- Journal List
- J Gen Intern Med
- v.22(6); 2007 Jun
- PMC2219858
Reducing Racial Bias Among Health Care Providers: Lessons from Social-Cognitive Psychology
Diana Burgess
1Center for Chronic Disease Outcomes Research (a VA HSR&D Center of Excellence), Veterans Affairs Medical Center, One Veterans Drive, Minneapolis, MN 55417 USA
2Department of Medicine, University of Minnesota, Minneapolis, MN USA
Michelle van Ryn
1Center for Chronic Disease Outcomes Research (a VA HSR&D Center of Excellence), Veterans Affairs Medical Center, One Veterans Drive, Minneapolis, MN 55417 USA
3Department of Family Medicine and Community Health, University of Minnesota, Minneapolis, MN USA
John Dovidio
4Department of Psychology, Yale University, New Haven, CT USA
Somnath Saha
5Section of General Internal Medicine, Portland Veterans Affairs Medical Center and Department of Medicine, Oregon Health & Science University, Portland, OR USA
Abstract
The paper sets forth a set of evidence-based recommendations for interventions to combat unintentional bias among health care providers, drawing upon theory and research in social cognitive psychology. Our primary aim is to provide a framework that outlines strategies and skills, which can be taught to medical trainees and practicing physicians, to prevent unconscious racial attitudes and stereotypes from negatively influencing the course and outcomes of clinical encounters. These strategies and skills are designed to: l) enhance internal motivation to reduce bias, while avoiding external pressure; 2) increase understanding about the psychological basis of bias; 3) enhance providers’ confidence in their ability to successfully interact with socially dissimilar patients; 4) enhance emotional regulation skills; and 5) improve the ability to build partnerships with patients. We emphasize the need for programs to provide a nonthreatening environment in which to practice new skills and the need to avoid making providers ashamed of having racial, ethnic, or cultural stereotypes. These recommendations are also intended to provide a springboard for research on interventions to reduce unintentional racial bias in health care.
Despite ample evidence of racial and ethnic disparities in health care1,2 progress in correcting these inequities remains elusive.1–3 Provider behavior has been identified as 1 important contributor to disparities in health care. There is significant evidence that (1) health care providers hold stereotypes—based on patient race, class, sex, and other characteristics4–17—that influence their interpretation of behaviors and symptoms, and their clinical decisions17–25; (2) application of such stereotypes frequently occurs outside conscious awareness26–31; and (3) providers interact less effectively with minority than with white patients.32–37 In response to this evidence, significant resources have been devoted to programs to prepare providers to better care for patients from diverse backgrounds. These programs, however, typically focus on improving providers’ cross-cultural communication skills and, as such, are likely to have only limited effects on the unconscious cognitive processes that result in stereotype activation and application. In fact, there has been relatively little discussion to date of how to mitigate the negative impact of unconscious racial stereotyping among health care providers, despite the acknowledged need to do so.32–35 This paper is intended to address this gap by drawing from a highly developed body of research from social cognitive psychology to recommend a set of evidence-based intervention strategies.
First, it is important to note that overt expressions of prejudice and negative racial stereotypes have declined substantially over time, as norms condemning bias and endorsing racial equality have become stronger and more widespread across the population.38,39 However, even consciously egalitarian people may hold negative ethnic and racial stereotypes and attitudes,26–31 of which they may not be fully conscious. For example, Whites tend to have unconscious stereotypes of Whites as intelligent, successful, and educated, and of Blacks as aggressive, impulsive, and lazy,40 and Whites spontaneously associate different health conditions with Whites and Blacks. 41 The consequences of this unintentional bias ultimately may be as adverse as more overt biases.42–44 Furthermore, as unconscious biases are “habits of mind”,31,45 learned over time through repeated personal experiences and cultural socialization, they are highly resistant to change.46 Given the realities of medical training and continuing education, our recommended strategies are not intended to change these associations directly,47 but rather to give providers strategies and skills to prevent unconscious attitudes and stereotypes from influencing the course and outcomes of clinical encounters in negative ways. It is important to note that these proposed strategies have not yet been tested in the realm of health care and should undergo rigorous evaluation before widespread adoption.
