Advertisement

Ignoring It Isn't an Option: Racial Bias in Emergency Medicine

      A cartoon published in the mock newspaper The Onion a few weeks after President Obama's electoral victory last fall featured beaming black and white Uncle Sam figures, hats off and elbows interlocked, dancing on a grave with a headstone labeled “Racism: Deceased, 2008.”

      “Kelly” (pseudonym for Ward Sutton). A funky dance to the year that was (cartoon). The Onion, December 15, 2008. Available at: http://www.theonion.com/content/cartoon/dec-15-2008. Accessed March 24, 2009.

      The decline of overt bigotry, capped by the election of our first African-American president, gives Americans legitimate grounds for pride—but as the cartoon's self-parodying tone implies, decline is not demise. Self-congratulatory celebrations, it seems, are a bit premature.
      One area where America's medical care system persists in needing improvement, say physicians who have studied the problem, is the equitable treatment of members of different racial and ethnic groups. Disparities in the quality of care, an unmistakable part of physicians' and patients' personal experiences, have been increasingly well documented in the research literature over about the past decade and a half. Although the field is still relatively new and empirical evidence is only beginning to accumulate in many areas, findings now substantiate the observation that members of minorities consistently receive subpar care in emergency departments (EDs) and other facilities for cardiac conditions, diabetes, renal disease, mammographic screening, vaccinations, and numerous kinds of pain.
      • Cone D.C.
      • Richardson L.D.
      • Todd K.H.
      • et al.
      Health care disparities in emergency medicine.
      • Pezzin L.E.
      • Keyl P.M.
      • Green G.B.
      Disparities in the emergency department evaluation of chest pain patients.
      Moreover, though differences in socioeconomic status and insurance coverage are associated with differential rates of ED usage by different groups,
      • Baker D.W.
      • Stevens C.D.
      • Brook R.H.
      Determinants of emergency department use: Are race and ethnicity important?.
      they do not account for the anomalies in care, nor do patients' expectations or communication styles. “When they control for all the associated confounding factors,” says Elise O. Lovell, MD, an emergency physician at Advocate Christ Medical Center in Oak Lawn, Illinois, who has studied racial assumptions among medical residents, “and they say, ‘What's the last thing remaining?' Well, the last thing remaining is physician bias.”
      Explicit bias is gradually fading, but implicit bias is harder to recognize (by definition, being unconscious) and harder to root out. Persons who would openly admit bias to an interviewer are becoming rarer in American society generally, and vanishingly rare in the medical profession, but given the prevalence of stereotyped representations, Dr. Lovell and others argue, it's no surprise that these unconscious associations persist. “We're not even aware that we have it, but it's a collection of all of the cultural messages that we've received over time,” she says. It's essential to draw clear distinctions, particularly between implicit biases and racism, which involves deliberate behavior. “The idea that you can have bias inside of you,” Dr. Lovell observes, “doesn't necessarily mean that you're a racist…. It's been shown that the first step toward decreasing your bias, and decreasing how it will influence your decisions, is simply to be aware of it.”

