Computational method for ranking task-specific exposures using multi-task time-weighted average samples

A method is presented for ranking task-specific exposures using time-weight ed average samples collected during the performance of multiple tasks. The task ranking can be used for purposes such as prioritizing further assessme nt or control, No a priori estimates of the individual task concentration d istributions are required, Sample concentrations and task-specific concentr ations are assumed to be log-normally distributed, and each known sample co ncentration is modeled as the geometric time-weighted average of the unknow n task concentrations. Since the task durations are usually not known, the task time-weights are estimated as a crude fraction of sampling period, Log transformed sample concentrations are aggregated based on the non-occurren ce of each task during some samples, resulting in a set of Linear equations which are solved to yield estimates of the log-transformed task median con centrations. The performance of the method was tested under a variety of co nditions using simulated sample data. The method was found to yield remarka bly reliable ranking of task median concentrations, especially for the high exposure tasks, provided that the number of samples was adequate and the t ask concentration distributions were not highly overlapped. The performance of the method can be easily modeled using simulated data over the range of plausible task concentration distributions for any number of samples and a ny job scenario, Even under the conservative assumption that some task conc entration distributions are highly overlapped, the assigned ranking can be usefully interpreted in the light of the modeling to determine whether a ta sk is relatively high exposure or low exposure. (C) 1999 British Occupation al Hygiene Society. Published by Elsevier Science Ltd. All rights reserved.