next up previous
Next: Comparison With Results Up: Dose-response Analyses For Previous: Effect of ''Missing

Results For Internal Radiation.

The factor IG was included as a crude indicator of the occupational hazard associated with internal radiation monitoring status in all dose-response analyses. The results for this factor are never impressive and no doubt mix selection factors (including facility selection) with any effect of heterogeneous exposures. It is likely that these estimates are affected by measurement error since the accuracy and completeness of internal exposure data varies considerably by facility and over time--see [2,3]. The EM coefficient for lung cancer is positive and may reflect the contribution of Y-12 workers to this category [35]. The coefficient for circulatory diseases is negative, suggesting that the positive EM coefficient for lung cancer is not merely a smoking-related phenomenon. The NE group has a large coefficient for external causes, which has been interpreted in other places as an SES effect. This is noteworthy because it is adjusted for paycode, facility and cohort.