-2The above calculations have ignored the possibility that some unknown proportion of person-years of observation are irrelevant because of the latent period between exposure and detectionof cancer. Subtraction of these person years would increase the difference between observed and expected values derived from non-exposed populations. (3). It should be borne in mind that the Marshallese have been more carefully observed for thyroid disease than the general populations from which expected values have been derived. (4) It is curious that the Utirik population evidences no increase in benign thyroid lesions relative to the unexposed population of Rongelap. |