Continuing on in my research for the 2nd Annual LGBTQI Health Symposium (see these posts to follow my progress), this is the first paper that connects unconscious (or “implicit”) bias to poorer health care. There are now 99 other papers that cite it, so it’s an area that’s gained a lot of attention since 2007.
Unconscious bias and how it’s measured
Implicit (unconscious) and explicit bias are differentiated by the fact that a person who may not have explicit bias (i.e. you ask them if they have a preference or positive feelings for a group different from them and they say they do not) may have implicit bias, as measured by reaction time using a special test called the Implicit Association Test (IAT). The theory is that when a person views images paired with either positive or negative descriptions, their response time changes. If their brains connect an image with a positive description, like “good,” “trustworthy,” etc, faster than they are more likely to associate those words with that image. Comparing response times gives a measure of implicit bias, or an unconscious tendency to favor one group over another.
You can take a series of these tests yourself at the Harvard Project Implicit web site. Several different types of bias are tested for. There is currently no test for gender identity bias (when will there be?).
There’s an enormous body of literature on this test and its validation, which I won’t go into here, because the attached paper shows the impact and connection to health care nicely.
Study: Greater unconscious bias results in care inequality
Implicit bias may not be recognized by the person who has it, but it may be recognizable in their behavior, maybe by themselves on reflection, or often by the people around them who are subject to the bias.
For example, members of an unfavored group may be counseled differently about health interventions or not offered them at all. Or care that is often denied altogether is not seen as abnormal by a clinician who denies the care, think the early days of HIV, or today’s transgender person health care.
This study takes a look at this issue via the recommendation to offer thrombolysis (clot dissolving) to a standardized patient who is either white or black (they viewed an image while reading the story). Resident physicians were asked first to diagnose coronary artery disease and then whether they would recommend thrombolysis. They were also asked to take the IAT discussed above and a test of explicit bias, and comparisons were made, with very interesting results:
- Resident physicians as a group displayed very little explicit bias – eg beliefs about ability to cooperate in care were stated as equal for white and black patients
- Resident physicians displayed significant implicit bias toward (against) black patients, if they were white, latino, or asian/pacific islander
- Resident physicians did not display implicit bias against black patients if they were African American
- The black patients were more likely to be diagnosed with CAD in general, but offered thrombolysis the same amount as white patients, in other words, they were less likely to be offered thrombolysis
- This association was correlated with implicit bias – the more implicit bias in the resident the more they were likely to not recommend treatment, even when they did not display explicit bias
See Figure 2 in the article for a comparison of the measured bias, and the more-confusing Figure 3, which shows that as implicit bias goes up, equal treatment goes down.
In other words, this type of bias can manifest and impact health without the physician or patient realizing it. The problem is realized, though, in communities and society, when a group of people experiences poorer health outcomes because of who they are.
What can be done to eliminate unconscious bias?
There are some very important additional findings with clues to this in the study:
Residents who figured out what the study was testing behaved more like someone without implicit bias, even if they tested with implicit bias. Awareness seemed to be protective. (The researchers went to some lengths to make the study’s purpose opaque to the subjects, and removed subjects from analysis if they figured it out)
Before completing the IAT section of the study, 60.5% of physicians agreed or strongly agreed with the statement: “Subconscious biases about patients based on their race may affect the way I make decisions about their care without my realizing it.” When shown the same statement after taking the IATs, 71.6% of physicians agreed or strongly agreed with this statement (difference in mean 5-point score=0.33, P<.001 by paired t test). Meanwhile 74.8% felt that taking IATs is a worthwhile experience for physicians, and 76.1% felt that learning more about unconscious biases could improve their care of patients.
A few important points, connection to LGBTQI Health
In the words of the authors:
…implicit biases may affect the behavior even of those individuals who have nothing but the best intentions, including those in medical professions.
This is reassuring, because it means that behavior that is commonly seen in the care of patients, including transgender patients (21% routinely denied care, 2% experience physical violence in medical offices) may not solely due to overt prejudice against a class of people that many clinicians have never interacted with. I’ll be posting more recent research that implicit bias can be modulated through direct contact or even through imagined contact.
Becoming aware of implicit bias and working to eliminate it has the potential to change the behavior of people with good intentions into intentional good behavior, with great (and equal) results for all patients who by definition deserve to achieve their life goals through optimal health. That’s what health care is here for 🙂 .