be to discontinue use of the old and divisive racial classifications and to recognize the scientific view that all humans are members of one race. The population would then be classified into ethnic groups. Under this arrangement, Hispanics and whites would not be divided. One may choose to be identified either as Hispanic or as white. Pending this change, NCI could simply omit the term ''racial" from its titles without doing any harm to the War on Cancer or acting in violation of OMB Directive No. 15. All Hispanics would then be classified as Hispanics or whites by self-identification. This would also allow NCI to avoid violating another important principle of classification: that the categories be mutually exclusive and totally exhaustive. It should not be possible for a person to be placed into two different groups simultaneously, and the system should accommodate the entire population.

This change would not be merely semantic, because it would avoid some of the ambiguity that currently decreases the reliability of calculations. Data for all ethnic groups could be reported in this way, but because of the increasing diversity of the U.S. population, it would still be necessary to aggregate small population groups into "macro-ethnic" groups. As noted earlier, "macro-ethnic" groups include, for example, Asian-American populations (whose subgroups include individuals of Chinese, Japanese, Korean, Indian, Southeast Asian, and other Asian descent). This becomes important because in the smallest population groups the number of cases of cancer will be small and the calculated rates will be unstable or unreliable. In addition, in reporting on ethnic minorities it is useful to present data on the rest of the population to make the classification complete for purposes of comparison.

The top margin of a summary table on cancer mortality could be as follows:

Ethnic Group A

Ethnic Group B

Ethnic Group C

Ethnic Group D

All Others

Total

Medically Underserved Populations

Data on the medically underserved population are not usually included in cancer statistics reports. The information could be obtained by revision of the source documents to include the necessary data or by special studies. Implementation of the required changes will take time, but the revision of source documents would have a more permanent effect. The source document for calculating incidence could be modified to include this information. Retrieval of the information from the death certificate may be more problematic. For this reason it may be better to obtain the



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement