A comparison of the direct and indirect methods of human exposure.

Surveys of human exposures to pollutants that incorporate biological measures of exposure and dosage are now being made to determine the distributions of exposures in the general population and in specific target groups. These surveys measure exposure directly by biological sampling and/or personal monitoring of subjects air, food and water intake and dermal exposure or indirectly by questionnaire evaluation of the subjects diet and activities as a function of time and location. Samples of breath, blood, feces and urine that give biological measures of past exposures can be evaluated and interpreted for severity of effect by the percentage of the population exposed above a guideline level (e.g. exposure leading to 2% of the population exceeding 20 micrograms/dl lead in blood). Because the complete population cannot be sampled with 100% accuracy, errors will exist in the percent of population with a given exposure and the exposure for a given percent of the population. This paper discusses the accuracy considerations of the direct and indirect methods of exposure assessment and recommends a procedure for designing the surveys.