A landmark report from the Institute of Medicine (IOM) in 2003 documented that from the simplest to the most technologically advanced diagnostic and therapeutic interventions, African American (or black) individuals and those in other minority groups receive fewer procedures and poorer-quality medical care than white individuals.1 These differences existed even after statistical adjustment for variations in health insurance, stage and severity of disease, income or education, comorbid disease, and the type of health care facility. Very limited progress has been made in reducing racial/ethnic disparities in the quality and intensity of care.2 The IOM report concluded that multiple factors contribute to racial disparities in medical care and that unconscious bias by health care professionals contributes to deficits in the quality of care. This Viewpoint discusses the potential contribution of societal racial bias to disparities in health care and health status.
Sources of Bias and Its Consequences
A possible role of racial bias is surprising to many people, given that large positive changes in the racial attitudes of white people have occurred in the last 60 years. However, research also reveals that implicit bias is one of the multiple ways in which racial bias and racism manifest but makes it less readily recognizable. In navigating their social world, all humans engage in conscious, deliberate cognitive processes, as well as implicit (unconscious) automatic evaluative processes based on images stored in memory. Negative beliefs about race are deeply ingrained in US culture, with many images in popular culture devaluing nonwhite racial populations. For example, greater exposure to television programs that portray black people negatively is associated with higher levels of racial prejudice toward blacks. Although blacks appear more frequently on television programs than in the past, more negative nonverbal behavior (facial expressions and body language) is directed toward black characters than to status-matched white ones, and exposure to nonverbal bias increases viewers’ racial biases, even though viewers are not consciously aware of these patterns. A study of the books, newspapers, and other materials that many college-educated US adults may read in their lifetimes found that the word “black” was most frequently paired with, in order of frequency, poor, violent, religious, lazy, cheerful, dangerous.3 “White” was most frequently paired with wealthy, progressive, conventional, stubborn, successful, and educated. Thus, negative stereotypes of black individuals (violent, lazy, and dangerous) reflect, in part, how often US adults have seen these words paired with black over their lifetime.
The presence of implicit biases that favor whites over blacks is consequential for the behavior even of individuals who sympathize with those who have experienced injustice and are committed to principles of racial equality. Higher levels of implicit bias among clinicians have been directly linked with biased treatment recommendations in the care of black patients, although the pattern is not uniform.4 Implicit bias by clinicians has also been associated with poorer quality of patient-physician communication and lower patient ratings of the quality of the medical encounter.
Race/Ethnicity and Health
Disparities in medical care contribute to the even larger challenge of pervasive and persistent racial/ethnic disparities in health status. In the United States, compared with white individuals, black individuals have earlier onset of multiple illnesses, greater severity and more rapid progression of diseases, higher levels of comorbidity and impairment throughout the life course, and increased mortality rates. Where data are available, similar patterns are evident for American Indians, Native Hawaiians and other Pacific Islanders, low socioeconomic status (SES) Asian populations, and US-born Latinos and those with long-term residence in the United States. These racial/ethnic disparities in health are costly to society in terms of loss of life in the most productive years of life. For example, black-white differences in mortality have been estimated to account for the premature deaths of 260 African Americans every day.5
Although racial disparities in access to care, as well as in the quality and intensity of care, contribute to racial/ethnic disparities in the severity and course of disease, most racial disparities in the onset of illness occur prior to the presentation of patients to receive health care. Racial/ethnic differences in SES are large and contribute to racial/ethnic differences in health. In 2013, for every dollar of household income white people earned, Hispanic households earned 70 cents and black households earned 59 cents (identical to the black-white gap in income in 1978).6 Socioeconomic status, whether measured by income, education, or occupational status, in the United States and globally is a central factor associated with variations in health. The opportunities to be healthy in the environments in which individuals live, learn, work, play, and worship are key determinants of health. In US data, SES tends to be a stronger factor related to variation in health than race, and SES disparities in health are evident within each racial group.
Discrimination and Health Effects
When the health of black and white people are compared at equivalent levels of income and education, racial disparities are reduced but remain evident at all levels of SES.7 A growing body of evidence suggests that societal racial bias contributes to these residual effects of race in multiple ways. Scientific evidence indicates that conscious and unconscious bias combine to create patterns of racial/ethnic discrimination in employment, bank loans, housing, purchasing a car, and hailing a taxi.8 Individuals who face discrimination are aware of some of these experiences, which are a source of psychosocial stress. A recent review9 documented that self-reported measures of discrimination were adversely related to multiple indicators of health (eg, hypertension, all-cause mortality, incident asthma, incident breast cancer, and poor mental health), several early indicators of clinical disease (eg, inflammation, carotid intima-media thickness, visceral fat, obesity, coronary artery calcification, shorter telomeres, and cortisol dysregulation), and health behaviors (eg, poor sleep quantity and quality, cigarette smoking, and substance use). In addition to exposure to discrimination, similar to other types of stressors, the threat of discrimination was also related to increased cardiovascular response, symptoms of poor mental health, and hypertension. Perceived discrimination has also been associated with lower levels of health care seeking and adherence behaviors, and research in the United States, South Africa, Australia, and New Zealand has revealed that discrimination makes an incremental contribution over SES in accounting for racial disparities in health.
