Validation of a new method for structured subjective assessment of past concentrations

A new method has been developed to assess the level of exposure from airborne hazardous substances. The method provides subjective exposure assessments using a structured approach, based on descriptive information about work activities and the work environment. Validation of the method has been carried out for 63 jobs involving five different agents: man-made mineral fibres, asbestos, styrene, toluene and mixed respirable dust. In three cases, there were two occupational hygienists who completed the assessments and in the remaining scenarios there was only one assessor. The subjective exposure assessments generally showed a statistically significant correlation with exposure measurements (correlation coefficient for the log-transformed data generally ranging from 0.31 to 0.93). The main exception to this pattern was the styrene data set where there was a small difference between the highest and lowest measured exposure level and this resulted in a much poorer association between the estimated and measured values. The assessments were also positively biased, with the ratio of the geometric mean estimated level to the measured level typically ranging from 1.3 to 2.2. Possible causes of the bias are discussed along with strategies to minimise its effect.