region. State economic areas experiencing variations in cancer risk among minority populations are also highlighted. Data from this study may provide leads about etiology and cancer risk that may be pursued by further epidemiologic research. The study has revealed noteworthy patterns, such as a higher rate of prostate cancer among African-American men in the south Atlantic states, increasing rates of stomach cancer among American Indians in the Southwest, and "limited declines" in cervical cancer among African-American women in the Southeast.
In addition, DCEG staff, in collaboration with the National Institute for Occupational Safety and Health and NCHS, have jointly sponsored a study of occupation and industry codes on death certificates for purposes of understanding cancer prevalence by occupational risk. This study now encompasses 24 states and includes more than 5 million records. NCI reports that data from this Category I and II study are available for whites, African Americans, and all minority populations combined.
NCI is also planning several case-control studies of specific cancers and cohort studies of noncancerous conditions that are disproportionately prevalent among African-American men using data from the U.S. Department of Veterans Affairs (VA) inpatient and outpatient medical records. These data are available for more than 1 million African-American male veterans, as well as 4 million white male veterans, and can be used to examine the risk of various cancers associated with serious medical conditions and procedures.
NCI is also collaborating with the Health Care Financing Administration (HCFA) to link SEER program data with Medicare data to assess costs and the use of selected screening procedures, diagnostic procedures, and treatment patterns for older patients (ages 65 years and older). Health claims data will be examined by race, income, education, and related variables to understand how screening and treatment patterns may differ for subpopulations. In addition, the data will be analyzed by using census tract information to detect differences by the socioeconomic status of geographic areas for this Category I and II study.
In 1994, NCI established the Breast Cancer Surveillance Consortium to study mammography access and utilization among women in community-based settings and the impact of access and utilization on cancer outcomes. This Category II study is expected to yield data on mammography screening practices and mammography performance. It will also provide data on screening among ethnic minority women, as well as other factors that influence mammography access and utilization, such as education, income, and urban-rural location.
In addition, NCI has funded five grants specifically related to special populations to assess the utility of health claims data for cancer surveillance. These Category I and II projects will explore the completeness and