Implicit Bias in Health Care ___________________________________________________________________
over Black faces. For single-race Black individuals, 45% had implicit preference for their own group. For biracial White/ Black adults, 23% were neutral. In addition, 22% of biracial White/Asian participants had no or minimal implicit racial biases. However, 42% of the White/Black biracial adults leaned toward a pro-White bias. In another interesting field experiment, although not specifi- cally examining implicit bias, resumes with names commonly associated with African American or White candidates were submitted to hiring officers [41]. Researchers found that resumes with White-sounding names were 50% more likely to receive callbacks than resumes with African American- sounding names [41]. The underlying causes of this gap were not explored. Implicit bias related to sex and gender is also significant. A survey of emergency medicine and obstetrics/gynecology residency programs in the United States sought to examine the relationship between biases related to perceptions of leader- ship and gender [42]. In general, residents in both programs (regardless of gender) tended to favor men as leaders. Male resi- dents had greater implicit biases compared with their female counterparts. In a scoping review of studies around the world, researchers identified 87 studies that assessed unconscious biases among healthcare professionals [109]. Racial implicit biases were most frequently studied. Physicians and nurses were included in the majority of the studies. Analysis of the included studies indicates that implicit biases remain prevalent among healthcare providers. Other forms of implicit bias can affect the provision of health and mental health care. One online survey examining anti-fat biases was provided to 4,732 first-year medical students [43]. Respondents completed the IAT, two measures of explicit bias, and an anti-fat attitudes instrument. Nearly 75% of the respondents were found to hold implicit anti-fat biases. Interestingly, these biases were comparable to the scope of implicit racial biases. Male sex, non-Black race, and lower BMI predicted holding these implicit biases. Certain conditions or environmental risk factors are associated with an increased risk for certain implicit biases, including [44; 45; 106]: • Stressful emotional states (e.g., anger, frustration) • Uncertainty • Low-effort cognitive processing • Time pressure • Lack of feedback • Feeling behind with work • High patient caseload • Lack of guidance • Long hours • Patient overcrowding • High-crises environments
Visit https://implicit.harvard.edu/implicit and complete an assessment. Does it reflect your perception of your own biases? Did you learn anything about yourself? inter active activity
Measuring implicit bias is complex, because it requires an instrument that is able to access underlying unconscious processes. While many of the studies on implicit biases have employed the IAT, there are other measures available. They fall into three general categories: the IAT and its variants, prim- ing methods, and miscellaneous measures, such as self-report, role-playing, and computer mouse movements [36]. This course will focus on the IAT, as it is the most commonly employed instrument. It is also important to note that the IAT is more a procedure and less a discrete measurement, because there is not a single IAT. Instead, each specific dimension (e.g., race, gender, age, disability) has its own set of items. After complet- ing the IAT, respondents are provided with results regarding their measured preference such as: “Your responses suggested a strong automatic preference for White people over Black people” [108]. The key term here is “preferences,” which does not necessarily mean implicit bias or negativity. The IAT is not without controversy. One of the debates involves whether IAT scores focus on a cognitive state or if they reflect a personality trait. If it is the latter, the IAT’s value as a diagnostic screening tool is diminished [37]. There is also concern with its validity in specific arenas, including jury selection and hiring [37]. Some also maintain that the IAT is sensitive to social context and may not accurately predict behavior [37]. Essentially, a high IAT score reflecting implicit biases does not necessarily link to discriminating behaviors, and correlation should not imply causation. A meta-analysis involving 87,418 research participants found no evidence that changes in implicit biases affected explicit behaviors [38]. EXTENT OF IMPLICIT BIASES AND RISK FACTORS Among the more than 40 million participants who have completed the IAT at the Project Implicit website, individuals generally exhibited implicit preference for White faces over Black or Asian faces. In addition, there is a general preference for heterosexual people over gay individuals, young over old individuals, thin over obese people, and cisgender over gender and sexual minorities (LGBTQ+) [107]. The Pew Research Center also conducted an exploratory study on implicit biases, focusing on the extent to which individuals adhered to implicit racial biases [40]. A total of 2,517 IATs were completed and used for the analysis. Almost 75% of the respondents exhibited some level of implicit racial biases. Only 20% to 30% did not exhibit or showed very little implicit bias against the minority racial groups tested. Approximately half of all single-race White individuals displayed an implicit preference for White faces
18
MDIL1526
Powered by FlippingBook