceding an outbreak exceeding the upper limit of the 95% reference range. The number of excessive events was determined for both the outbreak year and the control year prior to the outbreak.

Between 1910 and 1999 in England and Wales, there were 111 outbreaks identified to be associated with consumption of drinking water. There were 22 outbreaks excluded because the location was not known, the date was not known or there was no available rainfall data. Therefore the study included 89 outbreaks however the exact date of onset was not known for 18 of these outbreaks. There was a significant association found between rainfall in outbreak years as compared to the control years for all four one week periods before the outbreak date and a significant association between cumulative rainfall of over 40 mm in the previous 7 days (days 1-7) and outbreaks (p = 0.001). There was an excess of outbreaks compared to controls for all four one week periods in the greater than 40 mm rainfall group. There was a significant excess of low weekly rainfall (less than 20 mm per week) for the days 8 to 28 in the outbreak compared to the control years (p =0.002). When cumulative rainfall was greater than ten and less than or equal to 20 millimetres the risk of outbreaks was lowest and when rainfall was greater than forty millimetres the risk was highest, irrespective of the time period. The rainfall exceedance during the period before the start of outbreaks was calculated. The odds ratios of rainfall exceedance between matched outbreak and non-outbreak years were not statistically significant for any single day before the outbreak. However, the cumulative odds ratio was higher than 1 for all days up to day 30, nearing statistical significance for days 6 and 10. The main exceedance over control years was found to occur in the first ten days.

There were 30.3% of outbreaks that had less than 20 mm rainfall in the three weeks prior to the week before the outbreak compared to 10.1% in control years. Also, 14.6% of outbreaks had a period of rainfall in excess of 40 mm in the outbreak week compared to 2.2% of control weeks. These results imply that the attributable fraction of outbreaks associated with a sustained period of low rainfall is 20%, while the attributable fraction for periods of heavy rainfall is 10%. Low rainfall may contribute to source water contamination in a number of ways including a relative increase in the proportion of sewage effluent in rivers, opening of water channels allowing groundwater contamination, and increased run-off in subsequent rain events due to drying of soils. Therefore when drinking water providers are formulating Water Safety Plans and assessing the health impacts of climate change, interventions should focus on periods of low rainfall as well as risks following heavy rain.

© Copyright Water Quality Research Australia Limited http://www.wqra.com.au/
Health Stream articles may be reproduced and communicated to third parties provided WQRA is acknowledged as the source. Literature summaries are derived in part from copyright material by a range of publishers. Original sources should be consulted and acknowledged.