      Ubiquitous But Measurable

      Nearly everyone finds the subject uncomfortable to discuss. No responsible physician or institution wants to admit to treating people differently for nonclinical reasons. Knox H. Todd, MD, MPH, an emergency physician and pain control specialist at New York's Beth Israel Medical Center, began observing ethnicity-based treatment disparities in his first post-residency position in San Antonio, Texas, then published his first study of the phenomenon (in this case, disparate uses of analgesia for Hispanic and non-Hispanic white patients with long-bone fractures) while working at UCLA.
      • Todd K.H.
      • Samaroo N.
      • Hoffman J.R.
      Ethnicity as a risk factor for inadequate emergency department analgesia.
      He recalls 2 reactions: “One was, ‘Of course that's true. We know that happens'—the underlying idea being that's implicit bias. The other reaction, which I see very often, is denial: ‘This can't occur. And certainly this doesn't occur in my institution.'” Even in the cosmopolitan and multicultural environment of Los Angeles, however, it was occurring. Responses by fellow emergency physicians and the hospital's administration to this airing of dirty laundry, Dr. Todd reports, were “extremely negative.”
      UCLA's hospital was by no means uniquely culpable. A series of single-site studies subsequently appeared, some replicating the results of Todd and colleagues' work and finding disparities, others not. (Todd mentions an occasional tendency for general improvement in the quality of pain management to correlate with improvement in disparities.) Enough related reports appeared by the early years of this decade that Congress asked the National Academy of Sciences to examine the issue; the subsequent report Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care
      exhaustively documented the clinical, systemic, and social factors that make it harder for members of minority groups to obtain appropriate care in a range of settings.
      The representation of different medical specialties in this overview reflected the patterns of care commonly provided to minorities. As knowledge about these phenomena took shape, Dr. Todd notes, “almost all of the analgesic ethnicity disparities work was from emergency medicine.” The specialty was also one of the first to include diversity as an explicit component of medical education, adding it to the American College of Emergency Physicians (ACEP)/Society for Academic Emergency Medicine core curriculum in 1996.
      Task Force on the Core Content for Emergency Medicine Revision
      Core content for emergency medicine.
      Since 2003, the Agency for Healthcare Research and Quality's annual National Healthcare Disparities Report (NHDR) have also tracked systemwide changes—both progress and regression. The most recent report (2007) finds, depressingly, that “the number of measures on which disparities have gotten significantly worse or have remained unchanged since the first NHDR is higher than the number of measures on which they have gotten significantly better for Blacks, Hispanics, American Indians and Alaska Natives, Asians, and poor populations.”
      National Healthcare Disparities Report.
      Oligoanesthesia is a highly nuanced area for disparities research, given the subjectivities involved: the patient's perception and interpretation of pain, communication with medical personnel, assessment and diagnosis, and treatment decisions. Todd notes that laboratory studies with experimental painful stimuli have found small differences in pain thresholds and tolerance between ethnic groups, but nothing clinically important.
      • Zatzick D.F.
      • Dimsdale J.E.
      Cultural variations in response to painful stimuli.
      • Greenwald H.P.
      Interethnic differences in pain perception.
      His work with a painful but relatively simple injury category, extremity fractures, shows a logical progression winnowing out various explanations for the disparate care provided. Observing that physicians' behavior, not patients' acculturation, is the strongest explanation for disparities in care, he quotes a maxim by psychologist Raymond C. Tait, PhD, of St. Louis University: “Pain assessment is a Rorschach test of the physician.”
      Hypothesizing at first that mediation by language might explain disparities, Dr. Todd was surprised to find at UCLA that it didn't. The next possibility was uneven pain assessment by either patients or physicians. In a sham study using visual pain scales,
      • Todd K.H.
      • Lee T.
      • Hoffman J.R.
      The effect of ethnicity on physician estimates of pain severity in patients with isolated extremity trauma.
      Dr. Todd says (completed before publication of the initial discrepancy report, and using a questionnaire designed to imply that the object of study was physicians' reasons for ordering x-rays), “there was absolutely no difference in the patients' expression of pain and the physicians' perception of pain by ethnicity.” Laypersons generally report pain at a higher level than physicians, who have of course witnessed much worse instances of it and thus “anchor” the maximal point on their visual scales higher than patients do, but the lack of ethnic differences within either the patient or physician assessments appeared to rule out the assessment step as the locus of bias. (Physicians' assessments differed slightly by patient ethnicity, but not significantly in either statistical or clinical terms.)
      “So physicians seem to be equally able to assess pain, no matter [about] ethnicity,” Dr. Todd continues. “One of the ideas behind ‘pain as a fifth vital sign' is [that] if we just make pain assessment mandatory, then disparities will disappear, right? But in fact that didn't happen.” A further study at Emory University examining black and white patients' experiences after long-bone fracture
      • Todd K.H.
      • Deaton C.
      • D'Adamo A.P.
      • et al.
      Ethnicity and analgesic practice.
      pointed directly to the ordering of analgesics as the step where ethnic disparities appeared; of the possible confounding variables assessed, including payer status (as a marker of socioeconomic status), only time since injury and black race, which roughly matched, predicted lack of analgesia in the ED. As additional support for physicians' decisions as the causal factor behind disparities, Dr. Todd cites a study by David Fosnacht's group in Salt Lake City indicating no difference between groups (white and Hispanic in this case) in their expectations for pain treatment,
      • Lee W.W.
      • Burelbach A.E.
      • Fosnocht D.
      Hispanic and non-Hispanic white patient pain management expectations.
      weighing against “disempowerment” explanations holding that patients from minority cultures might be reluctant to report symptoms or seek care.
      More recently, the hope that a rising tide of better pain treatment might lift the boats of the undertreated populations received a blow from a large analysis of 13 years' worth of National Hospital Ambulatory Medical Care Survey (NHAMCS) data on EDs' opioid use
      • Pletcher M.J.
      • Kertesz S.G.
      • Kohn M.A.
      • et al.
      Trends in opioid prescribing by race/ethnicity for patients seeking care in us emergency departments.
      : although quality improvement initiatives by The Joint Commission (TJC) have coincided with dramatic general increases in opioid prescribing for pain in ED visits, the white-nonwhite treatment gap persists. “You really have to look at large data sets to see these trends,” Dr. Todd comments. “Single-site studies are something that started this series of work but are inadequate to answer what's going on.”

