CONTENTS INTRODUCTION I. BACKGROUND II. THE SOCIAL SCIENCE RECORD ON IMPLICIT BIAS IN HEALTH CARE III. THE SHORTCOMINGS OF THE SYMBOLIC INTERACTIONISM PERSPECTIVE IV. TOWARD A NEW PARADIGM CONCLUSION INTRODUCTION
Recently, a group of researchers reported that physician implicit race and ethnicity biases do not affect their hypertension treatment for minority patients nor do such biases impact health outcomes for these patients. (1) These findings are counterintuitive. Moreover, they are contrary to the weight of the emerging empirical record that has suggested that physician implicit bias is inversely related to the quality of doctors' treatment decisions, communication with, and perceptions of their minority patients. Examples include findings that implicit bias affects treatment decisions for heart disease, (2) pediatric urinary tract infections, (3) and diseases stereotypically associated with minority patient groups. (4) The profound concern that this recent study raises has less to do with the danger that some may erroneously and prematurely celebrate the fact that physician bias is unrelated to the estimated 83,000 deaths of minority patients annually due to discriminatory health care. (5) Rather, the real concern is that this study will join the copious body of social science literature on implicit bias in health care, which completely overlooks the fundamental structural nature of unconscious racism and its contribution to racial and ethnic inequality in the U.S. health care system. In other words, finding that a group of physicians' implicit biases are or are not associated with inferior treatment decisions for individual patients with a single disease is not the point if eliminating racial and ethnic health inequality is the goal. The persistent health disparities phenomenon is a structural problem, and therefore implicit biases that contribute to disparities must be structurally dismantled. Moreover, the racial discrimination that causes disparities is so fundamentally associated with poor health outcomes that finding an attenuated association between bias and hypertension treatment does not alter the structurally causal relationship between bias and health disparities overall.
In this article, I sketch out the broad contours of a new theoretical approach to the problem of health disparities. I assert that unconscious racism in medicine is an avoidable and reparable injustice that requires incentive and norm-changing solutions in order to radically disrupt the context in which medicine is currently practiced and under which minority patients currently suffer. Reforming the anti-discrimination legal regime is the solution explored here, (6) but there are other structural solutions to consider that are also important to achieving health equality. For example, fixing systemic educational inequality, housing segregation, and the lack of universal health care coverage would go much farther toward equalizing health outcomes than changing discrimination laws. However, I believe that legal reform is also essential to bringing about health equality. Law has the effect of expressing and influencing shifts in social norms, which can permeate systems to affect structural change. Therefore, this discussion centers on reversing the trend toward acceptance of implicit bias as an inevitable, harmless fact of life. Put bluntly, I assert that unconscious racism produces invidious discrimination and an odious inequality that should be prohibited and punished by law. However, as long as the discussion of unconscious bias in health care continues to be framed in terms that examine only individual, cognitive contributions to the problem, the systemic solutions to the health disparities will fail to emerge. Researchers will continue to chase increasingly narrow observations about the hidden attitudes that pass stealthily between and among individual actors in the health care system, instead of pursuing the systemic resolutions for the fact that racial and ethnic discrimination at every level of health care delivery, financing, and organization, is a fundamental cause of poor health outcomes.
I submit that as a fundamental and theoretical matter, the question of whether physician bias is related to medical decisionmaking for individual diseases is far too small an inquiry. The implicit bias work by social psychologists to date has been defined and limited by a symbolic interactionism framework. This framework has permitted only de-contextualized, ahistorical, and individualized consideration of the broadly systemic and institutional problems that produce health care disparities and health inequality. In place of the individualized inquiries that have dominated the implicit bias discourse, I bring a critical theory perspective to bear on the problem of health disparities in general, and more specifically, on the question of whether individual and institutional providers' implicit biases contribute to these disparities. From this perspective, I analyze the political economy in which health care disparities occur. I apply constructs from structural violence theory to better understand the context in which physician bias operates, the structural inequality and racism that has produced this bias, and the inadequacy of cognitive and behavioral solutions alone to address it. I conclude by proposing a new theoretical construct that I call "structurally derived discrimination." I offer this construct to add a broader theoretical perspective to the implicit bias discourse. I contend that without this perspective, the U.S. health care system will never eradicate, or even meaningfully reduce, health disparities caused by unconscious racism. (7)
Racial and ethnic health disparities--the clinically unsupportable differences between health care and health outcomes experienced by minority as compared to white patients--are both deadly and financially costly. In 2005, Dr. David Satcher estimated that 83,570 deaths occur each year as a result of racial and ethnic health disparities. (8) In addition, researchers have estimated that over 30 percent of the direct medical costs that African Americans, Latinos, and Asian Americans incur are excess costs due to health inequities--amounting to over $230 billion during a three-year period. (9) Moreover, the most recent evidence suggests that most disparities are becoming worse or remain unchanged. (10)
In 2011, the U.S. Department of Health & Human Services' Agency for Healthcare Research and Quality (AHRQ) released its ninth annual report on National Healthcare Quality and Disparities. (11) A key function of the report is to describe the progress that has been made in reducing disparities in the U.S. health care system. (12) The 2011 report recorded few positive changes in disparities over the five-year period from 2002 to 2008 by racial and ethnic groups based on measures of quality, such as the number of deaths due to cancer, heart attacks, and the incidence of end stage renal disease due to diabetes. (13) By most measures, disparities in access to health care remained unchanged among African Americans, Native Americans, Latinos, and white Americans. (14) Well over 90 percent of the measures that describe disparities in the quality of health care that African Americans, Latinos, Asian Americans, and Native Americans receive as compared to whites have remained unchanged. (15) Thus, notwithstanding isolated metrics that show a narrowing health disparities gap such as overall life expectancy, for most outcome and quality measures, the efforts by health care providers, scientists, and policymakers to "eradicate disparities" has failed. The evidence continues to confirm that minorities spend more money to get inferior care, suffer poorer health outcomes, and die earlier than whites in the United States.
Perhaps the least understood aspect of racial and ethnic disparities is the role that unconscious racial attitudes (i.e. implicit biases) play in contributing to inequity in health and health care. As in much of the discourse about disparities, the current literature merely focuses on implicit bias in the individual clinical encounter. As such, social science researchers have assumed a theoretical paradigm similar to theorists who posit that health disparities are due to behavior, lifestyle, and genetic differences between races. (16) Indeed, some scholars have gone so far as to suggest that racial bias is patently irrelevant to racial and ethnic disparities in health and health care. (17) Unfortunately, behavioral and biomedical theories divert attention from the systemic injustice of discrimination in health care. Even the nascent social science literature that confirms the association between physicians' unconscious racism and health inequalities also overlooks the structural context from which unconscious racism emanates. While the researchers behind this implicit bias literature are to be commended for their work, much more is needed in order to understand and effectively combat the systemic influences of bias on health inequality. Therefore, it is important to consider the limitations of the existing social science literature on physician implicit bias in order to shed light on the true nature of the problem, and only then develop broader interventions to address it.
THE SOCIAL SCIENCE RECORD ON IMPLICIT BIAS IN HEALTH CARE
Twenty-five years of social science research confirms that implicit, anti-minority biases are pervasive among Americans generally, and among physicians in this country specifically. (18) Using implicit association test (IAT) data collected from hundreds of thousands of voluntary visitors to Harvard University's Project Implicit website, researchers concluded after analyzing data from over 2,500 test takers who self-identified as "MDs" that the tested physicians exhibited the same preferences for whites over blacks as those found among the general population....