Article Text
Abstract
CAREX Europe provided insight in numbers of workers exposed to occupational carcinogens. The database contains estimates of numbers of workers occupationally exposed to carcinogens by industry in 15 European countries for 1990-93. Amendments were made for four countries joining the EU in 2004 covering exposure data from 1997. Even more recently (e.g. WOODEX, CAREX-Canada) prevalences of exposures were combined with exposure intensities.
Combining CAREX with existing exposure measurement databases would allow for adding a quantitative dimension to estimates of number of workers exposed to carcinogens. Several large exposure measurement databases exist (e.g. institution-based databases like MEGA, COLCHIC, NEDB or research databases like ExpoSYN).
Linking CAREX to these exposure databases is in principle feasible, but would require standardisation of coding of industries and occupations. It would also require adjustments of exposure distributions given for instance reason for sampling, measurement strategy applied (e.g. worst-case sampling), measurement devices used which might often lead to heavily biassed exposure distributions. Linkage will therefore not be straightforward nor easy.
For improving usefulness of CAREX for estimating the global burden of disease arising from occupational carcinogens it will be essential to combine number of workers exposed with their actual levels of exposure incorporating temporal trends and regional and economical differences on a global scale (upper-income economies versus upper-middle and lower-middle-income economies). Irrespective of this enormous challenge we should realise that only for a limited amount of well-studied agents (e.g. asbestos, crystalline silica, welding fume, wood dust) coverage across industries and time will be sufficient. But even for these agents extrapolation to developing countries will be needed. For other less well-studied agents expert judgement will remain essential. Exposures like those occurring among informal and illegal workers will not appear in official statistics and databases and will require other approaches than the standard CAREX and measurement database approach.