and ethnic minority populations, data should be cross-tabulated to reveal information on "medically underserved" and "non-medically underserved" populations within each ethnic group. Improving data collection and obtaining the necessary information can be obtained by collecting data on several indicators of SES, along with the respondent's ethnic self-identification, native language, and information on the birthplace of the respondent's parents and their ethnic identities, where possible. This more comprehensive information is especially needed for clinical and epidemiologic work. Perhaps the simplest way to begin would be to focus on access to health services (both treatment and prevention) as a data element in surveillance.
One of the important pieces of information lacking in the current cancer surveillance system is the linkage of screening data or related information on cancer prevention activities to cancer incidence. Such data include access to care, which can be defined as the timely use of personal health care services to achieve the best possible health outcome. This definition relies on both the use of health care services and successful outcomes as a measure of access to care, a key aspect of cancer control among ethnic minorities and medically underserved individuals. This issue has recently been addressed by the Surveillance Implementation Group, formed by NCI in March 1998. The purpose of this group is to provide advice on proposed implementation plans, research directions, and priorities for expanding the NCI Surveillance Research Program and to recommend such plans, directions, and priorities. Defining the needed databases and surveillance systems is one of the many activities to be performed by this group in the near future.
An example of a group that links screening data to outcomes in the Breast Cancer Surveillance Consortium, which links national mammography screening data to an outcomes database (Ballard-Barbash et al., 1997). By standardizing data collection and linkage mechanisms for mammography and cancer registry data, this program will assess and improve the effectiveness of screening mammography. This important work should be continued and should serve as a model for additional programs.
As noted earlier in this chapter, information on survival rates of cancer patients can help in the identification of potential problems in either access to or quality of cancer screening and treatment services. The most