Table 3.

Priorities for improving data inputs used by population models of cancer equity

Limitation of current data	Priorities for future data resources
Omission of data from people with limited use of health care and participation in research studies	Cancer risk factor information, patterns of care, and health outcomes among uninsured and underinsured groups and other disenfranchised populations living in geographically diverse areas
Lack of data on exposures relevant for underrecognized groups	Identify relevant exposures underrepresented in national surveys and other data sources; expand questionnaires of national surveys to capture relevant risk factors
Coarse or broad race and ethnicity categories that conceal important variation in health outcomes	Disaggregate race and ethnic groups, especially for Alaska Native, Asian American, Hispanic, Native American, and Pacific Islander persons; gather necessary information to characterize persons at the intersection of multiple identities including those who identify as multiracial
Poor data quality or unavailable data identifying disenfranchised populations in clinical databases	Include standard discrete data with robust data confidentiality protections for race and ethnicity, sex, and gender beyond male and female, disability, sexual orientation, immigration, incarceration, and language preference
Unknown eligibility for screening tests and method of cancer detection; unclear whether tests are for screening or diagnostic follow-up	Include risk factor data to determine eligibility for cancer screening tests (eg, pack-years of smoking), method of detection (symptoms, screening), and purpose of tests (screening or diagnostic follow-up) in cancer registries and medical records
Self-reported data that are susceptible to misclassification and sampling bias; lack of individual-level data on social determinants of health in clinical datasets	Facilitate geocode linkages between medical records and claims with area-based measures of social determinants of health; facilitate linkage between surveys and medical records or claims for individual-level measures of social determinants of health, for example, National Health Interview Survey, National Health and Nutrition Examination Survey, and Medical Expenditure Panel Survey linked with Medicare; strengthen health information technology policies and procedures that allow data linkages while preserving patient confidentiality
Absence of data on measures of systemic racism	Increase availability of data on factors that reflect the effects of systemic racism including income, education, health literacy, employment, health insurance coverage, medical debt, residential segregation and mortgage lending practices, neighborhood factors (resources, violence), environmental quality (air, water), voting participation, local media and advertising exposure, and individual experiences of discrimination