Full Description

In the May 2002 issue of Journal AWWA,Stewart and co-workers summarized an investigation thataddresses the potential health consequences related toconsuming water from a reservoir at which body-contactrecreation is permitted (Stewart et al, 2002). The investigationused a water quality model to estimate pathogenloading from recreationists and to calculate predictedpathogen concentrations at the outlet tower of the reservoir.Based on those data, risk to drinking water consumerswas predicted. The authors reported that the annualrisk of waterborne illness would increase three times abovebackground, despite conventional water treatment beforedistribution as potable water.Although the authors used a powerful combination ofmodeling techniques for assimilating a wide range of disparatedata, the substantial uncertainty associated withthe model parameters most directly affecting risk is notarticulated and/or analyzed in the findings of the investigation.The unspecified effect of uncertainty in this investigationcasts doubts on the conclusions. These uncertainty issues, furthermore, suggest that the use of the Stewart et al document as guidance for policymakers and stakeholders who are examining issues associated with body-contact recreation and drinking water should be reexamined. The authors raised the fact that high levels of uncertainty may affect the analysis. They report: "The modelwas highly sensitive to parameters associated withpathogen loading" (page 87). Pathogen loading to thereservoir was assumed to occur in two ways: sheddingof residual fecal material from recreationists' bodies uponcontact with the water (shedding); and, accidental fecalreleases (AFRs) and therefore was computed as a functionof the number of infected individuals recreating in thereservoir. The percentages of recreationists assumed to beinfected in each simulation varied by pathogen as follows:Cryptosporidium 0-5%, rotavirus 5-20%, Giardia0-10%, and poliovirus 5-20%, and the frequency of anAFR was 0-0.2%. This discussion focuses on parameterestimates for Cryptosporidium prevalence, rotavirusprevalence, and AFRs as well as the risk managementimplications of those ranges. Includes 13 references.

Product Details

Edition: Vol. 95 - No. 1 Published: 01/01/2003 Number of Pages: 3File Size: 1 file , 200 KB