Background: Health care workers are potentially exposed to a number of carcinogens. Studies among women in this field have focused on white nurses; however, workers in many health care occupations share exposures experienced by nurses.
Methods: Cancer mortality was examined among female health care workers using death certificate data collected in 24 U.S. states from 1984 through 1993. Cancer mortality odds ratios (MORs) were calculated by race (white, black) and age group.
Results: White nurses had a 30% elevation of mortality due to liver cancer and myeloid leukemia. White registered nurses (RNs) had a small excess and white licensed practical nurses (LPNs) had a small deficit of mortality due to breast cancer. Ovarian cancer was in excess among RNs, but decreased among LPNs. Among black nurses, excesses of death due to kidney cancer (MOR = 1.7) and multiple myeloma (MOR = 1.3), and a significant 50% deficit in mortality due to cancer of the esophagus were found. Black RNs, but not LPNs, had an excess of breast cancer (MOR = 1.3; 95% CI = 1.0-1.5). Ovarian cancer was elevated by 30% in both RNs and LPNs. Excess deaths due to cancers of the breast, ovary, and uterus occurred among white physicians. Among black physicians, lung cancer was significantly elevated (MOR = 2.8). White pharmacists had significant excesses of breast (MOR = 1.5) and ovarian (MOR = 2.4) cancers, and myeloid leukemia (MOR = 2.0). White clinical laboratory technicians had excess deaths from several cancers. The greatest excess was for myeloid leukemia (MOR = 2.3; 95% CI = 1.5-3.4). Excesses among radiologic technologists included cancers of the lung, pancreas, breast, uterus, and ovary.
Conclusion: Several findings reported here warrant further investigation. In particular, excesses of myeloid leukemia among nurses, pharmacists, and clinical laboratory technicians and liver cancer among nurses should be investigated in studies with data on occupational and other exposures. Patterns of mortality from breast and ovarian cancer found in this study must be evaluated further in studies with data on reproductive history.