The elimination of the existing pattern of large and widespread racial and ethnic disparities in health and health care is an official goal of American health policy, but the causes of these disparities are only partly understood. This project examines a potentially important source of disparities that has received little attention--racial and ethnic differences in the source of medical care. Focusing on residents of New York City and surrounding counties in New York and New Jersey, the study will use the HCUP state inpatient data base to compare hospital usage of whites, African Americans, Hispanics, and Asian Americans. The primary focus will be on comparing use of high volume hospitals for conditions and procedures for which previous research has demonstrated a positive relationship between volume and outcomes. Our preliminary analyses of the use of New York City hospitals by residents of New York City found that whites were about twice as likely as African-Americans to use a high volume hospital for carotid endarterectomy (64% vs. 32%) and for cancer surgery (42% vs. 23%), with rates for the other two ethnic groups falling in between. The project is designed to explore the magnitude of racial/ethnic disparities in the use of high volume hospitals for the 22 conditions and procedures for which a positive volume-outcome relationship has been found, to identify factors that help account for these disparities, and to explore some dimensions of potential policy significance. Using measures of individual patients or of their neighborhood (based on patient's zip code), the analyses will explore the role of age, gender, type of insurance, and proximity to the hospital, including how patterns within New York City compare to those of its suburbs. Inclusion in the analysis of two comparison procedures with no known volume-outcome relationship will permit us to assess whether racial/ethnic disparities in the use of high-high volume hospitals is related to the existence of evidence that volume is associated with better outcomes. Separate analyses will examine the relationship of race/ethnicity and use of high volume hospitals within the Medicaid and Medicare populations. We will also give special attention to whether disparities are lower when cases have become concentrated in fewer hospitals as a result of public policies--certificate-of-need, which has concentrated cardiac surgery, and a quality-enhancing program that has attracted large numbers of patients to """"""""designated care centers"""""""" for AIDS. The study has both substantive and methodologic aims. Substantively, it opens inquiry into processes that may lead to differential distribution of racial/ethnic groups that are likely to have different quality outcomes. This type of sorting may be an important and largely unrecognized source of health disparities in the U.S. Methodologically, we will explore a number of measurement issues, including the effects of alternative ways of operationalizing high volume and proximity.
