Improving exposure estimates by combining exposure information

Abstract

Objectives: Any exposure estimation technique has inherent strengths and limitations. In an effort to improve exposure estimates, this study developed and evaluated the performance of several hybrid exposure estimates created by combining information from individual assessment techniques.

Methods: Construction workers (n = 68) each completed three full-shift noise measurements over 4 months. Three single exposure assessment techniques [trade mean (TM), task-based (TB), and subjective rating (SR)] were used to estimate exposures for each subject. Hybrid techniques were then developed which incorporated the TM, SR, and TB noise exposure estimates via arithmetic mean combination, linear regression combination, and modification of TM and TB estimates using SR information. Exposure estimates from the single and hybrid techniques were compared to subjects' measured exposures to evaluate accuracy.

Results: Hybrid estimates generally were more accurate than estimates from single techniques. The best-performing hybrid techniques combined TB and SR estimates and resulted in improvements in estimated exposures compared to single techniques. Hybrid estimates were not improved by the inclusion of TM information in this study.

Conclusions: Hybrid noise exposure estimates performed better than individual estimates, and in this study, combination of TB and SR estimates using linear regression performed best. The application of hybrid approaches in other contexts will depend upon the exposure of interest and the nature of the individual exposure estimates available.