Occupational exposure assessment in the general population: improvements, innovations, and impact

Abstract

Studies performed in the general population are important for understanding the relationships between occupational exposures and human diseases. However, estimating occupational exposure in population-based studies is challenging due to various factors, including the large number of participants, variety of included occupations, and the lack of personalised exposure measurements.

Through one systematic review and five original research studies, this thesis explores how existing studies estimated exposure in population-based studies, where potential improvements may exist for current approaches, what innovations may be on the horizon, and how these advances may be harvested for higher scientific and societal impact. Two important results from the thesis are highlighted in this summary.

First, the thesis concludes that approaches used to estimate occupational exposures in population-based studies in the last 30 years have generally stayed the same. Fundamental innovations are hampered by the scarcity of exposure information in population-based studies, which are inferred from subject responses to questionnaires on their work history. However, a key area of potential advancement exists in using artificial intelligence and natural language processing to automate, standardise, and adaptively improve how we obtain and process exposure-related information from study subjects.

Second, by applying quantitatively estimated exposures in population-based studies, this thesis reports that work exposures to crystalline silica and diesel engine exhaust, even at relatively low levels, were associated with increased risks of lung cancer. The numbers of additional lung cancer deaths associated with different exposure scenarios were calculated to make the results directly applicable for informing risk assessment and policymaking.