      Evidence on the Unconscious Level

      New methods are also bringing rigor to observations at smaller scales. A powerful instrument for detecting and overcoming bias, developed at Yale in 1998 on a demonstration basis, has become a standard component of the social psychology repertoire. The online Implicit Association Test (IAT) measures subjects' reaction times in making connections between words or images with positive or negative associations and faces or names of people representing different races, nationalities, ages, sexes, body types, and other categories, along with other objects of interest such as religious symbols and US presidents. The test asks subjects to give the same keystroke response to pleasant terms or images (eg, joy, flowers, food) and a series of people from one group, and a different handed response to unpleasant ones (war, vomit, insects) and people from another group; during further iterations of the test, the group associations are reversed. The system randomizes keyboard positions and sequences to minimize artifacts.
      The IAT builds on the tendency for certain words to be commonly associated with other words: lightning with thunder, day with night, salt with pepper, and so forth. Subjects connect such pairs faster than pairs of unrelated terms: pairing “lightning” and “pepper,” say, is confusing and requires more deliberation than the intuitive response involved in pairing “lighting” with “thunder.” A subject who recurrently takes longer to associate one group with positive concepts than another group, or more quickly associates negative concepts with that first group, implicitly finds a positive impression of the first group nonintuitive and is correspondingly likely to have an unconscious bias against them. (The test works whether subjects follow association instructions accurately or make mistakes; hesitation time is the measured variable.)
      Aided by a 2003 National Institute of Mental Health grant and now administered by Project Implicit (https://implicit.harvard.edu/implicit/), a collaborative virtual network of researchers and technicians hosted at Harvard, the IAT has undergone extensive methodologic critique, revision, and expansion into an ongoing, worldwide series of studies. Hundreds of research publications have used variants of the IAT, and data from some 6 million instances of the test are now in Project Implicit's collection.

      Banaji M, Greenwald A. The Implicit Association Test: A talk with Mahzarin Banaji and Anthony Greenwald. Edge, Feb. 12, 2008. Available at: http://www.edge.org/3rd_culture/banaji_greenwald08/banaji_greenwald08_index.html. Accessed March 24, 2009.

