Racial and ethnic identification practices in public health data systems in New England.

OBJECTIVE: Efforts are underway to standardize "racial" and "ethnic" identification in public health data systems under the Revised Minimum Standards for the Classification of Federal Data on Race and Ethnicity issued in 1997. This study analyzed the racial and ethnic constructs and labels used in public health data systems maintained by the six New England states in light of these standards. METHODS: The authors surveyed public health officials responsible for ongoing individual-level data systems and reviewed relevant documents. RESULTS: Information was obtained on 169 of 170 identified data systems. Ninety-one systems (54%) conformed to the federal standard in having separate "race" and "ethnicity" fields, yet many of these did not conform to the standard in other respects. Fifty-five systems had only a race field; of these, 20 included no identifiers corresponding to Hispanic and/or Latino ethnicity. Three systems used only an ethnicity field. The systems used various lists of racial and/or ethnic categories, and overlapping but not fully comparable labels. Few systems allowed for identification of ancestry groups not included in the revised federal guidelines but with large populations in New England, such as Brazilians. Some definitions and coding instructions seemed inconsistent with social and geographic reality. CONCLUSIONS: These public health data systems used inconsistent methods for classifying people by race and ethnicity. Standardization according to federal standards would improve comparability, but would limit options for defining and including some ethnic groups while forcing other groups to be aggregated in single race categories, perhaps inappropriately. Fundamental reconsideration of racial and ethnic categorization is called for.