Use and analysis of exposure monitoring data in occupational epidemiology: An example of an epidemiological study in the Dutch Animal Food industry.
Heederik-D; Boleij-JS; Kromhout-H; Smid-T
Appl Occup Environ Hyg 1991 Jun; 6(6):458-464
A discussion was presented of partitioning of exposure variability for two types of monitoring strategy, and the use of one strategy in a study of Dutch animal feed workers was reported. One strategy involved monitoring all members of the study population at various times. The magnitude of underestimation of exposure response relationships was determined from the ratio of within worker and between worker variability. This strategy was most easily applicable to situations where personal dosimetry was technically simple. The second strategy involved grouping workers into homogeneous exposure categories for which only a sample of workers was monitored. This was favorable when monitoring methods were time consuming or expensive but was subject to inappropriate subgroup definition and variation in work practices and environmental conditions. For both methods, exposure response relationships could be underestimated due to within and between worker variances. The ratio of such variances determined the magnitude of bias. A study of the grouping method was done in eight Dutch animal feed production and storage facilities to assess division into homogeneous exposure groups, use of data to calculate cumulative exposure estimates and use of measurements to estimate exposure in 14 facilities. Eight occupational categories based on job title were used. A total of 530 measurements (some repeats) were made of personal inspirable dust and endotoxin levels. For both measures, occupational category rather than facility explained most variance. Between worker variance was somewhat larger overall than within worker variance. Crane drivers and other production workers categories demonstrated inhomogeneity due to task differences. Cumulative and current dust and endotoxin exposures correlated well with lung function measures. The authors recommend that analysis of variance be applied to repeated measurements to optimize measurement strategy or exposure grouping.
NIOSH-Publication; NIOSH-Contract; Contract-91-38661; Humans; Livestock-foods; Lung-function; Occupational-exposure; Organic-dusts; Respiratory-system-disorders; Epidemiology; Exposure-levels; Analytical-methods; Statistical-analysis
Applied Occupational and Environmental Hygiene