We like to think that physicians make diagnoses and prescribe treatments based on empirical observations of symptoms. Patients should be treated based their condition, not based on their personality, or worse, their race or socioeconomic status.

Most physicians strive for this sort of objectivity, but a new study suggests that internalized stereotypes and prejudices have more influence over physician decisions than we assume.

The study was conducted by providing a case study of a 50-year-old man called "Mr. Thompson," a smoker with a history of hypertension, "who presents to the emergency department with chest pain. He appears to be in a lot of pain describing it as 'sharp, like being stabbed with a knife.' "

Some physicians were told Mr. Thompson was white, and others were told he was black. The participants were then asked to decide whether the pain was the result of coronary artery disease and whether to prescribe clot-busting drugs.

The results are surprising:

Doctors were more likely to think "Mr. Thompson" was having a heart attack when he was black than when he was white. But they did not prescribe treatment to reflect this -- physicians who thought a black Mr. Thompson was having a heart attack prescribed thrombolysis less often than when they thought a white Mr. Thompson was having one.

This may help explain some of the racial disparities in medical care, and why the death rate is 29% for African Americans and the death rate from stroke is 40% higher.

Researchers concluded that the bias wasn't the result of "harboring deliberate ill will," but rather based on internalized racial stereotypes. The good news, however, was that physicians were at least "willing to open their subconscious minds for inspection," which often isn't the case in situations like these, such as when researchers recently suggested there may be racial bias in NBA refereeing.