Deriving safe short-term chemical exposure values (STEV) for drinking water

Abstract Small and brief exceedances of chemicals above their guideline values in drinking water are unlikely to cause an appreciable increased risk to human health. As a result, short-term exposure values (STEV) can be derived to help decide whether drinking water can still be supplied to consumers without adverse health risks. In this study, three approaches were applied to calculate and compare STEV for pesticides. The three approaches included basing a STEV on the acute reference dose (ARfD) (Approach 1), removing conventional attribution rates and uncertainty factors from current guideline values (Approach 2) and extrapolating 1 d and 7 d no observed adverse effect levels (NOAEL) from existing toxicity data using a log-linear regression (Approach 3). Despite being very different methods, the three approaches produced comparable STEV generally within an order of magnitude, which often overlapped with other existing short-term exposure values such as short-term no adverse response levels (SNARL) and health advisories (HA). The results show that adjusting the current guideline value using standard extrapolation factors (Approach 2) often produced the most conservative values. Approach 2 was then applied to two other chemical classes, disinfection by-products (DBPs) and cyanotoxins, demonstrating the wider applicability of the approach.