Health Care
Implicit Bias Among Physicians
A second race-bias study with important findings r is Alexander R. Green’s Implicit Bias Among Physicians and its Predictions of Thrombolysis Decisions for Black and White Patients. This study was the first to apply the IAT to issues of health care using actual health care providers rather than college students. The question was simple but crucially important: does implicit bias leads doctors to make different health care decisions based on the race of a patient?
The researchers found that explicit racial bias was irrelevant to medical decisions – but that implicit bias can directly influence the professional decisions they make. After evaluating doctor’s attitudes as measured by the IAT, Green discovered that as the degree for antiblack bias on the race preference IAT increased, certain treatment recommendations for black patients decreased.
This study had two phases. In the first phase, subjects (all of whom were doctors) took three separate IAT tests. One was a basic Race-IAT, one was a Race Cooperativeness IAT and one was a Race Medical Cooperativeness IAT. In the first test, reaction times were measured using associations between positive and negative words and racial representations. In the second test, the positive and negative words were specifically chosen to be about cooperativeness (“easygoing” vs. “stubborn” etc.) while for the third test, the cooperativeness language was narrowed down further to focus on cooperativeness with medical recommendations.
In the second phase, doctors were given a hypothetical “patient” complaining of symptoms of coronary artery disease (CAD). The patient information appeared on a computer screen along with a randomly generated image of either a white or black person. The pictures used were designed to appear friendly and were created by blending together stock photo images of different people to make sure that no image corresponded to a real person. The physicians were asked to “rate the likelihood that the chest pain was because of CAD”, “whether they would give the patient thrombolysis” (a pharmacological treatment of CAD) and the “strength of their recommendation”, all on a scale of 1-5. The physicians were also asked about their explicit biases towards African American patients. Thrombolysis was specifically chosen becausetreatment with medication (unlike surgery) requires cooperative, dependable patients who can be relied upon to follow a doctor’s instructions
With regard to explicit bias, doctors showed equal preference for both White and Black Americans and also perceived both White and Black patients as equally cooperative with medical procedures. The IAT, however, showed consistently strong anti-Black associations across all three IATs. As for the hypothetical chest-pain sufferers, “physicians were more likely to diagnose black patients… than white patients… with CAD”, but recommended thrombolysis for the same percentage of patients of each racial category. In other words, black patients were less likely to receive a recommendation for thrombolysis than white patients, as more black patients were diagnosed with CAD.
More importantly, Dr. Green and his associates determined that physicians self-reported feelings about African Americans and their cooperativeness had little-to-no predictive value. Going by explicit measurements only, we may conclude that a doctor’s level of racial bias has no influence on the actual decisions they make. However, the story here is that a doctor’s level of implicit bias can directly influence the professional decisions they make. Looking at the IAT, Green discovered “as the degree for antiblack bias on the race preference IAT increased, recommendations for thrombolysis for black patients decreased”.
A important side note for future research: a very small portion of the doctors knew why they were being tested and what the end research goals of the study were. Those doctors actually prescribed thrombolysis more often to black patients than white patients, perhaps due to self-conscious overcompensation.
