risk starting with the time the person first worked in one of the at-risk departments of the plant and ending at death or in 1982. Death rates of U.S white males and Ohio state mortality rates served as comparisons. The level of exposure in this cohort was relatively low: 94 percent of reported values were below the limits of detection, 5.1 percent were between 10 and 50 μg/L, and 0.6 percent were greater than 50 μg/L. Only 50 percent worked at the plant more than 5 years, and the turnover of employees was 15–25 percent per 5 years.
A study by Dupree and colleagues (1987) examined mortality among workers at the Linde Air Products Company Ceramics Plant in Buffalo, New York. The plant converted uranium ore to uranium tetrafluoride from 1943 to 1949. The intermediate products in the conversion process were insoluble uranium oxides. Workers also had exposure to other toxic chemicals, including sulfuric acid. Uranium ores with high content of radium-226, an external gamma radiation hazard, were present in the plant for 18 months. Workers were allowed to work for only 2 hours a day with these ores.
The study cohort (n = 995) consisted of white males who had worked at the plant for at least 30 days. The authors created an employment roster from company records and cross-checked it against IRS records (93–95 percent concordance). Worker’s individual job histories were reconstructed from security records, medical examination records, and earnings records. The authors tracked the vital status of the members of the cohort through the SSA and other sources. They obtained death certificates from the Department of Energy’s Death Certificate Retrieval Office and coded the cause of death. Vital status was ascertained for 94.3 percent of the workers, and death certificates were obtained for 94.6 percent of the deceased. The authors compared the death rates of the cohort with death rates for all U.S. white males and for white males in Erie and Niagara counties of New York, the counties in which the workers lived. The period during which a worker was at risk of death began at the date of hire and ended at death or on December 31, 1979, the close of the study.
The authors estimated the dose ranges for each job in the plant using information on airborne radon and uranium monitoring, surface contamination, and urine uranium levels and created a model that assumed a distribution of inhaled particle size. The jobs were grouped into three categories with estimated annual lung doses of <10 mSv, 10–100 mSv, and 100–1,000 mSv. For reference, the occupational limit for lung dose is 150 mSv/year. The authors classified jobs by external radiation dose, using film badge records. The estimated external dose during the 18 months in which radium-226 was present (<20 mSv/year) was considerably less than the allowable whole-body dose limit (50 mSv/year). Therefore, they classified workers only according to internal uranium exposure.
The cohort was relatively young, with 64 percent aged 16–35 years at the date of hire. The vast majority began work from 1943 to 1945, and 56 percent