Unjust and Unwell

This article originally appeared on PSMag.com on Dec. 16, 2014.

By Paloma Toledo

Sometimes a tragedy needs to be made public in order to enact change.

Social media allowed us to see Eric Garner, a black man in New York City, die at the hands of a white police officer as he uttered, “I can’t breathe.” Protesters all over the country are decrying the grand jury’s decision not to indict the officer. These protests are shining a light on critical disparities in the American police force.

The protesters want justice and equality, but just as bias is evident in the criminal justice system, it is evident in medicine as well.

There are widespread racial and ethnic disparities in health care. While there are many possible causes, including lack of access to timely care and lack of education, bias among health care professionals may be one of the reasons health outcomes vary. Studies suggest that physicians and health care workers may be biased, and their biases may affect the decisions they make when caring for patients.

While studying attitudes about race and racism is challenging in a clinical setting, some studies have suggested that a patient’s race does affect how doctors treat them. Bias typically appears in two forms—overt bias and subconscious or so-called implicit bias toward one group over another. The Implicit Association Test is commonly used to measure implicit bias. In this computer-based test, the user is asked to sort a series of words, and a measure of bias is generated based on reaction times.

To date, this test has been taken by more than one million people. In a racially and ethnically diverse sample of test takers from the general population, nearly 70 percent of people had an implicit pro-white bias.

In 2009, investigators from the University of Washington evaluated data for test takers who identified themselves as physicians. Similar to the general population, a pro-white bias existed among the 2,535 physicians who had taken the test. Medical students and resident physicians have also been shown to have a pro-white bias.

In a now classic study published in 1999 in the New England Journal of Medicine, primary care physicians were asked to evaluate two patients presenting with chest pain and to offer treatment recommendations. The physicians were shown photographs of their patients, with the only difference being their race and gender. Researchers found that female and black patients were less likely than men and whites to be referred for what is considered the best treatment (cardiac catheterization, in this case).

In another study published recently by investigators at Harvard University, physicians were asked to manage a hypothetical patient presenting with chest pain. As they read the symptoms, the physicians saw a picture of either a black or a white patient. Bias was measured using the Implicit Association Test, and while there was no apparent explicit race preference, a pro-white implicit bias existed. This bias affected treatment recommendations. As the degree of pro-white bias increased, the likelihood of treating the black patient decreased.

As a physician, I find it hard to imagine that other physicians would make treatment decisions based on a patient’s race, but bias has other insidious effects. It can affect expectations about patients and their health care choices.

For example, I am an obstetric anesthesiologist. Patients can choose to use an epidural to manage labor pain. Studies have shown that minority women are less likely to use epidurals than non-minority white women. I have heard physicians say that if they think a patient is likely to refuse a treatment, they do not offer as much detail about that treatment as they might if they think patients are likely to want the treatment.

Whether this is efficiency or bias is not known, but it has the potential to affect treatment and possibly health outcomes, specifically if physicians are counseling patients differently based on their race or ethnicity.

The Institute of Medicine has recommended diversifying the physician workforce as one way to reduce racial and ethnic disparities in care, as the demographics of the current physician workforce do not reflect the demographics of the general population.

According to the most recent census, 30 percent of the U.S. population is either Hispanic or black, yet only 12 percent of physicians are. In an outpatient setting, a patient can look for a doctor of the same race/ethnicity as themselves, but in situations such as labor, women don’t get to choose the race or ethnicity of the anesthesiologist on call.

So what can be done?

It’s time we talk about race in medicine. Talking about race is uncomfortable. Few doctors want to think of themselves as biased, or that their colleagues are biased. But unless we address the issue, change will not occur.

Physicians often tend to assume that patients are responsible for health care disparities, because they choose not to seek treatment, or they choose not to take pain medications. But we too may be part of the problem. Physicians must begin to talk more openly about how we, consciously or unconsciously, contribute to health care disparities. We need to learn about how our biases may be affecting the care we give.

No one is going to post a video of bias occurring in the clinical setting. Medical tragedies rarely leave the walls of the hospital. But just as we are demanding equality in the criminal justice system, we should demand equal treatment for our patients in order to end health care disparities.

- Paloma Toledo is an assistant professor of medicine at Northwestern University.