Objectives: We assessed the extent to which injury rates among American Indians in Oregon are underestimated owing to misclassification of race in a surveillance system.
Methods: The Oregon Injury Registry, a population-based surveillance system, was linked with the Indian Health Service patient registration file from Oregon, and injury rates for American Indians were calculated before and after correcting for racial misclassification.
Results: In 1989 and 1990, 301 persons in the Oregon registry were coded as American Indian. An additional 89 injured persons who were coded as a race other than American Indian in the registry were listed as American Indian in the Indian Health Service records. The age-adjusted annual injury rate for health service-registered American Indians was 6.9/1000, 68% higher than the rate calculated before data linkage. American Indian ancestry, female sex, and residence in metropolitan counties were associated with a higher likelihood of concordant racial classification in both data sets.
Conclusion: Injury rates among American Indians in an Oregon surveillance system are substantially underestimated owing to racial misclassification. Linkage of disease registries and vital records with Indian Health Service records in other states may improve health-related data regarding American Indians.