Differences, disparities, and biases: clarifying racial variations in health care use.

Studies documenting racial differences in health care use are common in the medical literature. However, observational studies of racial differences in health care use lack a framework for interpreting reports of variations in health care use, leading to various terms, ranging from "variations" to "bias," that suggest different causes, consequences, and, ultimately, remedies for such variations in treatment. We propose criteria to assess racial differences in health care use by using a clinical equity (equal treatment based on equal clinical need) framework. This framework differentiates between initial reports of racial differences and subsequent classifications of their findings as racial disparities or racial bias in health care use. Racial variations in health care use may be considered disparities after demonstrating that racial differences are not attributable to treatment eligibility, clinical contraindications, patient preferences, or confounding by other clinical factors and are associated with adverse consequences. Racial bias with adverse consequences in health care may be inferred if a racial variation in treatment that has been characterized as a disparity persists after accounting for health care system factors (for example, type of hospital at which the patient was treated). We apply this framework to published reports of racial differences in treatment to determine which studies provide evidence of differences, disparities, and bias. We discuss the use of such a framework in directing policy interventions for alleviating inappropriate racial variations in health care use.