Additional notes on the case-control study in western Washington on the cancer risk from asbestos in drinking water.

It is important to note that my presentation was a summary of our work and that a paper with a fuller description of procedures and detailed tables and risk estimates will be published in another journal (1). One questioner asked what factors determined "no exposure" and what proportions of the cases and controls had no exposure? For each place of residence or work, the following factors could have led to an absence ofexposure to Sultan River drinking water: (1) the place was located outside the boundaries of the Sultan River distribution system; (2) the place was located inside the boundaries of the Sultan River distribution system at a time when the Sultan River was not on line as a source ofwater for that area; (3) a subject worked at a job that, though based in the Sultan River area, involved extensive travel away from the place of employment; and (4) the subject did not use a community water supply, as water was obtained from a well, bottle or other sources. There was no estimated exposure to asbestos in drinking water in 1% of the cases and 0.6% of the controls. The asbestos concentration in the drinking water from the Sultan River was specifically measured. Twenty-two samples were taken from the distribution system from June 1978 to April 1979. There was quite a large seasonal variation in concentration. After correction for seasonal variation, we found that the concentrations varied only randomly across the Sultan River area. The average of all uncorrected concentrations was used as