Racial bias also affects health through institutional mechanisms. Although segregation has been illegal since 1968 and black individuals show the highest preference for residing in integrated neighborhoods, declines in segregation in recent decades have been very small and have had negligible effects on the residential concentration and isolation of most African Americans and the geographic concentration of urban poverty. Segregation affects health by restricting socioeconomic attainment through limiting access to quality educational and employment opportunities. Segregation also leads to residence in poorer-quality housing and in neighborhood environments with elevated risk of exposure to acute and chronic psychosocial stressors and toxic chemicals and reduced access to resources and amenities that enhance health, including medical care.
Some physicians are unaware that racial disparities exist and question the evidence of disparities.2 Successfully addressing the possibility of clinician bias begins with awareness of the pervasiveness of disparities, the ways in which bias can influence clinical decision making and behavior, and a commitment to acquiring the skills to minimize these processes.
Medical schools, health care organizations, and credentialing bodies should pay greater attention to disparities in health and health care as a high national priority. These organizations should redouble their efforts to increase awareness of disparities, enhance diversity in the health professions, and work toward eliminating discrimination and its adverse effects on health and health care. Considerable evidence is available to guide in the implementation of interventions to reduce racial/ethnic differences in health and health care.2 Moreover, the United States is not unique. Similar patterns of racial health disparities are found in other countries, such as the United Kingdom, Australia, Canada, New Zealand, South Africa, and Brazil. Leadership on racial equity to address health disparities in the United States could have positive national effects and additional potential effects on stigmatized racial populations around the world.
The health care system cannot eliminate racial/ethnic disparities in health. Health care professionals need to collaborate with other sectors of society to increase awareness about the health implications of social policies in domains far removed from traditional medical and public health interventions. Much of the contemporary disease burden is linked to behaviors that are potentially modifiable with access to timely information and the necessary resources and opportunities to facilitate the challenge of behavioral change. Many individuals live, learn, work, and play in disadvantaged contexts where it is nearly impossible to pursue healthy choices. Multilevel policies and interventions in homes, schools, neighborhoods, workplaces, and religious organizations can help remove barriers to healthy living and create opportunities to usher in a new culture of health in which the healthy choice is the easy choice.
Focusing only on racial disparities in health, in which the health of white people is used as the reference, obscures a major challenge that the United States faces in improving health. A recent IOM report indicated that people in the United States have poorer health than individuals in other high-income countries and that even the most advantaged individuals had worse health than their peers in other affluent nations.10 Health policy initiatives in the United States are needed to improve the health of all, even while those policies seek to enable those farthest behind to improve their health more rapidly than the rest of the population so that the large gaps in health by race and SES will ultimately be reduced. Large social inequities in health are unacceptable in a nation founded on the principles of liberty, equality, and justice for all, and there is inadequate recognition that dismantling racial bias in all of its forms is likely to be a potent health intervention.
Corresponding Author: Ronald Wyatt, MD, MHA, The Joint Commission, One Renaissance Blvd, Oakbrook Terrace, IL 60181 (firstname.lastname@example.org).
Conflict of Interest Disclosures: Both authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Correction: This article was corrected for byline order on August 11, 2015.
AR. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC: National Academy Press; 2003.
et al. The impact of racism on clinician cognition, behavior, and clinical decision making. Du Bois Rev
. 2011;8(1):199-218. PubMedGoogle ScholarCrossref
N. A systematic review of the extent and measurement of healthcare provider racism. J Gen Intern Med
. 2014;29(2):364-387.PubMedGoogle ScholarCrossref
et al. Black-white inequalities in mortality and life expectancy, 1933-1999. Public Health Rep
. 2001;116(5):474-483.PubMedGoogle Scholar
C. Race, socioeconomic status, and health. Ann N Y Acad Sci
. 2010;1186(1):69-101.PubMedGoogle ScholarCrossref
DR. Self-reported experiences of discrimination and health. Annu Rev Clin Psychol
. 2015;11:407-440.PubMedGoogle ScholarCrossref