P096 Statistical modelling and development of a quantitative job exposure matrix for wood dust in the wood manufacturing industry. (1st September 2016)

• Record Type: Journal Article
• Authors: Basinas, Ioannis; Liukkonen, Tuula; Sigsgaard, Torben; Andersen, Nils T; Vestergaard, Jesper M; Galea, Karen; Wiggans, Ruth; Vincent, Raymond; Eduard, Wijnand; Kolstad, Henrik A; Vested, Anne; Kromhout, Hans; Schlünssen, Vivi
• Is Part Of: Occupational and environmental medicine. Volume 73(2016)Supplement 1
• Page Start: A152
• Page End: A153
• Publication Date: 2016-09-01
• Subjects: Medicine, Industrial -- Periodicals; Environmental health -- Periodicals; 616.980305
• DOI: 10.1136/oemed-2016-103951.417
• Languages: English
• ISSNs: 1351-0711

Abstract

Abstract : Aim: To develop a quantitative job exposure matrix for population-based cohorts for use in a large epidemiological study on Chronic Obstructive Pulmonary Disease (COPD) and asthma (http://codust.com ). Methods: An occupational hygiene database of wood dust measurements made during wood processing and handling has been established using measurement data from Denmark, Finland, France, Germany, The Netherlands, Norway, and the UK. Both personal and stationary measurements were included to elaborate empirical models of exposure to wood dust. Model development is based on Linear mixed effect regression with job codes (ISCO-88) treated as the random effects. Fixed effects are the year of measurement and factors that potentially confound the existing time trends. Results: Preliminary results from a model using data from 25, 450 measurements covering a time period from 1978 to 2007 suggest wood dust exposure declined by almost 6% per year in the European countries included. The model was established using country, year of measurement, sampling duration (<60 min versus >60 min), size fraction and type of measurement (personal versus stationary) as fixed effects. Perspectives: Currently the analytical strategy is further optimised. This includes applying hard correction factors to adjust for differences between sampling devices and integrating, for every job code in the matrix, an exposure prior (none, low, high) as determined by expert judgment to reduce uncertainty and allow extrapolation where the number of exposure measurements are low or non-existent. Final models and the resulting quantitative job-exposure matrix will be presented together with the results of a validation exercise partly based on comparisons with the results from measurements not included in the modelling process.