The trusted source for
healthcare information and
Are there subconscious racial bias and myths?
A study from the Georgetown University Medical Center in Washington, DC, reports patients’ race and sex may influence a doctor’s decision about whether to refer someone for cardiac catheterization — even when other information such as descriptions of symptoms and types of insurance are the same.
The trend is most dramatic in black women, who received just 40% of the referrals of white men with identical conditions. Overall, females received 60% of the referrals given to male patients. Blacks received 60% of the referrals that white patients received.
"It’s not my belief that the difference is due to overt bias," says researcher Kevin A. Schulman, MD, of the university’s Clinical Economics Research Unit. "It’s more likely to be subconscious."
Schulman says doctors may not know specifics about particular ethnic groups and may instead operate on stereotypes. Medical schools don’t teach students about how people are raised and grow within an ethnic group, so the stereotypes are not corrected.
The problem is a familiar one, says George A. Mensah, MD, chief of cardiology and the head of cardiovascular care at the Veterans Affairs Medical Center in Augusta, GA. Medical training may be perpetuating racial myths from one generation of physicians to the next because even though accurate data gained from recent studies are available, they make their way into practice very slowly.
Some doctors may not think African-Americans are vulnerable to coronary artery disease, Mensah says. The misconception began with studies in the 1960s that compared how blacks and whites developed heart disease. Poor populations of African-Americans, who could not afford to eat a lot of fatty snack food or cheese, or buy cigarettes, were compared to more affluent whites, who could.
The black subjects’ rates of coronary heart disease were much lower, and the flawed findings went on to help shape misconceptions on how to recognize heart disease in black people. These early studies needed to control such confounding factors.
Schulman’s study "is a very good one," Mensah says, because it is designed to control many of the confounding factors that could themselves contribute to the decision whether to recommend catheterization.
More than 700 doctors attending the 1996 meeting of the American Academy of Family Practice and the 1997 meeting of the American College of Physicians were shown video footage of actors portraying patients. The patients were white or black, a man or woman, and either 55 or 70 years old. They gave identical scripted descriptions of three different types of chest pain: definite angina, possible angina, and non-anginal pain. The descriptions were approved by a panel of four cardiologists. Insurance information was the same for respective age groups as well as occupations.
Patients were considered either high- or low-risk for coronary heart disease. They were assigned characteristics mimicking the subjects found in different cardiovascular disease risk percentiles of the Framingham Study (70% to 80% for high risk and 20% to 30% for low risk).
Doctors first were asked to characterize the chest pain and estimate the chance it represented coronary disease, defined as at least a 70% blockage of a coronary artery. From there, they could make these four choices:
1. Do not get a stress test.
2. Get a regular stress test.
3. Get a stress test with thallium.
4. Order other tests like echocardiography.
If they ordered a test, one of three results was given, according to the angina groups. The doctor was asked again for the probability the patient had significant arterial disease and if he or she would send the patient for catheterization.
Schulman says it was essential to show the doctors actual footage of the patients. Presenting a situation similar to an office visit triggered thought patterns that may not have happened if the practitioner was given written pages to read.
"Our whole hypothesis was that this is something visual," he says.
"Really, what it takes to fix the issue of bias on recommendation is you have to have dialog going on across the country," Schulman says. This dialog needs to discuss what biases are at work in making clinical decisions as well as what patients should expect from their doctors during an office visit. Identifying the bias and replacing it with information that supports better decision making, however, probably won’t be a comfortable process.
"Physicians never think they are biased," says Gail L. Daumit, MD, a researcher from Johns Hopkins Medical Institutions in Baltimore who studies racial differences in access to cardiac care among kidney patients. "Sometimes it takes a study to open their eyes. That can be a painful thing to do, but it’s important."
"Unfortunately, until it’s fixed, I think patients have to be more aggressive," Schulman says. Minorities should not feel intimidated about telling doctors about chest pain, he says, and should be more insistent about getting the right care. If patients feel their complaints are not being taken seriously, they should seek a second opinion.
Doctors in the study were internists and family physicians. Schulman says while this study can’t be extrapolated to cardiologists, he would expect specialists to show similar recommendation patterns as the generalists. "They all went to the same medical schools," he says. "They all trained together."
1. Schulman, et al. The effect of race and sex on physicians’ recommendations for cardiac catheterization. New Engl J Med 1999; 340(8):618-625.