Use and Analysis of Exposure Monitoring Data in Occupational Epidemiology: An Example of an Epidemiological Study in the Dutch Animal Food Industry

Abstract

No general, theoretical framework exists for exposure monitoring strategies for epidemiological purposes nor for the analysis of these data. It is, however, recognized that the variation in exposure over time and between workers is an important element of consideration in the design of measurement strategies and the analysis of exposure data. In this paper, the partitioning of exposure variability is discussed for two types of monitoring strategies. First, a monitoring strategy is considered in which all members of the study population are monitored on various occasions. In this strategy, the ratio of within and between worker variance gives information about the magnitude of underestimation of the exposure-response relationship. The second strategy discussed is one based on “homogeneous” exposure categories. This strategy is illustrated with a study among animal feed workers in The Netherlands. It is shown that the grouping which results from this strategy can be applied for epidemiological purposes successfully if the ratio of the within and between category variance in exposure is relatively small. This implies that it is important to optimize the ratio of these variances in a given study population instead of composing strictly homogeneous exposure categories. In both strategies, the analysis of variance of repeated exposure measurements on the same individuals plays a crucial role. It is recommended that repeated measurements on the same individual should be taken routinely. The analysis of variance should be used as a tool to analyze such data in order to optimize a measurement strategy or a categorization of the population in exposure groups.