FlashReport
Aversive racism and medical interactions with Black patients: A field study

https://doi.org/10.1016/j.jesp.2009.11.004Get rights and content

Abstract

Medical interactions between Black patients and nonBlack physicians are usually less positive and productive than same-race interactions. We investigated the role that physician explicit and implicit biases play in shaping physician and patient reactions in racially discordant medical interactions. We hypothesized that whereas physicians’ explicit bias would predict their own reactions, physicians’ implicit bias, in combination with physician explicit (self-reported) bias, would predict patients’ reactions. Specifically, we predicted that patients would react most negatively when their physician fit the profile of an aversive racist (i.e., low explicit–high implicit bias). The hypothesis about the effects of explicit bias on physicians’ reactions was partially supported. The aversive racism hypothesis received support. Black patients had less positive reactions to medical interactions with physicians relatively low in explicit but relatively high in implicit bias than to interactions with physicians who were either: (a) low in both explicit and implicit bias, or (b) high in both explicit and implicit bias.

Introduction

Although there has been dramatic improvement in the health of all people living in the United States over the past 50 years, the level of disparities between Black and White Americans for several key indicators of health remains essentially unchanged (National Center for Health Statistics, 2006). Among the suggested contributors to this disparity are health providers’ racial prejudice and stereotypes (Institute of Medicine, 2003, van Ryn et al., 2006). The potential impact of such provider bias on the health care that Black patients receive is substantial: Approximately 75% of all medical interactions for Black patients in the US are “racially discordant” – that is, they involve nonBlack health care providers (Penner, Albrecht, Coleman, & Norton, 2007). Moreover, relative to racially concordant medical interactions, racially discordant interactions are characterized by less patient trust (Cooper et al., 2003), less positive affect (Johnson, Roter, Powe, & Cooper, 2004), fewer attempts at relationship building (Simonoff, Graham, & Gordon, 2006), and less joint decision-making (Koerber, Gajendra, Fulford, BeGole, & Evans, 2004). Although provider bias has been proposed as a contributor to such outcomes in racially discordant interactions, it has not, as far as we know, been directly investigated. Thus, the present research investigated the impact of physicians’ explicit and implicit racial bias on medical encounters with Black patients.

Behavior toward Blacks is influenced by both explicit racial attitudes, traditionally assessed with self-reports, and by implicit attitudes, which are automatically activated typically without conscious awareness (Greenwald, Poehlman, Uhlmann, & Banaji, 2009). Moreover, explicit and implicit measures of bias tend to predict different responses (Dovidio, Kawakami, Smoak, & Gaertner, 2009). Explicit measures predict blatant discrimination, whereas implicit measures predict more subtle expressions of discrimination that often occur unintentionally, such as nonverbal behavior and negative decisions in complex situations in which bias could be attributed to factors other than race (McConnell and Leibold, 2001, Son Hing et al., 2008). For instance, Green et al. (2008) found that when presented with vignettes about patients with symptoms of a myocardial infarction, physicians higher in implicit bias were less likely to recommend appropriate drugs for Black patients.

Whereas people are aware of their overt and deliberative (e.g., verbal) behaviors, which relate to explicit measures of their attitudes, they may be unaware of their subtly biased and spontaneous (e.g., nonverbal) behaviors, which relate to implicit measures (Dovidio et al., 2002, McConnell and Leibold, 2001). As targets of these behaviors, however, Blacks and members of other disadvantaged groups attend closely to these subtly biased behaviors, which critically shape their impressions of intergroup interactions (Dovidio, Kawakami, & Gaertner, 2002). The inconsistency between positive overt expressions and negative subtle displays may be particularly problematic because this kind of mismatch is generally perceived to reflect deceitfulness (beyond even a mismatch between negative overt and positive subtle behaviors; Eskritt & Lee, 2003), which can be especially detrimental in interracial interactions that are often characterized by intergroup mistrust (Dovidio, Gaertner, Kawakami, & Hodson, 2002).

Indeed, Dovidio and Gaertner, 2004, Gaertner and Dovidio, 1986) proposed that a subtle form of bias, “aversive racism,” can have a particularly detrimental influence on interracial interactions. An aversive racist is a person who is low in explicit bias but who harbors implicit racial biases against Blacks. Aversive racism research has traditionally focused on Whites, but it also applies to the orientations of members of other groups (e.g., Asians; Kawakami, Dunn, Karmali, & Dovidio, 2009) toward Blacks. The mixed messages conveyed by aversive racists during interracial interactions can interfere with effective social coordination and jointly affect Blacks and nonBlacks’ abilities to work together successfully. For example, dyads consisting of a Black participant and a White aversive racist performed less effectively than dyads involving Blacks with Whites who had consonant explicit and implicit attitudes and ironically, even those with high explicit and high implicit bias (Dovidio, 2001).

