By Dustyn Addington
Part II of Implicit Bias in Health Care
In Part I, we explored the concept of implicit bias. To see how the issue connects to health care, we have to start with justice.
If justice is a matter of people getting that they deserve through fair and equitable processes, then we must ask what people are getting and if there are arbitrary and harmful disparities in what they receive.
Philosopher Iris Marion Young describes the common definition of a structure as “a confluence of institutional rules and interactive routines, mobilization of resources, and physical structures.” Bringing the concept of justice to bear on structures, Young argues that “structural injustices are harms that come to people as a result of structural processes in which many people participate.” Structural injustices concern the way people are treated in a system of processes, institutions, practices, and rules.
As a complicated and high-stakes system, health care must therefore be evaluated in terms of structural injustices.
One point of increased focus must be implicit bias, due to its track record of producing systemic harms through processes and in institutions, including legal, economic, and interpersonal contexts.
Systems-level thinking is crucial here. A recurring error in the accounting method for your personal accounts affects far fewer people and far fewer dollars than an accounting error in a national chain of banks. Similarly, systems magnify the effects of implicit bias. A recurring process, like an interview process, that is unguarded against implicit bias can replicate the flawed decision-making procedure indefinitely. Biases can therefore be amplified in systems and structures (however, if properly ordered, some systems can work to minimize bias, like scientific processes).
Several studies have shown the impact of implicit bias on health care systems. Here I focus on three categories of effect: the impact on doctors, the effect on screening and diagnosis, and the resulting disparities in treatment.
Impact on Doctors
One might hope that because doctors are well-educated, scientifically-minded, and have well-developed reasoning skills, that they are immune to the effects of implicit bias. Unfortunately, the research is clear: doctors are vulnerable to implicit bias. One study showed that even with doctors who had “reported no explicit preference for white versus black patients,” a difference in patient perception and treatment remained. Using the Implicit Association Test and evaluating the rate of treatment of thrombolysis, the study reports: “As physicians’ prowhite implicit bias increased, so did their likelihood of treating white patients and not treating black patients with thrombolysis.” It does not stop there—even bedside manner can be deteriorated by implicit bias. These doctors are not racists in the sense that they hold racist views explicitly. However, they implicitly hold prejudicial associations that affect their judgments, decisions, and behaviors. They are not alone: the research appears to show that we are all subject to implicit bias.
Screening and Diagnosis
Implicit bias acting at the level of screening and diagnosis means that the ailments of some will be detected and some will not, for no other reason than their gender/sex, race/ethnicity, or belonging to another social category.
Examples are numerous: “Black patients with chest pain are referred for advanced cardiac care less often than white patients with identical symptoms.” “Asian Americans had 34 percent lower odds of being screened [for diabetes] than whites.” Even evaluations of patients’ ability to “stay still and calm” during an examination are subject to implicit bias.
Reductions in screening and appropriate diagnoses due to implicit bias mean that treatment can be delayed or never occur, making health problems endure, worsen, or even become fatal when earlier diagnosis and treatment would have improved the patient’s health. This is a kind of harm, one that is avoidable and unfair, making health differences due to implicit bias fall squarely under the definition of health disparities.
Not only are non-white patients diagnosed more irregularly, but the treatment received is often different from that received by white patients. Below is a selection of findings from recent research on implicit bias and treatment.
- “Non-white patients receive fewer cardiovascular interventions and fewer renal transplants.”
- “As physicians’ prowhite implicit bias increased, so did their likelihood of treating white patients and not treating black patients with thrombolysis.”
- “Black women are more likely to die after being diagnosed with breast cancer.”
- “Non-white patients are less likely to be prescribed pain medications (non-narcotic and narcotic).”
- “Black men are less likely to receive chemotherapy and radiation therapy for prostate cancer and more likely to have testicle(s) removed.”
Differences in treatment due to implicit bias produces an increase in suffering and mortality. It is of small consolation that the harm is unintended. Implicit biases produce structural injustices by systematically harming groups of individuals engaging in the processes and institutions of health systems.
The effects of implicit bias are pernicious, subtle, and harmful. Implicit bias infects many areas of health systems; the ones listed above are but a few. Differing admission rates to medical school, the effect of hiring rates and wage differences on the ability to afford healthcare, and the violence resulting from the biased judgments of police officers are of substantial consequence for public health as well. Changes in norms, processes, and systems are needed. In Part III, I explore empirically-backed strategies for mitigating implicit bias.
Dustyn Addington is the Assistant Director of Learning and Strategy at the Foundation for Healthy Generations. He is also a graduate student in the Department of Philosophy at the University of Washington, researching bias, knowledge, and judgments.