Skip to main content



The work performed in the frame of this Concerted Action was devoted to the development of methods to characterise occupational exposures, the identification of their potential and their limitations, the evaluation of their epidemiological performance, and their validation.
The concerted Action on retrospective evaluation of occupational exposures in cancer epidemiology addresses 3 main topics: job exposure matrices (JEM); general and adhoc exposure questionnaires; statistical analysis if data.
Project participants working on the use of JEMs have reached several methodological conclusions: it is necessary to consider the complete occupational history and not just the job held at a point in time; the total duration and level of exposure must be taken into account; purely dichotomic information is of limited value, the probability of exposure in a given job should be taken into account; latency periods are important; it may be useful to use a more precise classification of jobs than the 2-digit International Labour Office code; the use of JEMs in specific industries should be encouraged, as very precise analysis of job situations is possible in such contexts; to allow valid comparisons, it is important to use standardised codes for jobs, industrial activities, and exposures.
Teams are developing ad hoc questionnaires for some specific exposures and standardising coding procedures for expert evaluation of questionnaire data. They will compare the performances of JEMs and questionnaires on the same data sets and develop quality control methods for use in case control studies in the general population.
Participants are using statistical methods to improve the intrinsic precision of JEMs, by explicitly taking into account the probabilistic nature of each cell in the matrix.
Specific projects include the use of JEMs in specific industries, several community based epidemiological investigations involving a small number of specific nuisances of current interest in cancer epidemiology (organic solvents, asbestos and polycyclic aromatic hydrocarbons, formaldehyde), and record linkage surveillance systems.
The first part of the Concerted Action work concerned industry based studies (cohorts or case-control nested within a cohort). These are situations in which detailed and objective measurements of exposure are possible and a limited number of exposures are generally considered. A variety of occupational settings (agriculture, paper industry, construction industry, shoe-making, electroplating, welding, electricity production, railroad) were studied in a variety of European countries (Denmark, France, Germany, Italy, The Netherlands, United Kingdom).

The second part of the Concerted Action work concerned case-control studies in the general population. This is a situation in which it is not feasible to gather objective information on life long occupational exposures for all the subjects in a study or even a statistically representative subsample of them. Based on several case-control studies involving a variety of chronic diseases (lung and nasopharynx cancers, glomerulonephritis, leukaemia, lymphoma) and a variety of exposures (asbestos, formaldehyde, organic solvents, polycyclic aromatic hydrocarbons) in a variety of European countries (France, Germany, Italy), a number of key issues were again investigated and discussed.

The third part of the Concerted Action work concerned record linkages studies in the general population. This is a situation in which life long occupational histories are most generally not available. The exposure information available being generally limited to a combination of job title and/or industrial activity at one point in time for a cross sectional general population sample.


Institut National de la Santé et de la Recherche Médicale (INSERM)
16 Avenue Paul Vaillant Couturier
94807 Villejuif