Us and Them: Breaking Down the Barriers

Dr Carmen Lawrence

Paper presented at Fulbright Conference: Healthy People Prosperous Country

July 11, 2008

It is not only bad people who are prejudiced, that would not have such a strong effect. Most people would not wish to imitate them – and so, such prejudices would not have much effect- except in exceptional times. It is the prejudices of good people that are so dangerous.

Vikram Seth, A Suitable Boy. London: Phoenix, 1993.

Introduction

Among the most universal and persistent inequalities in health and well-being within developed countries are those based on race and ethnic group membership. In Australia, the extreme health disadvantage suffered by Indigenous people continues despite sporadic attempts by governments to address what are seen as the underlying causes and heroic efforts by many service providers to deliver appropriate health care.

Explanations for these racial and ethnic disparities – here and elsewhere - range from those which point to innate group differences, for example in genetic susceptibility to disease, to those which implicate the history of dispossession and discrimination to which such groups have been exposed. Poverty, and the diminished living conditions associated with it, is frequently invoked as a major causal factor. In particular, reference is often made to increased exposure to environmental hazards, poorer diet, overcrowded housing and the effects of lower education levels on health literacy and health related attitudes and behaviours (e.g. smoking).

Racial differences in health care

The sneaking suspicion that one of the key factors contributingto the poorer health status of people from racial and ethnic minoritiesis the fact that they actuallyreceive a poorer quality of health care because of their race or ethnicity did not, until fairly recently provoke sustained investigation or analysis. But in 1999, repeated exposure in the media led the U.S. Congress to commission the Institute of Medicine (IOM) to examine the question of whether and to what extent racial and ethnic minorities received a lower quality of healthcare.

The initial assumption was that such differences as had been observed were likely to be the result of access related factors such as insurance status, ability to pay and the availability of health services. But theIOM was also asked to examine whether there was any reason to believe that overt or subtle bias or prejudice[1] on the part of healthcare providers might also be affecting the quality of care. They were also requested to suggest intervention strategies to remedy whatever deficiencies were identified.

They reviewed over 100 studies[2] which had examined the quality of healthcare for various racial and ethnic groups while holding constant variations in things like insurance status, patient income and other access-related factors. Some had also controlled for potential confounding factors such as the severity of disease, the presence of other illnesses, where care was received and demographic variables like age and gender.

The IOM reported that it was “struck by the consistency of findings” which showed that “minorities are less likely than whites to receive needed services, including clinically necessary procedures” (p 2). The disparitiesreported were not restricted by disease type, appearing inthe treatment of cancer, cardiovascular diseases and mental illness; nor did they very with the type of procedure, involvingmajor interventions as well as routine procedures for common health problems. Studies in the U.K have also indicated under-representation of non-whites in access to renal transplantation and a pattern of more coercive treatment in mental health facilities.[3]

Similar results have been obtained in Australian research. Despite excess mortality and morbidity amongst Indigenous people they receive less specialist healthcare both as inpatients and outside hospitals. Using National Hospital Morbidity data, Cunningham[4] showed that patients identified as Indigenous were less likely than other patients to have a major procedure recorded, even after adjusting for patient, episode and hospital characteristics. The results were also apparent for most diseases and conditions and led the author to conclude that there were “systematic differences in the treatment of patients identified as Indigenous”.

Similarly, Coory and Walsh[5] compared the rates of percutaneous coronary interventions[6] and bypass surgery after acute myocardial infarction and found that the rates were significantly loweramong Indigenous patients than among non-Indigenous patients even after adjusting for age, sex, SES status, hospital characteristics and other illnesses.Other studies have also shown differences in the application of hospital procedures, including for cancers, and inequalities in access to cardiovascular healthcare. An exploratory study[7] in W.A. reported in 2004that, in general, Aboriginal people with lung or prostrate cancer were less likely to receive a surgical procedure for their cancer than the non-Indigenous population.

