AOL oe PHYS. MED. BIOL., 1976, VoL. 21, No. 6, 903-919. © 1976 The Observation and Analysis of Cancer Deaths among Classified Radiation Workers J. A. RETSSLAND, vip., aaanst.p., P. KAY, pup., a.r.c.s. and G. W. DOLPHIN, rim.p., p.sc. National Radiological Protection Board, Harwell, Oxon, ULI. Received 29 ALarch 1976, tn final form 5 July 1976 Ausrracr. The extent to which occupational radiation exposure contributes to cancer mortality is an influence on future world energy policy. It is also a factor in deciding the level of expenditure to reduce radiation levels expericneed by workers. Here we discuss some of the difficulties In analysing the situation and present the results of some calculations which estimate the expected age-specific radiation mortalities from all inducible cancers and also from leuknemian separately, Using » high vatue for the average occupational exposure and a conservative estimate of the associated risk, we find that a survey of molality among radiation workers must run over many years before sufficient dati would be accumulated to resolve the effects of radiation-induced neoplasins from those arising from other causes. We show tho advisability: of determining the cause of death both of persons who remain employedin the industry and all persons who enter and subsequently leavo tho industry, perhaps being employedin it for only a short time. Our estimates are based on maintenance of an occupationally exposed dose of one rad per person per year during the period of the survey which may extend over several decades. However, scaling of the estimates to any other exposure rates is casily performed. . We nlso give estimates of the lowest risk cocllicients delectable in a given observation time. Smee for a work force of 3000 these lowest detectable values ave an outer of magnitude larger than those expected, it is clear (hat only a national or international survey can produce dato adequate for even modest objectives. J. Introduction The widespread use of radioactive isotopes, X-ray equipment, neutron generators and the rate of expansion of the nuclear power industry has raised questions about the carcinogenic effects of radiation at the levels experienced by operational workers. We are not concerned in this paper with exposure resulting from accidents, neither are we concerned with the mechanisms of carcinogenesis. Our prime objectives here are to assess the prospects of drawing any conclusions from a survey of causes of death of radiation workers; to identify the factors which influence these prospects; to estimate the effects of latency of radiation cffects on the age-specific death rate; and to provide basic information from which the relative magnitude of vadiation-induced and natural cancers can be estimated. All of these factors arc relevant considera{ions in the sebling up of a survey. The ultimate value of a survey will be determined by the information contained within the collected data. Before establishing the data base it is prudent to attempt to anticipate the demands which will bo made on it. This paper reports such an atlempe6. 33 BEST COPY AVAILABLE