      Researchers Mahzarin Banaji, PhD (now at Harvard), Anthony Greenwald, PhD (University of Washington), and Brian Nosek, PhD (University of Virginia) have found that the test can not only elicit bias that subjects were unaware of—their own personal biases included
      • Vedantam S.
      See no bias.
      —but also predict conversational discomfort, political opinions, criminal sentencing decisions, and social fear responses (corroborated by amygdala activity on functional MRI testing). Disturbingly, it shows that members of minority groups have internalized some of the same distorted assumptions as majorities. And in one application of particular interest within the medical community, the test has connected unconscious biases to clinical decisions in an emergency setting.
      • Green A.R.
      • Carney D.R.
      • Pallin D.J.
      • et al.
      Implicit bias among physicians and its prediction of thrombolysis decisions for black and white patients.
      Alexander R. Green, MD, MPH, associate director of the Disparities Solutions Center at Massachusetts General Hospital and lecturer at Harvard Medical School, has developed a variant of the IAT customized for physicians, using clinical vignettes. His group found that while physicians expressed no explicit preference for white versus black patients in perceived cooperativeness, their IATs indicated otherwise, and this hidden assumption had consequences beyond personal ideation: higher white-preference IAT scores were signficantly associated with a likelihood that physicians would treat white patients, but not black patients, with thrombolysis, given identical clinical presentations suggesting an acute coronary syndrome.
      “It's pretty clear from the research now that there are links between your score on the IAT and certain differences in the way you behave,” says Dr. Green. “But there's also evidence that you can override your unconscious biases.” He links use of the IAT with the key term “solutions” in his center's name: it can educate physicians about disquieting aspects of their own thinking and help them voluntarily adjust it (along with their clinical decisions) in a more professional and progressive direction. “As we're trying to improve quality,” Dr. Green says, “we need to put special emphasis on not only general quality improvement, but focused quality improvement for those disparate populations who need it most.”
      The IAT has drawn critics, some of whom express methodologic concerns and/or find it politically ominous.
      • Arkes H.R.
      • Tetlock P.E.
      Attributions of implicit prejudice, or “Would Jesse Jackson ‘fail' the Implicit Association Test?”.
      To those inclined toward conspiracy theories, technophobia, and perhaps imaginative extrapolations from the fiction of George Orwell and Aldous Huxley, the IAT would be a useful conformity enforcing device: too many milliseconds' delay on a test of one's ideological “correctness” on certain questions, and one becomes an enemy of the people. Such an abuse would be the antithesis of the test's purpose, say its developers and proponents. In the extensive scholarly scrutiny and journalistic coverage the IAT has received to date, the researchers are careful to specify that the unconscious preferences it detects and quantifies are not the same thing as racist actions, and that the IAT should never be used for social sorting functions such as screening job applicants or jurors. (Banaji told the Washington Post that she would testify in court against anyone who tried to apply it for such purposes.) Instead of serving as a purportedly objective prejudice meter that might somehow separate legitimate from illegitimate beliefs–a science fiction misapplication that would lie well outside both the test's capacities and any claims its developers have made–the IAT best functions as an educational tool augmenting the conscious mind's control over unconscious components. (Malcolm Gladwell, author of Blink,
      • Gladwell M.
      Blink: The Power of Thinking Without Thinking.
      noted that if someone's IAT indicates a white-over-black bias, it's possible to improve the score by thinking of admired Africans and African-Americans like Nelson Mandela or Martin Luther King before retaking it.)
      Green and colleagues received a detailed critique of their thrombolysis study from prominent opponents of the test
      • Dawson N.V.
      • Arkes H.R.
      Implicit bias among physicians.
      and responded to their critics on grounds of clinical realism, statistical integrity, and parsimony of claims. “Most colleagues were surprised at the attack,” Dr. Green says. “I think [the opponents'] fear is that this IAT methodology will become something that gets widely disseminated and used without a lot of evidence behind it. The truth is there's a tremendous amount of evidence, and no one is trying to do anything with it that is beyond its capability so far.”
      For the purposes of improving ED care, Green believes it is healthier to use, study, calibrate, and continue to refine the IAT than to eschew it because of potential abuses. “There are a lot of groups around the country that are using this methodology now,” he says, “and it won't be long before there'll be a lot more studies backing up the results we showed—or maybe not. Maybe they'll show that ours were specific to thrombolysis, or that in fact the results weren't quite as strong as what we found. We don't know what's going to happen yet, but certainly it's worth looking at.”

