A method for evaluating the mean exposure from a lognormal distribution

Abstract

OSHA's risk assessments, made in support of new standards, implicitly assume that workers are exposed at the level of the PEL on the average over their working lifetimes. This approach seems to be justified at least in cases where the toxicant is eliminated slowly from the critical site in the body. Thus, there are situations where the industrial hygienist should seek to evaluate worker mean exposure overtime. A simple testing procedure is proposed for determining whether the mean exposure of a lognormally-distributed exposure series, mu c, is less than the Permissible Exposure Limit (PEL) at desired levels of significance and power. With the assumption that the exposures are described adequately by a longnormal distribution, the method allows realistic exposure scenarios where mu c less than PEL to be declared acceptable with between two and fifty 8-hr TWA measurements, depending upon the value of muc and the level of variability. Application of the methods to three industrial data sets indicates that the approach is viable.