Homogeneous Exposure Groups

Homogeneous Exposure Groups

The constitution of so-called Similar or Homogeneous Exposure Groups (S/HEGs),i.e. an estimate of the distribution of exposure over classification bands (as a percentage of the Occupational Exposure Limit), is a key factor in the exposure monitoring/surveillance of workers in the South African Mining Industry. However, analysis of data from the mining industry reveals that very few H/SEGs met the requirements for homogeneity, resulting in either under- or overestimation of exposure levels for individual workers.In international measurement strategies, e.g. the American Industrial Hygiene Association and the European Standardisation Committee, much attention is paid to the constitution of S/HEGs, where exposure determinants (or exposure models), historical data,professional judgement, and process characteristics are major information sources. The current practice in the South African mining industry is to constitute S/HEGs using a step wise process: first, the mine is subdivided into sampling areas; second, the sampling area is subdivided into activity areas, e.g. continuous coal mining.In industry, Bayesian Hierarchical Frameworks(BHFs) have been developed, which accommodate a formalised combination of the various types of information to informative priors which are used to determine the prior distribution of exposure.1,2 Currently, such BHFs have not been developed for exposure to dust in the mining industry. The envisioned constituents of informative priors would be ‘historical data’ and an‘exposure model’.Recently, a PhD project was designed to improve the accuracy of the constitution of S/HEGs with the development of a BHF for exposure to coal dust, using an in-depth analysis of ‘historical’ exposure data as one of the building blocks. Since an exposure model for coal dust exposure is lacking, and the determinants of exposure have not yet been identified, it is envisioned that the development of a Bayesian Belief Homogeneous Exposure Groups Network (BBN) will form the second building block. A BNN will accommodate a structured expert elicitation from geologists, mine and ventilation engineers,drill experts, occupational hygienists, etc., to identify determinants of exposure to coal dust in South Africa.The BBN consists of nodes connected by links, where the nodes represent the relation among different bits of knowledge; the conditional dependencies are quantified in a set of conditional probability tables.An envisioned PhD thesis will focus on the development of such a BNN to form the backbone of a future exposure model. We invite all interested candidates to submit applications to undertake a PhD in the Wits School of Public Health, as part of the aforementioned project on development of a BHF. A bursary is available from the Anglo American Endowment to Wits,which will cover registration fees and research seed funding. Please contact Prof. Derk Brouwer, e-mail:derk.brouwer@wits.ac.za, for further information.

REFERENCES

1. Hewett P, Logan P, Mulhausen J, Ramachandran G, Banerjee S. Rating exposure control using Bayesian decision analysis. J Occup Environ Hyg. 2006;3:568-581.

2. Sottas PE, Lavoué J, Bruzzi R, Vernez D, Charrière N, Droz PO. An empirical hierarchical Bayesian unificationof occupational exposure assessment methods.Stat Med. 2009;28:75-93.