      Cultural Competence as an Aspect of Quality

      The physician who is aware of the problem of disparate treatment, say informed commentators, can prepare for more balanced practice well in advance and apply that knowledge at key points in clinical encounters. Some anti-bias procedures are matters of common sense, like having interpreters available to overcome language barriers. Others may require considerable cognitive effort, as when a patient's adherence to spiritualist beliefs or folk treatments conflicts with the scientific paradigms of Western medicine. Patients may communicate better and view physicians with more trust if they are seen by someone from their own culture. The unifying theme is to develop a broad cultural competence appropriate to the populations seen in one's practice and to cultivate respect and empathy for patients instead of exercising top-down paternalism, regardless of differences that may clash with one's own beliefs.
      Sheryl Heron, MD, MPH, associate professor and associate residency director at Emory's Department of Emergency Medicine, finds that medical education is the optimal point for developing healthier attitudes. The academic medical environment, often chronically short on demographic balance (and, in certain institutions, nearly devoid of people of color), includes many people who “want to do the right thing” but don't know where to begin. Addressing colleagues and giving grand rounds in various settings, she often encounters residents who are the only member of their group present, are seeking a position, and don't want to rock boats, but face myriad subtle recruitment barriers. Better to mention ethnic concerns head-on, she believes, than to hem and haw about them: “Saying nothing is not the right idea.” She recommends cultivating an atmosphere that's overtly welcoming toward colleagues, staff, applicants, visitors, and patients alike. “I'm from Jamaica,” she adds, “and if I go somewhere and there are no Jamaicans or people of color, I would want somebody at least to tell me where I can get some Jamaican food.” (In some hospitals she has visited, the best source for this information is the nonphysician staff.)
      Dr. Heron has also used a simple “walk test” in group settings such as residency retreats to elicit awareness of the range of backgrounds and experiences they may encounter in practice. Residents stand on one side of the room, listen to a series of statements, and silently walk across the room to face colleagues briefly each time a statement is applicable to them personally. Asking them questions that move sequentially from the commonplace to the rarer and more troubling (“Have you ever given blood?” “Have you seen or heard racial slurs?” “Do you know anyone exposed to HIV?” “Is there anyone in your family with mental illness?”), having them “walk to” the questions alongside colleagues, and then debriefing them after the exercise, Heron has found that this self-selection process in the presence of colleagues is “a powerful exercise that enables people to walk in somebody else's shoes.” The humility and honesty that she sees in their responses, she believes, have constructive effects on the care they will deliver later in their careers.
      Working with colleagues at several institutions, she has also developed a publicly accessible online cultural competence monograph with guidelines and pertinent case studies.
      • Heron S.
      • Kazzi A.
      • Martin M.L.
      Monograph on cultural competency.
      Here, clinical narratives involving unfamiliar phenomena such as cao gio or “coining” (a Vietnamese traditional healers' practice of treating children for fever by rubbing hot coins and warm oil on the back) provide background information that can make the difference between helpful communication and, in this case, an unwarranted call to child protective services over harmless skin marks interpreted as signs of abuse.
      Dr. Green finds that a systems approach, using clinical guidelines “that make bias not an option,” can be useful. “You obviously don't want to take control completely away from the physician,” he adds, “because there are a lot of situations that algorithms don't apply to, but the physician has to be responsible for checking the algorithms to make sure they're working for that particular patient. They could override them.” This orderly approach can often be a time-saver, he adds, opening up scarce minutes for discussions with patients.
      Discharge instructions that account for cultural differences, says Dr. Green, are particularly important. A patient treated for congestive heart failure may be fully aware of the need to banish salt from his diet, for example, but if his wife isn't, and cooks him a pot of salty soup, he'll be back in the ED with fluid overload. Knowing about cuisine and other everyday practices in some detail makes communications more effective.
      In the eyes of some members of the African-American community, the US medical establishment may never entirely escape the shadow of the Tuskegee experiments. The belief that an institution is conducting secret and unethical research on groups of patients, Dr. Lovell finds, remains “pretty pervasive, and obviously really discouraging to run up against.” Counterfactual and counterproductive narratives, she says, can be extremely hard to debunk: “In the last couple of months I was speaking with someone who told me he was convinced there was a cure for AIDS and that the government wouldn't release [it] because that would mean a loss of revenue” for drug manufacturers. The short-term encounters that emergency physicians have with many patients are rarely conducive to frankness when patients view a hospital with suspicion, she notes, but trusted community leaders outside the institution can serve as valuable allies.
      “Our specialty inhabits an interesting moral space within the US health care system,” says Dr. Todd. “We're there all the time, we are not supposed to be checking wallets… and I think we have, with the public, a white hat.” Some of the chronic pain patients he encounters in his role on the board of the American Chronic Pain Association have offered extremely positive views of emergency physicians' reliability and respect for patients, he finds. That moral authority creates an opportunity to make headway against both the reality and the perceptions of bias.
      “Emergency physicians look at this, and they go through sort of the Kübler-Ross stages,” he adds. “First it's denial, then it's anger, then it's bargaining: ‘Well, maybe my institution does this.' And then there's a certain amount of depression that sets in, perhaps, and then there's some acceptance. I do think we need more leadership in this area, and honest discussions. I do think that equity is a part of quality…. Part of being a professional is to accept the fact that these judgments exist; they're part of being human.” The empathy that helps physicians overcome them, Dr. Todd and others are convinced, is a teachable skill–even if it's often “one of the things we train out of med students.”
      • Hojat M.
      Empathy in Patient Care: Antecedents, Development, Measurement, and Outcomes.

