Obstruction of valid race/ethnicity data acquisition by current data collection instruments

Methods of collecting race/ethnicity data affect the validity of conclusions based on them as do other factors such as lack of consensus and inadequate definitions for terminology; and misclassification or miscounting of patients. Current data collection instruments do not identify mixed heritage. We propose improving data measurement and collection by including a quantified multiracial/multiethnic heritage category for persons of mixed heritage; a don't know category; and a refuse to reveal category. The resulting problems affect all uses of racial and ethnic data, including health care provision, research, and health planning. More precise race/ethnicity categories should be encouraged by such means as a 16-category checklist (with instructions to check all which apply) currently in development. The Methods section of all clinical, epidemiologic, and pharmacologic reports should include descriptions of racial measurement and reasons for including or excluding clearly defined populations.