COUNTERACTING UNCONSCIOUS PREJUDICE AND STEREOTYPES THROUGH INDIVIDUATION
Promising evidence in social cognitive psychology indicates that with sufficient motivation, cognitive resources, and effort, people are able to focus on the unique qualities of individuals, rather than on the groups they belong to, in forming impressions and behaving toward others.48,49 Even automatically activated prejudice and stereotypes can be inhibited when people are perceived more in terms of their particular qualities than primarily as members of social categories.50 Interventions to reduce bias in provider judgment, behavior, and decision making should therefore promote the cognitive strategy of individuation, in which the provider focuses on the individual attributes of a particular patient, as opposed to categorization, in which the provider perceives the patient through the filter of his or her group membership (e.g., race). Based on this premise, we outline the elements of a training program that builds upon psychological research on unconscious prejudice and stereotypes. Our recommendations are rooted in an evidence-based conceptual model, depicted in Figure 1. It recognizes the importance of motivation, information, and skills, which have been identified as key ingredients for successful interventions in other domains,51 but also considers the importance of emotion in interracial interaction.
Enhance Internal Motivation and Avoid External Pressure to Reduce Bias
As we have discussed earlier, much of contemporary racism occurs unintentionally, fueled by unconscious prejudice and stereotypes. To the extent that many people are unaware of their biases, there is little motivation for change. However, for people who believe that they are unbiased and aspire to be nonprejudiced, it may be possible to capitalize on their good intentions to motivate efforts to reduce their unconscious biases once they become aware of them. Techniques that lead people to recognize their unconscious biases include exercises where they are prompted to compare what they would do and what they should do in a variety of intergroup circumstances,45 and recently developed procedures (such as the Implicit Association Test52,53) that can reveal unconscious prejudice and stereotypes. These procedures can engender negative emotional states54,55 that motivate people to become more sensitive to and attempt to counteract the effects of unconscious prejudice and stereotypes. Promoting awareness of the social and historical context of race, the evidence documenting racial disparities in the quality of health care,2 and the evidence that provider bias may contribute to those disparities15 may also help in nurturing internal motivation to reduce bias, based on a sense of fairness.
A critical element of this process is self-discovery within a nonthreatening, private context. Anticipated public censure can have paradoxical effects.57,58 Efforts to reduce racial/ethnic disparities should avoid imposing a “politically correct” agenda, but instead appeal to providers’ desire to provide the best possible care to all patients.
Enhance Understanding of the Psychological Basis of Bias
In contrast to the traditional psychological perspective that “pathologized” prejudice and stereotyping, current understanding posits that the cognitive strategy of categorization that gives rise to stereotyping and racial prejudice is a normal aspect of human cognition. Helping providers understand this may allow them to approach their own potential biases in a more informed and open way. This can be accomplished with selected readings, demonstrations of unconscious stereotyping using web-based tools, and guided discussion.59,60
One direct consequence of this approach of openly acknowledging stereotypes is that it can facilitate sensitivity to the negative impact of unconscious prejudice and stereotypes, whereas denial of their existence and direct attempts to suppress their activation can have further negative consequences. Stereotype suppression (e.g., trying to push away any stereotypic thoughts that arise while in the presence of an African-American patient) can lead to a number of negative, unintended consequences. Experiments testing the impact of instructions to avoid stereotyping have shown very short-term benefit with a “rebound effect,” in which the stereotype later recurred at a higher rate than among the control groups who were not instructed to suppress.61–64 Stereotype suppression also can result in increased social distancing, thus potentially undermining the interpersonal quality of the encounter.61 In addition, suppressing stereotypes requires effort,65 which can deplete cognitive resources and adversely affect the ability to solve problems and make correct decisions.66 It is therefore important to reinforce that stereotypes—even negative racial stereotypes—are a natural phenomenon in our society, and that it is better to recognize and use strategies (discussed below) to counteract them, rather than to try to actively suppress them.