      Two-Way Traffic in Bias

      Empathizing with certain patients, admittedly, calls for more than the customary level of aplomb. Dr. Heron recalls several encounters where patients' own prejudices threatened to derail clinical procedures. One man being evaluated for chest pain stopped the exam to tell her, “You know, you're one of the smartest black women I've ever met; you really are pretty articulate.” She had to excuse herself from the room just long enough for some healthy wall-kicking, then returned to complete the workup and politely educate the patient about the achievements of a group of people he'd simply had little exposure to.
      Another recently tried to be complimentary by telling a resident, “Isn't it great we've got a nigger in the White House?” When Dr. Heron saw this patient (homeless, alcoholic, and white), she avoided bringing up the slur but talked with him for 30 minutes about their common humanity in terms of transfusions and blood types. “The guy actually cried and said ‘You are the nicest doctor I've ever met; no one's ever spoken with me with such a degree of care.' He shook my hand, and the race card just came out of the situation.”
      A third, a trauma patient sporting a swastika tattoo, declared belligerently, “I don't want to be treated by no niggers.” With a department full of distraught residents, Dr. Heron had to defuse this situation authoritatively. “I simply went into the room and told the gentleman, ‘Sir, we treat everyone the same. Notwithstanding what we may or may not believe, we will not accept unprofessionalism, rudeness, or disparaging remarks. It is up to you to determine whether or not you're going to be treated by one of us here, and you are welcome to leave if you are not comfortable…. We are who you get. You take it or leave it.'” He paused, looked at her momentarily, stayed, had his injuries treated, and learned something.

      References

      1. “Kelly” (pseudonym for Ward Sutton). A funky dance to the year that was (cartoon). The Onion, December 15, 2008. Available at: http://www.theonion.com/content/cartoon/dec-15-2008. Accessed March 24, 2009.

