ACL OS * ’ In the population of Utirik (158 persons, all ages) followed for 21 years, the accumulated person-years of observation are estimated at 3140. In such a group, expection of thyroid cancer would be as follows: (1) Based on an average for all registries listed in Cancer Incidence on Five Continents! (2.5/100,000 per year): ‘ .08 cases, upper 95% confidence limit .66 (2), Based on incidence rates in the Marshal Islands, (dates), excluding cases in exposed individuals (2.0/100,000 per year): .06 cases, upper 95% confidence limit 54 (3) Based on estimates of the risk per rad thyroid dose per year , observed in Rongelap and Ailingnae combined (3.9/1,000,000/ rad/year, Tperson-yoars-at—risk-est-Si4O] thyroid dose 50 rads )= 157,000 person-~rad-years): -61 cases, upper 95% confidence limit 2.17 Conclusions: G) The occurrence of even 1 case of thyroid cancer in Utirik is unlikely to be due to chance at conventional Levels of statis- tical significance - if the rates do not exceed normal population rates. The occurrence of 2 cases by chance is extremely unlikely. (2) | The observed number of thyroid cancers is higher than would be expected on the basis of experience in the exposed populations cot ; ; of Rvand A, but does not exceed the upper 95% confidence limit of the expected value. Caveats (1) In computing expected valucs, no account has becn taken of age and sex differences between populations. The incorporation of adjustments for these factors is unlikely to change the above conclusions. 5 Usk g |