Comparison of standardized mortality ratios (SMRs) obtained from use of reference populations based on a company-wide registry cohort to SMRs calculated against local and national rates

The DuPont Company has maintained a mortality registry for all active and pensioned U.S. employees since 1957. Standardized mortality ratios (SMRs) for each plant site in the U.S. can be calculated based on the comparison with the entire U.S. DuPont population or with a regional subset of DuPont employees. We compared the SMRs derived from a large, international cohort mortality study of chloroprene workers (IISRP study) with those derived from the entire DuPont Registry and appropriate subpopulations of the registry for two U.S. neoprene plants--Louisville (Kentucky) and Pontchartrain (Louisiana). SMRs from the IISRP study for the Louisville cohort based on national rates for all causes of death, all cancers, respiratory cancer, and liver cancer are higher than those based on local mortality rates. Both the national and local comparisons (several counties surrounding each plant) for all-cancer SMRs are lower than 1.0, the local comparison being statistically significantly reduced. In contrast, the SMRs based on the total U.S. DuPont worker mortality rates for all causes of death (1.13), all cancers (1.11), and respiratory cancers (1.37) are statistically significantly increased. The SMR for liver cancer (1.27), although elevated, is not statistically significant. SMRs based on DuPont Region 1 were closer to 1.0, and the SMR for all cancers was no longer significant. Stratification of the Louisville subcohort of males using the same cumulative exposure categories used in the IISRP study yielded SMRs calculated against DuPont Region 1 that were generally higher than those calculated against U.S. and local rates. Only the third exposure category showed SMRs statistically significantly above 1.0 for all cancers and for cancer of bronchus, trachea, and lung. However, there does not appear to be an exposure-response trend. The SMRs from the IISRP study for the Pontchartrain cohort based on national rates are higher than those based on local rates for all causes of death, but all are less than 1.0. The all-cause SMRs for both local and national comparisons are significantly reduced. There were no deaths from liver cancers observed in this cohort. Comparisons of the Pontchartrain cohort against the total U.S. DuPont worker mortality rates resulted in higher SMRs for all causes of death (0.98), all cancers (1.03), and respiratory cancer (1.08), but none were statistically significant. SMRs based on DuPont Region 2 showed very little change from those based on the total registry. The use of reference rates based on regional workers in the same large company produces SMRs lower than those based on the entire company population (regional socio-cultural effects) but higher than those based on geographically closer local general populations (healthy worker effect). The healthy worker effect is seen in cancer mortality rates as well as in other chronic diseases.