anthropologists and other social scientists (Marshall, 1992). Attention to cultural diversity is made possible by this theoretical shift of focus.
Nonetheless, philosophical studies of the relevance of cultural difference to bioethics and decisionmaking are in their infancy. Pellegrino et al. (1992) and Veatch (1989) have both edited volumes that catalog the array of religious and cultural perspectives relevant to clinical bioethics. However, these works do not address the issue of potential conflicts in the United States when these myriad traditions collide (Orr et al., 1994). More fundamental is the question of whether an "ethnic perspective" on bioethics is philosophically justifiable, or desirable. African American philosophers express opposing views in a recent volume (Flack and Pellegrino, 1992).
Specific studies of how culturally diverse patients respond to innovations such as advance directives are only beginning to appear.9 Clinical case reports demonstrate the potential for conflict when patients and providers have conflicting expectations (Meleis and Jonsen, 1983; Muller and Desmond, 1992). Garrett and colleagues have demonstrated that African Americans differ from European Americans in their willingness to complete advance directives and desires about life-sustaining treatment (Garrett et al., 1993). Using survey research to investigate wishes about life-prolonging treatment, Caralis and colleagues found that significantly more African Americans and Hispanics, "wanted their doctors to keep them alive regardless of how ill they were, while more … whites agreed to stop life-prolonging treatment under some circumstances" (1993, p. 158).
Although results from surveys asking hypothetical questions about patient preferences have inherent limitations, the findings are intriguing and confirm the importance of political tensions. The probability is high that populations traditionally underserved by the health care system will find it hard to trust that physicians will make decisions in their best interest. Numerous studies have documented the higher morbidity and mortality of U.S. minority populations, as well as lack of access to care (Krieger, 1993; Adler et al., 1993). Empirical evidence documents a lower rate of organ donation by minority groups, perhaps another indication of lack of trust (Kasiske et al., 1991; Kjellstrand, 1988). With the increasing prevalence of managed care, suspicions will mount; economic incentives for the use of life-sustaining interventions will likely be transformed. Mistrust of government programs—and the health professionals seen as government agents—is justified by the historical record (see Jones, 1981, on the Tuskegee syphilis study).
Bioethics research paradigms, because of failure to include analysis of the social context within which end-of-life decisions are made, have failed to account for significant power differentials between patients and provid-