Abstract Details

Title: Modelling of Historical Exposure Level of Carcinogens Using Canadian Workplace Exposure Database
Authors: Chaojie Song; Victoria H Arrandale; Chris B McLeod; Cheryl E Peters; Hugh W Davies; Jérôme Lavoué; France Labrèche; Joanne Kim; Paul A Demers
Session: Large Scale Occupational Exposure Databases: Applications in Research and Prevention - St. Patrick, 18 Oct 2016, 10:30 AM - 11:45 AM

Background & Objectives

There is a general assumption that levels of exposure to many workplace carcinogens have decreased over time due to progress in hazard control (e.g. lower occupational exposure limits, improved engineering controls). However, the magnitude and pattern of change in exposure level is less well known. In order to explore empirically the time trends in exposure, we modelled the historical exposure levels of three prevalent occupational carcinogens using industrial hygiene measurements collected in various Canadian provinces from the 1980’s to 2000’s and stored in the Canadian Workplace Exposure Database (CWED).

Methods

Among the top ten most prevalent workplace carcinogens, we selected silica, wood dust, and formaldehyde for their relatively large numbers of observations in the CWED. For each carcinogen, a mixed-effect model was developed, with industry as a random effect term. The fixed effect terms were: year of measurement, sampling type (area or personal), sampling duration (minutes), location of measurements (province), and exposure level group (low, medium, high) assigned by the CAREX Canada job-exposure matrix. A fixed effect interaction term between sampling year and exposure level group was tested to examine whether trends differed by exposure level group.

Results

The number of available CWED measurements for modeling of silica, wood dust, and formaldehyde exposures was 5234, 7390, and 4759, respectively. For silica, the model indicated an overall decreasing trend in exposure level of -4.3% per year. For wood dust, interaction between year of measurement and exposure level group was statistically significant, with trends varying by exposure level group (high: -3.4%/year, medium: -1.0%/year, low: +3.7%/year). For formaldehyde, the model showed an overall increasing trend of +2.6% per year.

Conclusions

Our analysis showed varying time trend patterns among different workplace carcinogens during the period covered by the CWED. The assumption that occupational exposure to carcinogens would be decreasing over time was not fully supported by the analysis. Shifts in sampling objectives (anticipation, recognition, evaluation or control) or in sampling focus (higher-risk tasks or jobs) over time may explain part of the results. The model enables us to extrapolate levels of exposure for various workplace carcinogens, which can be used for exposure assessment in occupational epidemiological studies.