Explanations

It is now generally agreed that the phenomenon of fewer and poorer health care services for racial and ethic minorities because of their status is real. The appropriate response obviously depends on the reasonsfor thesedifferences. In considering possible sources of the health care disparities the IOMreviewers first speculated that there may be subtle differences in the way racial and ethnic groups respond to treatments, since there was some evidence that African Americans were slightly more likely to reject recommended treatments. However, the size of the difference was not sufficient to explain the discrepancies observed. Nor were differences in help-seeking attitudes or treatment preferences observed

Instead the IOM Committee considered it more likely two sets of causal factors were implicated:

  • Systemic or institutional factors which systematically disadvantage racial and ethnic minorities: those relating to the operation of the health care system and the legal and regulatory climate, including cultural and linguistic barriers (e.g. lack of interpreters), fragmentation of the health care system, incentives for physicians in low cost insurance plans to limit services and the fact that minority groups are more likely to receive care in a hospital setting. This is sometimes described as institutional racism and represents the “collective failure of an organisation to provide an appropriate and professional service to people because of their colour, culture and ethnic origin.”[8] Although the health care system appears to have been the most systematically investigated and is the subject of my paper today, it is clear that such failure occurs in many areas of service provision. And we should ask, “why”?
  • Individual factorsoperating in the clinical encounter - particularly provider bias or prejudice against minorities which may take the form of clinical uncertainty when interacting with people from different racial or ethnic groups and beliefs (stereotypes) held by providers about the behaviour or health of minorities. The IOM argued that such provider behaviour might, in turn, cause patients to act in ways that contribute to the observed disparities, for example by discontinuing treatment. They concluded that although there was little direct evidence of how these factors might influence care, there was evidence suggestive of bias, prejudice and stereotyping on the part of health care providers.

In the absence of such direct evidence the IOM research team drew on mix of theory and relevant research to try to understand what might be happening. They pointed out that clinical decisions are surrounded with uncertainty and care providers, like the rest of us, bring prior beliefs based on age, sex, income and ethnicity, to help them make sense of behaviour in the clinical encounter. They pointed out that these beliefs, if based on evidence and epidemiology, may facilitate efficient practice, but if they are inaccurateracial or ethnic stereotypes which biasperception of the individual’s condition, then real harm can be done.[9]

Racial stereotyping

There is an existing and extensive literature in the social sciences which has explored the nature of stereotypes, how stereotypes evolve and persist and how they shape expectations and affect interpersonal interactions.It shows, amongst other things, that individuals who hold negative stereotypes of a group will discriminate against members of that group, treating them less favourably.

Since the IOM report, there has been an explosion of research and commentary on therole played by stereotyping and prejudice in the clinical encounter. As a result, there is now a considerable body of evidence from a variety of countries that health care providers hold stereotypes based on patient race, class, sex and other characteristics and that these stereotypes do influence their interpretation of patient communications and symptoms and, inexorably their clinical decisions, often to the detriment of the minority group member. One U.S. study found that doctors rated black patients as less intelligent, less educated, more likely to abuse drugs and alcohol, less likely to comply with medical advice, lacking in social support and less likely to participate in cardiac rehabilitation.[10]

While overt expressions of prejudice and negative racial stereotypes appear to have declined in recent decades, probably because of the greater social opprobrium attached to such views, the fact remains that unmistakable racism is still expressed and it is still a daily experience for many Indigenous Australians. A recent Australian survey by Pederson and her colleagues[11] found that racist attitudes and behaviour are still common. Fifty two percent of urban residents in W.A. and 69% in a regional community expressed prejudice against Aboriginal Australians. Further research in W.A. indicates that such attitudes translate into the experience of racially based harassment for many Aboriginal people.[12]

Even people who express explicitly egalitarian views may hold negative ethnic and racial stereotypes of which they are unconscious – and that probably includes us. Such unconscious biases- sometimes called aversive racism - may be just as destructive as those which are more overt.Few people are prepared to admit, even to themselves that they may be racially prejudiced. Harvard psychologist Banaji[13] used an ingenious approach to uncover such tendencies. She showed subjects a sequence of positively and negatively toned adjectives, each of which was then paired with a stereotypically white or black name (e.g. Meghan and Keisha). As each pair appeared, the subject was asked to press a key to register whether the adjective was good or bad. Their reaction times were measured. Banaji found that most people reacted faster when white names were paired with positive adjectives and black names were paired with negative adjectives. She suggested that there was a pre-existing link between goodness and whiteness and badness and blackness in the minds of her subjects so that their brains more quickly assimilated the information to the stereotypes. The effects were evident in those who claimed not to be racially prejudiced.

Usually, such attitudes do not result in expressions of hatred or open contempt, but rather in anxiety and discomfort, which lead to avoidance. Whites tend not to recognise when their actions are racially biased but they inevitably provide nonverbal cues which may signal negativity and produce distrust. One of the common findings in the U.S. research on prejudice is that whites often report feeling anxious while interacting with blacks, leading them to seek to avoid such occasions.

