next up previous
Next: Facility Comparison Analyses Up: ORNL-6785_Appendix Previous: ORNL-6785_Appendix

APPENDIX- A Mortality Study Of Employees of The Nuclear Industry in Oak Ridge, Tennessee


A major goals of this report is to consider various methodological and technical issues related to combining data from multiple facilities to evaluate the potential adverse health effects of low-level occupational exposure to ionizing radiation. A detailed description of the data collection and validation procedures, as well as tabular and graphical summaries of the resulting data base are presented in by Watkins {et al.} in companion reports [13,12,14]. The analysis files (which are SAS data sets) are the starting point, see Figure A1 (below), for this report and are available through [2] CEDR. The demographic analysis ( NEED LINK TO VARIABLE LIST AND DESCRIPTION OF EACH VARIABLE, i.e. META DATA ) file contains an exhaustive roster of all persons hired at one or more of the Oak Ridge facilities before December 31, 1982. The facility comparison results presented in Tables II-V of this report were obtained using data in the demographic analysis file. Workers employed less than thirty days or whose data contain certain "critical errors" were excluded from the mortality analysis. The radiation dose-response results presented in Tables VI--XI were restricted to white males employed at either X-10 or Y-12 and use the yearly external and internal radiation exposure data contained in the three additional analysis files ( NEED LINK TO VARIABLE LIST AND DESCRIPTION OF EACH VARIABLE, i.e. META DATA) shown in Figure A1. In both the facility comparison and the dose-response analyses two or more steps are required to generate the Analytic Data Structure (ADS) which provides stratified tables for Poisson regression. This Appendix provides a detailed outline of the steps in the data analysis process.