on African Americans); Native Americans; and Hispanics from Cuban, Puerto Rican, and similar ancestries" (National Cancer Institute, 1997a, p. 22–23). To date, this recommendation has not been fully implemented, although it is under study by NCI.
With regard to the current sampling configuration, it is apparent that the ethnic minority populations are not equally distributed across the SEER program regions (Miller et al., 1996). For example, most of the SEER program Hispanic population lives in California (69 percent), and Mexican Americans account for the majority of the Hispanic population in that area. Therefore, results from the SEER program may not reflect the life-style, environment, or cancer burden among Hispanic groups in other geographic areas such as the Northeast and Florida. Similarly, two-thirds of the Chinese population covered by the SEER program lives in California, as does 54 percent of the Korean population. Half of the SEER program's Japanese population lives in Hawaii, and 40 percent lives in areas of California. The SEER program's urban African-American population is more evenly distributed across the country, with 28 percent in Los Angeles, 25 percent in Detroit, 19 percent in Atlanta, 12 percent in San Francisco, and 8 percent in Connecticut. However, current representation of African Americans in rural areas is limited primarily to the 10 rural counties in Georgia. Numbering approximately 50,000, these African Americans are poorer and less well-educated than whites in the same counties. The majority of the American Indians in the SEER program registry live in New Mexico and Arizona, although the American Indian population resides primarily in other regions of the United States.
The overall percentage of people living below the poverty level and the number of high school graduates in the SEER program registry are similar to those among adults in the United States (13 percent and 78 percent, respectively). Ethnic minority groups are intentionally overrepresented in the SEER program to ensure adequate numbers for statistical purposes (i.e., to allow the enumeration of cancer rates). The populations covered by the SEER program areas, based on the 1990 Census, are shown in Table 2-5 (Miller et al., 1996).
However, even with this expanded coverage, the data for smaller populations are less precise than those for larger groups and must be viewed with caution. Another difficulty in interpreting information from the SEER program is the manner in which racial or ethnic group membership is determined. SEER program data on cancer cases are based on information