How I Learned to Treat My Bias

At our hospital in Tennessee not long ago, I saw my picture on the hallway message board alongside those of other doctors in a display thanking us for our service. My Asian-Indian complexion set me apart -- it's something that I am rarely conscious about in everyday life. It got me thinking: When I walk into the room, do my patients see me as a foreigner?

Then I wondered: When I walk into a room, how do I see my patients?

For the next few days I observed myself whenever I entered a hospital room to see a new patient. To my surprise, I realized that in the initial glance I viewed patients as an "elderly black man" or a "Hispanic worker" -- and all the baggage that comes with their race, gender and ethnicity. My prejudices had kicked in.

Unfortunately, the entire health system sees patients by race, gender and ethnicity, and it has a profound effect on how care is delivered.

The Institute of Medicine in its 2002 report "Unequal Treatment" cited some provocative statistics. Black patients, for example, tend to receive lower-quality care for cancer, heart disease, HIV, diabetes and other illnesses. Black men are 40 percent more likely to die of cancer than white men. These differences often persist even after accounting for age, severity of illness and delays in seeking treatment among different groups.

How can this happen in America in 2007? It's simple. Social psychology shows that stereotyping is a universal human mental function. We use social groups (race, sex and ethnicity) to understand people -- to gather or recall information about people from our minds.

The mental processing goes something like this:

When I enter the room in which a patient is waiting for me, I do four things.

First, in the seconds before our initial greeting, I automatically and often unconsciously activate my stereotype. Thus, I assume a young Hispanic man is likely to be an uninsured construction worker.

Second, even though I believe that I do not judge people based on stereotypes, the data show it is very likely that I do. When I see an elderly black woman I am more likely to ask her about church as a support structure than I am to ask a white man the same question because I assume she is church-going.

Third, after the encounter, my stereotyping affects how I recall and process information. A white man complaining of pain receives more attention than a Hispanic woman with the same complaint because I stereotype white men as being more stoic.

(Remember that stereotyping is different from medical profiling based on disease epidemiology. A young black woman with anemia is more likely to have sickle cell disease than an elderly white man is, based on biology and racial background.)