*28 IV. Discussion
This study examines the progress achieved in reducing the racial divides in one of many possible indicators of health care access--the site of medical care. The study provides evidence that the vast majority of Americans, regardless of race/ethnicity, currently identify an office-based setting as a regular source of care. Moreover, only a small fraction of Americans rely on a hospital ER as a regular source of care. However, African Americans and Latinos, regardless of insurance status, continue to be far more reliant than whites on what some consider to be “non-mainstream” sources of care, with African Americans and Latinos being about twice as likely as whites to rely on a hospital-based provider as a regular source of care. The uninsured also were more likely than the insured to rely on a hospital-based provider as a regular source of care.
While the finding regarding the uninsured is consistent with other research, the continuing role of race/ethnicity as a factor associated with where an individual obtains health care was a less predictable finding. Studies in the 1980s had shown that minority Americans were more likely to use community or hospital-based clinics, but these studies left unanswered whether utilization patterns were a function of racial/ethnic differences in insurance coverage or income. This study provides strong evidence that race--independent of insurance coverage and income--continues to be associated with where ambulatory health care is obtained. The study findings counter the perception that whites, African Americans, and Latinos obtain health care from the same types of providers. While that fact is true for the vast majority of the population, there is a sizable subset of African Americans and Latinos who show a pattern of accessing the health care system that is different from the patterns observed in most Americans.
These findings are consistent with those of a recent study by Gaskin, which examines use patterns of inpatient hospital care. Analyzing 1994 hospital discharge data from nine states, Gaskin found that residents of racial and ethnic minority neighborhoods were more likely than the general population to use public hospitals and major teaching hospitals. Taken together, the findings provide evidence that racial/ethnic background continues to shape choices regarding the site of medical care. It also is conceivable that the findings may understate racial differences in the sites of medical care since respondents who identify community health clinics (private or public) as a regular source of care are defined as having an office-based provider.
*29 As previously noted, structural or institutional factors of varying settings of care may affect the content of care. These factors may explain some of the racial/ethnic differentials in care that have been observed. Research has shown that the organizational setting of care can affect the cost, quality, and patient satisfaction associated with care. Other factors, however, such as an individual's health and social needs, should also be considered in evaluating the content and appropriateness of care provided by a health care setting. A physician's office, for example, may be more conducive to a satisfying doctor-patient relationship but less convenient for some diagnostic tests. A hospital-based outpatient clinic might provide more technically sophisticated care than a physician's office but may have less potential for the development of a strong provider-patient relationship. Questions about differences in the quality of care in various settings, including various types of office-based settings, deserve to be systematically explored in future research and the findings included in the dialogue on possible factors contributing to racial/ethnic differences in health care.
This study raises a number of other issues for further investigation. Perhaps most important among these issues is the question of what factors explain the effect that race/ethnicity continues to have on where an individual obtains health care. Race/ethnicity might be a proxy for any number of factors such as the availability of private physicians in minority communities, patterns of residential segregation, or financial barriers such as co-payment requirements. It also might reflect preferences of patients for the flexible hours or other conveniences of hospital-based sources of care, a possibility consistent with the findings of a study that compared the characteristics of regular users of hospital OPDs and regular users of private physicians. The findings also might reflect historical patterns of utilization or choices made by patients because some sources of care may be perceived as more welcoming or culturally competent. These two factors may be linked since an individual may initially choose a source of care based on family tradition, but is unlikely to remain with that source of care solely for that reason. In sum, the finding could reflect barriers to care, patient preferences, or, of course, some combination of these factors.
The finding that race/ethnicity continues to exert strong influences on where individuals receive health care raises a multitude of questions. Further work is needed to explore the incentives and disincentives for obtaining care from different sites. It also will be important to assess whether there are systematic differences among the different sites in the content of care or the patient-provider relationship (e.g., communications *30 and trust), and whether these differences have implications for the health care outcomes of African Americans and Latinos.
Marsha Lillie-Blanton is a Vice President at the Kaiser Family Foundation and Director of Access to Care for Vulnerable Populations.
Rose Marie Martinez is the Director of the Institute of Medicine Division of Health Promotion and Disease Prevention.
Alina Salganicoff is a Vice President at the Kaiser Family Foundation and the Director of Women's Health Policy. The authors gratefully acknowledge funding support for this s project from the Henry J. Kaiser Family Foundation and computer programming support from Ase Sewall of Sewall Inc.