        • Cone D.C.
        • Richardson L.D.
        • Todd K.H.
        • et al.
        Health care disparities in emergency medicine.
        Acad Emerg Med. 2003; 10: 1176-1183
        • Pezzin L.E.
        • Keyl P.M.
        • Green G.B.
        Disparities in the emergency department evaluation of chest pain patients.
        Acad Emerg Med. 2007; 14: 149-156
        • Baker D.W.
        • Stevens C.D.
        • Brook R.H.
        Determinants of emergency department use: Are race and ethnicity important?.
        Ann Emerg Med. 1996; 28: 677-682
        • Todd K.H.
        • Samaroo N.
        • Hoffman J.R.
        Ethnicity as a risk factor for inadequate emergency department analgesia.
        JAMA. 1993; 269: 1537-1539
      2. Smedley B.D. Stith A.Y. Nelson A.R. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Institute of Medicine, Washington, DC2003
        • Task Force on the Core Content for Emergency Medicine Revision
        Core content for emergency medicine.
        Ann Emerg Med. 1997; 29: 792-811
      3. National Healthcare Disparities Report.
        (AHRQ Publication No. 08-0041) Agency for Healthcare Research and Quality, Rockville, Md2008 (Available at: http://www.ahrq.gov/qual/qrdr07.htm. Accessed March 24, 2009)
        • Zatzick D.F.
        • Dimsdale J.E.
        Cultural variations in response to painful stimuli.
        Psychosom Med. 1990; 52: 544-557
        • Greenwald H.P.
        Interethnic differences in pain perception.
        Pain. 1991; 44: 157-163
        • Todd K.H.
        • Lee T.
        • Hoffman J.R.
        The effect of ethnicity on physician estimates of pain severity in patients with isolated extremity trauma.
        JAMA. 1994; 271: 925-928
        • Todd K.H.
        • Deaton C.
        • D'Adamo A.P.
        • et al.
        Ethnicity and analgesic practice.
        Ann Emerg Med. 2000; 35: 11-16
        • Lee W.W.
        • Burelbach A.E.
        • Fosnocht D.
        Hispanic and non-Hispanic white patient pain management expectations.
        Am J Emerg Med. 2001; 19: 549-550
        • Pletcher M.J.
        • Kertesz S.G.
        • Kohn M.A.
        • et al.
        Trends in opioid prescribing by race/ethnicity for patients seeking care in us emergency departments.
        JAMA. 2008; 299: 70-78
      4. Banaji M, Greenwald A. The Implicit Association Test: A talk with Mahzarin Banaji and Anthony Greenwald. Edge, Feb. 12, 2008. Available at: http://www.edge.org/3rd_culture/banaji_greenwald08/banaji_greenwald08_index.html. Accessed March 24, 2009.

        • Vedantam S.
        See no bias.
        in: Washington Post. 2005: W12 (Jan. 23)
        • Green A.R.
        • Carney D.R.
        • Pallin D.J.
        • et al.
        Implicit bias among physicians and its prediction of thrombolysis decisions for black and white patients.
        J Gen Int Med. 2007; 22: 1231-1238
        • Arkes H.R.
        • Tetlock P.E.
        Attributions of implicit prejudice, or “Would Jesse Jackson ‘fail' the Implicit Association Test?”.
        Psychological Inquiry. 2004; 15: 257-278
        • Gladwell M.
        Blink: The Power of Thinking Without Thinking.
        Little, Brown, NY2005
        • Dawson N.V.
        • Arkes H.R.
        Implicit bias among physicians.
        J Gen Int Med. 2008; 24 (letter) (Response by Green AR: J Gen Int Med. 2008; 24: 141): 137-140
        • Heron S.
        • Kazzi A.
        • Martin M.L.
        Monograph on cultural competency.
        University of Virginia Health System, Charlottesville, VA1998-2008 (Available at: http://www.med-ed.virginia.edu/courses/culture/. Accessed March 24, 2009)
        • Hojat M.
        Empathy in Patient Care: Antecedents, Development, Measurement, and Outcomes.
        Springer-Verlag, NY2006