Determinants of exposure to 2-butoxyethanol from cleaning tasks: a quasi-experimental study

Abstract

Background: The quantitative assessment of airborne cleaning exposures requires numerous measurement methods, which are costly and difficult to apply in the workplace. Exposure determinants can be used to predict exposures but have yet to be investigated for cleaning activities. We identified determinants of exposure to 2-butoxyethanol (2-BE), a known respiratory irritant and suspected human carcinogen, commonly found in cleaning products. In addition, we investigated whether 2-BE exposures can be predicted from exposure determinants and total volatile organic compounds (TVOCs) measured with direct reading methods, which are easier to apply in field investigations.

Methods: Exposure determinants were studied in a quasi-experimental study design. Cleaning tasks were performed similarly as in the workplace, but potential factors that can impact exposures were controlled. Simultaneously for each task, we measured concentrations of (1) 2-BE according to the National Institute for Occupational Health and Safety 1430 method and (2) TVOC with photoionization detectors (PIDs). Simple and multiple linear regression analyses were performed to identify 2-BE exposure determinants and to develop exposure prediction models.

Results: Significant determinants from univariate analyses consisted of product type, tasks performed, room volume, and ventilation. The best-fit multivariable model was the one comprised of product type, tasks performed, 2-BE product concentration, room volume, and ventilation (R(2) = 77%). We found a strong correlation between the 2-BE and the TVOC concentrations recorded by the PID instruments. A multivariable model with TVOC explained a significant portion of the 2-BE concentrations (R(2) = 72%) when product type and room ventilation were included in the model.

Conclusions: Our results suggest that quantitative exposure assessment for an epidemiologic investigation of cleaning health effects may be feasible even without performing integrated sampling and analytic measurements.