The conversion of dust components is of high importance for retrospective evaluations of exposure levels, of occupational diseases or the time trend of occupational dust exposure. For this purpose, possibilities to convert nickel concentrations from inhalable to respirable dust are discussed in this study. Therefore, 551 parallel measurements of nickel concentrations in inhalable and respirable dust fractions were extracted from the exposure database MEGA (maintained at the Institute for Occupational Safety and Health of the German Social Accident Insurance) and investigated by linear regression analysis of ln-transformed concentrations. Inhalable dust is the most important predictor variable, showing an adjusted coefficient of determination (adj. R2) of 0.767 (R2 adjusted to sample size). Since multilinear regression analysis, cannot be applied, further description of data is gained by splitting the whole dataset into working activity groups (e. g. 'high temperature processing', adj. R2=0.628,' filling/transport/storage' adj. R2=0.741, 'machining/abrasive techniques', adj. R2=0.777). From these groups, four task restrictive subgroups, so-called 'heuristic groups', can be derived by pooling similar working tasks with similar regression coefficients and enhanced quality measures (adj. R2 between 0.724 and 0.924): 'welding (grinding time fraction [GTF]<5%)', 'welding (grinding time fraction [GTF]>5%)', 'high temperature cutting' and 'grinding'. For the working activity group 'high temperature processing' and the heuristic group 'welding' the determination of the grinding time fraction and its inclusion or exclusion from a dataset has a huge impact on the description of data and whether a transformation of nickel concentrations using the natural logarithm (ln) is necessary or not. In case of GTF < 5%, the conversions functions are linear, all other conversion functions are power functions with exponents between 0.713 and 0.986. It is possible to develop conversion functions for estimating the nickel concentration in the respirable dust fraction (cR(Ni)) out of the nickel concentration in the inhalable dust fraction (cI(Ni)). For the estimation of Nickel in respirable dust other studies, it is recommend to use the conversion functions of the heuristic trial and error groups. Limitations of the possibility to use the conversion functions are discussed.
Read full abstract