Feasibility Issues Encountered in Drinking Water Studies

It has long been suspected that endemic waterborne illness not associated with outbreaks is an important but unrecognized health risk. To address this concern, we conducted two community intervention studies of waterborne disease, both using daily illness diaries collected before and after an intervention that dramatically improved the quality of treated drinking water. The first was conducted in two Texas cities near the United States-Mexico border. Because the groundwater source for Texas Site TA was found to be subject to periodic severe surface water contamination, membrane filtration of the groundwater was added. Texas Site TB is located approximately 60 miles from TA, uses conventionally filtered and disinfected surface water and is similar in size and demographics to TA. The second study included two distinct geographic areas (NWA and NWB) of a northwest city that used disinfected but unfiltered surface derived from two very well-protected watersheds. NWA had filtration and ozonation added to chlorination, while NWB retained chlorination as its only water treatment. As with the Texas study, reported illness rates for the two areas before and after the intervention were compared. These two community intervention studies illustrate the challenges in conducting studies in ethnically and socio-economically diverse populations. More importantly, they illustrate the difficulty in presenting findings that may challenge both conventional wisdom and predictions of mathematical microbial risk models.