Cancer mortality and exposure to crocidolite.

unsafe ground in relying heavily on a single, relatively short period, occupational hygiene survey of the Wittenoom crocidolite mine and mill' in formulating various models relating risk for asbestos diseases and airborne exposure levels. When he retired in 1984 Major gave me the long running thermal preci-pitator slides that formed the basis for the estimates of exposure for the few occupational groups he was able to study. I have reassessed some of these using modern light and electron microscopy and the results will be the subject of a future publication but it is clear that the estimates made 20 years ago underestimated the exposure considerably up to a single order of magnitude. Major informs me that the large number of non-fibrous compact particles on the LRTP slides degraded the performance of the (early) phase contrast microscope he used which, no doubt, reduced its resolution of the fine crocidolite fibres. At the time he had no established counting rules for guidance and is not surprised that his 27 fpml could become closer to 200 fpml when a microscope satisfying modern resolution criteria and currently agreed counting rules is used. On recounting, the highest light microscopy result indicated exposures of some 300 f/ml greater than 5 pm in length and in addition some 800 f/ml less than 5 pm in length. Results from the general area outside the plant produced up to 2 f/ml greater than 5 um in length. For comparison with contemporary fibre counts additional factors such as the collection efficiency of the LRTP relative to that of the modern membrane filter method and static/lper-sonal sample differences must be considered ; these also would increase the estimate of airborne fibres. De Klerk et al apply Major's published estimates of airborne fibre for a few occupants in one crocidolite mill to 87 job classifications in another mill (discontinued in 1957) believed to have been dustier. The authors acknowledge that this could be a considerable source of error but taken with the underestimates of dust conditions measured in 1966 and my experience in measuring other exposures in the occupational environment , it is my view that the quantitative data in their papers cannot be supported. Authors' reply: We would like to make the following comments on Rogers' letter. (1) The main emphasis in our dose response level estimates relies on the internal comparisons within the cohort and depends more on the ranking …