In the clinical encounter, this may result in White doctors engaging in avoidance behaviour, including averting their eyes and spending less time with the patient. The discrimination is subtle yet systematic, influencing judgments and interaction, including cues to friendliness of lack of it. Such responses are especially likely to be activated under time pressure and substantial cognitive demands, a common experience of many practitioners. If the patient is alert to signs of prejudice, then such avoidance will be interpreted as reflecting hostile attitudes and so compromise the patient-provider relationship.

Davidio et al have reported[14] that aversive racism in this more subtle and indirect form is most likely to produce discriminatory action when “situational demands are unclear or when norms for appropriate actions are weak or ambiguous”. They support this conclusion with evidence that disparities in treatment are greatest when doctors are required to exercise considerable discretion – such as recommending a test or making a referral – and least when little discretion is involved – such as emergency surgery. For example, Black women are less likely than White women to be tested for osteoporosis and less likely to receive appropriate medication, once they have been diagnosed.

A study by Schulman et al[15] found that doctors were less likely to refer black female patients - actually videotaped actors trained to display the symptoms of cardiac disease – for cardiac catheterization than white males and females or black men “exhibiting” the same symptoms. Green et al[16] presented doctors with hypothetical descriptions of cardiac patients (race was systematically varied) and examined the relationship between their racial attitudes and the treatment they recommended. While doctors reported no explicit biases, they had more implicit negative attitudes toward blacks and stronger stereotypes of blacks as unco-operative patients. The more negative the attitudes, the less likely they were to recommend appropriate drug therapy.Very often the providers are unaware that they are applying stereotypes which compromise the quality of the care they are providing.

A review[17] of studies of consultations involving minority group members concluded that such patients are less likely to engender empathic responses from doctors, less likely to receive adequate information and less likely to be engaged as a partner in medical decision making. Comparisons of videotaped recordings of White doctor, Black patient interactions with White on White interactions have also shown shorter consultations, greater verbal dominance by the doctor, the provision of less information and fewer attempts to engage the Black patient in joint decision making.[18]

Given such experiences, it is not surprising that members of minority groups have been shown to have less trust in the health care system and in health care providers than the rest of the community. This is all the more important because there is a good deal of research which shows that the worse the relationship, the poorer the recall of medical information, the poorer the adherence to recommended treatment and the poorer the health outcomes.

Added to this are the direct effects that experiences of racism and discrimination have on health and wellbeing – and most Indigenous Australians experience racism on a regular basis. A recent review[19] of 138 empirical studies of the association between self-reported racism and health showed that the most consistent findings related to negative mental health outcomes and health-related behaviours such as smoking and alcohol consumption. Furthermore, the longitudinal studies indicate that the experience of racism precedes ill health rather than vice versa. Research by Larson et al[20] examined whether the experience of interpersonal racism had a measureable effect on the health of Aboriginal West Australians. They found higher rates of racially based negative treatment for Aboriginal people than for others interviewed (40% within the previous four weeks) and such experiences were significantly associated with poorer physical health and mental health.

Some researchers have suggested that the mechanism for this relationship is the same as for many chronic psychosocial stressors: changes to the neuroendocrine, autonomic and immune systems which ultimately compromise the individual’s health. Laboratory studies and some epidemiological investigations have found that perceived discrimination is associated with increases in blood pressure and hypertension. When the perceived discrimination is persistent, it appears to predict coronary artery calcification and coronary events.[21]

One of the other consequences of being subjected to pervasive negative stereotypes is that such views are actually internalised by members of the minority group. This means that those who are subjected to constant negative stereotypes come to accept as valid the dominant culture’s views about their inferiority. Researchhas shown that those who do internalize these views about themselves are more likely to consume alcohol to excess, to exhibit psychological distress and psychological problems such as low self esteem, feelings of isolation and identity crises.[22]

Although I have focused on health, this is not to condemn healthcare providers, many of whom work very hard under challenging conditions to ensure that patients get the best care possible, nor to suggest that they are alone in expressing such bias. The same tendencies almost certainly exist with other professionals who have been less extensively studied. It is certain that stereotypes and prejudice toward Aboriginal people, both overt and covert, operate in most areas of policy development and service delivery. This is nowhere more evident than in the NT intervention and the public discussion surrounding remote Indigenous communities.