According to a small University of Pittsburgh School of Medicine trial, published in the January issue of The Journal of Pain and Symptom Management, doctors give more compassionate nonverbal cues when treating seriously ill white patients than black patients.

African Americans are about 1.5 times more likely to say they would want more life-sustaining measures than white patients. But they are two to three times more likely to receive end-of-life treatment. The study was conducted to examine this disparity.

"So there's kind of a gap. We know they're a little more likely to want it, but they're a lot more likely to get it," says the study's senior author Amber Barnato, associate professor of clinical and translational medicine. "One of the things I wondered was, is there something about the way doctors interact with black and white patients that might contribute to this observation."

For the trial, researchers conducted a simulation where 33 attending physicians interacted with pretend patients. The actors in the simulation were diagnosed with terminal metastatic cancer and all read the same script. Doctors were not told the study was about race. 

"A simple hypothesis I had was, what if doctors just assume that black patients want life support so they don't ask about it," says Barnato. "That would be a statistical discrimination problem. They're basically saying, 'I think I've noticed in my lifetime that blacks are more likely to have life support, so I'm going to guess this next one wants it.' But that's not at all what we found."

Instead, the researchers found that the doctors had the same verbal communication with each patient whether they were black or white. However, they noticed that there was a difference in nonverbal communication.

"With the white patients, they were more likely to stand right at the patient's bedside and touch them in a sympathetic manner than when it was a black patient," says Barnato. "If body language is a significant tool for building trust or rapport, then if they're standing further away, that could lead to this cascading misunderstanding that could result in the patient and their family not trusting their doctor."

Barnato says the racial bias demonstrated in the study could play a role in the difference between how likely black and white patients are to receive life-sustaining measures. But it's not unique to end of life care or even the medical profession, where racial bias has long been found in doctors interacting with patients who are of a different race than their own.

"No matter how good we are as people, and how thoughtful we are, we all have implicit biases. And the thing about implicit biases is they operate really fast. They're not under our conscious control," says Barnato. "And that's why there's been all this research done trying to understand why police are more likely to shoot unarmed black people than they are white people."

While Barnato says the problem is complex, she says there are steps that could be taken in the medical profession to address implicit bias. And these kinds of practices could be used in other fields as well.

"What we could do is we could train doctors to have some intention. This is like a mindfulness process. It takes a habitual breaking of these fast unconscious behaviors," says Barnato. "These are things that individual physicians can do. But the first thing they have to do is accept the possibility that they would [carry a bias]."