Enhance Providers’ Confidence in their Ability to Successfully Interact with Socially Dissimilar Patients
Psychological research has shown that Whites often feel anxious when interacting with Blacks, because of a lack of positive experiences with interracial encounters, leading them to avoid such interactions.67 In the context of the clinical encounter, this may translate into White providers’ engaging in avoidance behaviors and spending less time with non-White patients, leading to poorer patient–provider relationships. Non-White patients, who may be particularly vigilant for signs of prejudice or rejection,68 may interpret signs of anxiety displayed by White providers as reflecting negative attitudes.69 Nonverbal behaviors associated with anxiousness overlap considerably with cues of dislike.70
The most successful way to alleviate intergroup anxiety and increase provider confidence is through direct “contact” with members of other groups.71,72 Thus interactive, facilitated discussions, particularly in which people interact in individualized ways, among colleagues of different race and ethnicity may be one method to enhance providers’ confidence in interracial interactions. Programs designed to enhance providers’ overall communication skills73 may also improve interracial clinical encounters, although this hypothesis has not been tested.
Enhance Emotional Regulation Skills Specific to Promoting Positive Emotions
Recent research suggests that providers who experience higher levels of positive emotion during clinical encounters may be less likely to categorize patients in terms of their racial, ethnic, or cultural group and more likely to view patients in terms of their individual attributes (i.e., individuation). For instance, in several studies, White subjects in whom positive emotions were induced more accurately recognized Black faces, suggesting that positive emotional states can lead to lower levels of racial categorization.74 Positive emotion has also been shown to lead to the use of more inclusive social categories, so that people are more likely to view themselves as being part of a larger group,28,75 which can facilitate empathy and increase the capacity to see others as members of a common “ingroup,” as opposed to “outgroup.”43 More generally, positive emotional states have been shown to broaden the scope of attention and the tendency to attend to new information,76 a mental state that is consistent with greater levels of individuation and lower levels of categorization.
Clearly, an individual’s emotional states vary in ways that are not always easy to regulate. However, simply being aware that stress and negative emotions may increase stereotyping could make providers more vigilant about when their interactions with patients are likely to be biased. Additionally, when time and circumstances allow, the use of stress-reducing techniques to enhance emotional well-being before patient encounters may help reduce racial bias. Teaching these methods—e.g., mindfulness techniques, meditation, Balint groups—may be an appropriate aspect of training programs that aim to reduce provider bias.
Increase Perspective Taking and Affective Empathy
Enhancing provider empathy is a particularly promising strategy for increasing the effectiveness of patient–provider interactions generally, and especially for interracial interactions. Although the development of empathic skill is a general aspect of medical training, providers may not recognize inherent obstacles to empathizing with patients from a different racial or ethnic group. People tend to have less empathy for others perceived as dissimilar,77 including members of other social categories,67 which typically result in low rapport.29,35,37,69 For instance, one study found that non-Hispanic White psychologists viewing videotapes of an actress playing a patient reported less empathy when the patient was described as Hispanic rather than White.78
Empathy has both cognitive and affective components.79,80 The cognitive component is perspective taking, in which a person is able to consider a situation from the position of another, e.g., imagining themselves in the other’s shoes. Controlled laboratory studies instructing subjects to adopt a perspective associated with another person and his or her situation have been shown to reduce bias toward a range of stigmatized groups,81 including Blacks,82,83 and to inhibit the activation of unconscious stereotypes and prejudices.84 Similarly, one study found that physicians rated as “empathetic/compassionate” by their peers had less stereotypic attitudes toward patients than physicians classified as low on empathy.85
The affective component involves empathic emotions. In practice, perspective taking and affective empathy are closely related: perspective taking arouses affective empathy, and affective empathy facilitates adopting the other’s perspective.79 Both ultimately produce more positive orientations toward the other person and a greater interest in the other’s welfare.86 Many studies have documented the positive effect of provider empathy on patients, which include increased patient satisfaction, adherence, self-efficacy, perceptions of control, less emotional distress, and better outcomes.87–90 Unfortunately, empathy has been shown to decline over the course of medical training91–94 and the course of physicians’ medical careers,85 suggesting that interventions to increase empathy might need to be repeated over time to be effective.
