LAUSR

Objectives To estimate proportion of pathologies attributable to occupational exposure, lifetime occupational exposure prevalence (LOEP) and relative risk are necessary. LOEP estimates are commonly used but often estimated with different methods. The method choice and the impact on estimates are rarely discuss in the literature. This study present and discuss the most widely used means of estimating LOEP and their respective impacts on estimates. Methods A sample of individuals representative of the French population from 2007 was linked with four Matgéné job-exposure matrices: flour dust, cement dust, silica dust and benzene. LOEP and the 95% confidence interval were estimated using five methods: the maximum exposure probability during the career (Method 1), four using individual exposure probabilities, three of which subdivide careers into job-periods (Methods 2–4) and one which subdivides them into job-years (Method 5). To quantify differences between methods, percentage of variation were calculated for prevalence values on Methods 2 to 5 versus Method 1. Results For each agent, LOEP estimated from the maximum probability during the career (Method 1) was consistently lower than prevalence taking account of job-periods or job-years. LOEP on Method 1 for flour dust, cement dust, silica dust and benzene were respectively 4.4%–95% CI [4.0–4.7], 4.3% [3.9–4.6], 6.1% [5.7–6.5] and 3.9% [3.6–4.2]. Percentage of variation ranged from 0% to 25.0% for flour dust, from 11.6% to 55.8% for cement dust, from 11.5% to 49.1% for silica dust and from 0% to 53.8% for benzene. Conclusions The present study provides a description of several LOEP estimation methods in the general population based on job-exposure matrices. It specifies the strong and weak points of each of the five chosen methods. For health monitoring purposes, LOEP should be reported as intervals, with low and high estimates obtained on different methods using job-periods (Methods 2–4).