Estimating Exposure Intensity Based on Odor

Area of Science:

• Environmental Science
• Occupational Health
• Toxicology

Background:

• Odor detection thresholds vary significantly within populations, often following a lognormal distribution.
• Estimating chemical exposure intensity retrospectively can be achieved by analyzing odor detection data.
• This method is applicable to single exposure periods with a single airborne contaminant.

Purpose of the Study:

• To develop a statistical method for estimating airborne chemical concentrations based on odor detection.
• To provide a framework for calculating confidence intervals for these exposure estimates.
• To assess the adequacy of odor as a warning property for airborne benzene.

Main Methods:

• Utilizing the proportion of individuals detecting an odor (k out of n) to estimate the chemical's concentration as the corresponding fractile of its odor detection threshold distribution.
• Calculating exact confidence bounds without normal distribution assumptions for small sample sizes or extreme proportions.
• Employing a parametric bootstrap procedure to incorporate uncertainty in odor detection threshold distribution parameters.

Main Results:

• The method was illustrated using benzene, with estimated population parameters: geometric mean of 37.8 ppm and geometric standard deviation of 1.92.
• In a hypothetical scenario with 2 out of 6 workers detecting benzene odor, the estimated exposure was 28.5 ppm, with a 95% lower confidence limit of 8.2 ppm.
• Benzene odor was found to be an inadequate warning for excessive airborne concentrations.

Conclusions:

• The developed method provides a quantitative approach to estimate chemical exposure from odor detection data.
• Statistical confidence intervals can be reliably calculated, accounting for sample size and parameter uncertainty.
• Odor detection is not a sufficient safety measure for airborne benzene, highlighting the need for objective monitoring.