Programs might promote routine use of the types of imagery strategies successfully used in experimental studies, such that providers spend a moment in each encounter “imagining [themselves] living in the shoes of the patient...feeling what he [/she] feels.” As with any behavior, repeated rehearsal can make the behavior routine. There is also some evidence that having providers take on the role of patients in role-play exercises can increase empathy among health care providers. This strategy might be useful in team discussions of difficult patients.95–97
Improve Ability to Build Partnerships with Patients
Partnership building can improve the effectiveness of patient–provider interactions by reframing the interaction as one between collaborating equals, rather than between one high-status person, the provider, and one low-status person, the patient.98 Whether people perceive another as a partner on the “same team” (i.e., as part of the same group, the “ingroup”) or as a member of a different groups (“outgroup”) has profound implications on their reactions to the other.42,98–101 These reactions are so ingrained and fundamental that they occur automatically, without awareness or intention. In the United States, race, ethnicity, and sex represent automatically activated categories that form the basis for ingroup–outgroup differentiation. In general, people retain information in a more detailed fashion, remember more positive information, and are more forgiving of behaviors for ingroup compared to outgroup members.102–104
Perceptions of common ingroup membership also increase psychological bond and feelings of “oneness” that facilitate perspective taking82 and the arousal of empathy in response to their needs or problems.105–107 Creating a partnership, which produces a common group identity, reduces conscious and unconscious racial biases among Whites in interracial interactions.98,108 Partnership-building between provider and patient would not only be expected to produce a more positive and responsive orientation of the provider to the patient, but also a more open and cooperative orientation of the patient to the provider. Developing a common identity through partnership building is conceptually similar to the “finding common ground” component of the patient-centered approach to care.109 Finding common ground requires that two parties reach agreement with regard to the nature of the medical problem and other priorities, the goals of treatment, and the roles of the doctor and patient. Studies of patient-centered care suggest significant benefits in terms of patient satisfaction and medical outcomes.110–114
CONCLUSION
Using lessons from research in social-cognitive psychology, we have made recommendations that we hope will inform and accelerate efforts to reduce unintentional bias among health care providers. We have proposed specific objectives, and have discussed the scientific rationale behind them, as a way of establishing a foundation on which current and future interventions to reduce provider racial bias might build. Those who choose to build on this foundation should be aware of several important caveats. First, although our recommendations are based on empirical work, most of that work has not been conducted in health care providers and settings. It will therefore be important to rigorously evaluate future interventions in terms of their success in reducing racial bias in patient–provider encounters, as well as to understand any unintended consequences. Second, it is important to recognize that the initial stages of changing old habits and developing new skills are likely to temporarily result in greater cognitive burden for providers during clinical encounters. Thus, it will take time and practice before providers are able to execute the strategies we have discussed in an automatic way, and to therefore fully realize their purported benefit. Third, not all of our recommendations should necessarily be construed as elements of single or stand-alone programs. Efforts such as enhancing empathy and partnership building are relevant not only to reducing provider bias but to improving patient–provider relationships in general. These efforts may already be incorporated into training programs and curricula that are not explicitly focused on reducing racial disparities. In such cases, our framework might serve as a guide to help educators and administrators determine to what extent their training environment includes, or does not include, components that are likely to contribute to reducing provider racial bias. Finally, our discussion is not meant to be a comprehensive review of all potentially important principles and strategies for reducing provider bias, but instead is intended to provide a starting point for the important task of reducing racial bias in health care.
Acknowledgments
Dr. Burgess is supported by a Merit Review Entry Program Award from VA HSR&D. Dr. Saha is supported by awards from the VA HSR&D Advanced Career Development Program and the Robert Wood Johnson Generalist Physician Faculty Scholars Program. The authors would like to thank Dr. Hanna Bloomfield for her helpful comments.
Conflicts of Interest None disclosed.
References
Articles from Journal of General Internal Medicine are provided here courtesy of Society of General Internal Medicine