Recently laboratory work on implicit bias has been extended to health providers, but whereas prior studies of physician bias focused on treatment decisions using retrospective or vignette methodologies, we investigated the relationship of nonBlack (i.e., Asian and White) physicians’ implicit and explicit racial bias to both physicians’ and Black patients’ responses to actual medical interactions in an inner-city primary care clinic. We predicted that physicians’ perceptions of their own behavior (involving the patient in the treatment decision and feeling on the “same team”) would relate primarily to physicians’ explicit (self-reported) prejudice. By contrast, we hypothesized that patients’ perceptions of the encounter would relate to physicians’ implicit bias, unintended activation of biased attitudes measured using the Implicit Association Test (IAT; Greenwald, Nosek, & Banaji, 2003), in combination with their level of explicit prejudice. Drawing on previous work demonstrating that dyads involving Blacks with Whites who fit the aversive racism profile perform particularly poorly on a cooperative task (Dovidio, 2001), we predicted a physician implicit bias × physician explicit bias interaction for Black patients’ perceptions of their involvement in the treatment decision as well as their personal responses to the medical encounter (perceptions of physician warmth and friendliness, feeling on the same team, and satisfaction with the visit). Patients’ reactions were expected to be least positive when physicians were low in explicit prejudice and high in implicit bias.

Section snippets

Participants

The patients were 150 Black patients (112 women, 38 men; average age, 43.63) at an inner-city primary care clinic in the Midwest. (There were no White patients at the clinic during 18 months of data collection.) Participants, who received $20.00 gift cards, were recruited consecutively. Seventy-three percent of the patients asked to participate agreed to do so; the sample closely matched the demographics of the clinic patient population.

The physicians, who received a $50.00 incentive for

Results

Because physicians interacted with more than one patient (i.e., patients are nested within physicians), to control for non-independence we used the General Estimating Equation (GEE) procedure, a form of multilevel modeling (Hanley et al., 2003, Hardin and Hilbe, 2003). We included implicit bias and explicit prejudice as main effects and their interaction in all equations. Also, all the measures in the equations were standardized by converting them to z scores; thus, the parameter estimates (β)

Discussion

Provider bias has been suggested as a contributor to health care interactions (Institute of Medicine, 2003), but there is only limited, indirect evidence for this hypothesis. In addition, although there is some evidence that physicians’ implicit bias predicts physician treatment decisions for Black patients (Green et al., 2008), no previously published research to our knowledge has directly linked physicians’ racial bias to reactions of both doctors and patients following medical interactions.

Acknowledgments

This research was supported by a Grant from the National Institute of Child Health and Development (1R21HD050445001A1) to Louis A. Penner, Principal Investigator, and a Grant from the National Cancer Institute (U01CA114583) to Terrance L. Albrecht and Peter Lichtenberg, Principal Investigators, and by a Grant from the National Science Foundation (BCS-0613218) to Samuel L. Gaertner and John F. Dovidio.

We thank Professor Brian Nosek and members of his laboratory at the University of Virginia for

References (38)

  • J.F. Dovidio et al.

    Aversive racism

  • A.R. McConnell et al.

    Relations among the Implicit Association Test, discriminatory behavior, and explicit measures of racial attitudes

    Journal of Experimental Social Psychology

    (2001)
  • L. Aiken et al.

    Multiple regression: Testing and interpreting interactions

    (1991)
  • J.C. Brigham

    College students’ racial attitudes

    Journal of Applied Social Psychology

    (1993)
  • D. Burgess et al.

    Reducing racial bias among health care providers: Lessons from social–cognitive psychology

    Journal of General Internal Medicine

    (2007)
  • L.A. Cooper et al.

    Patient-centered communication, ratings of care, and concordance of patient and physician race

    Annals of Internal Medicine

    (2003)
  • L.F. Degner et al.

    The control preferences scale

    Canadian Journal of Nursing Research

    (1997)
  • J.F. Dovidio

    On the nature of contemporary prejudice. The third wave

    Journal of Social Issues

    (2001)
  • J.F. Dovidio et al.

    Why can’t we just get along? Interpersonal biases and interracial distrust

    Cultural Diversity & Ethnic Minority Psychology

    (2002)
  • J.F. Dovidio et al.

    Implicit and explicit prejudice and interracial interaction

    Journal of Personality and Social Psychology

    (2002)
  • J.F. Dovidio et al.

    The roles of implicit and explicit processes in contemporary prejudice

  • R.A. Epstein

    Disparities and discrimination in health care coverage: A critique of the Institute of Medicine study

    Perspectives in Biology and Medicine

    (2005)
  • M. Eskritt et al.

    Do actions speak louder than words? Preschool children’s use of the verbal–nonverbal consistency principle during inconsistent communications

    Journal of Nonverbal Behavior

    (2003)
  • S.L. Gaertner et al.

    The aversive form of racism

  • S.L. Gaertner et al.

    Reducing intergroup bias: The common ingroup identity model

    (2000)
  • A.R. Green et al.

    The presence of implicit bias in physicians and its predictions of thrombolysis for Black and White patients

    Journal of General Internal Medicine

    (2008)
  • A.G. Greenwald et al.

    Understanding and using the Implicit Association Test: I. An improved scoring algorithm

    Journal of Personality and Social Psychology

    (2003)
  • A.G. Greenwald et al.

    Understanding and using the Implicit Association Test: III. Meta-analysis of predictive validity

    Journal of Personality and Social Psychology

    (2009)
  • J.A. Hanley et al.

    Statistical analysis of correlated data using generalized estimating equations: An orientation

    American Journal of Epidemiology

    (2003)
  • Cited by (0)

    View full text