Implicit Bias in Health Care: 2020  

The year 2020 was momentous in healthcare due to the sudden and acute onset of the COVID19 global pandemic. Our data collection strategy resulted in a total of 25 studies that were published in 2020 about implicit racial bias. Similar to our findings in 2019, the largest number of articles (11) were focused on mitigation strategies, followed by COVID19, and pediatrics. Aside from pediatrics, specialties were again represented broadly instead of deeply, with articles at either the broad cross-sectional level ​(Bonica et al., 2020; Chuang et al., 2020; Hagiwara et al., 2020)​ or as a single study in emergency medicine (Manchanda & Macias-Konstantopoulos, 2020), Neurology ​(Charleston et al., 2020)​, Obstetrics and Oncology ​(Lowe et al., 2020a; Lowe et al., 2020b)​, or Radiology ​(Maxfield et al., 2020)​. As was the case with articles published in 2018 and 2019, the findings from these studies will empower us to chart a clearer map for understanding how implicit bias may have detrimental effects on patient care, including specific dimensions like provider-patient communications, treatment recommendations, and health outcomes more broadly.  

Impact of Healthcare Providers’ Implicit Racial Bias on Patient Care  

Pediatrics  

Although there is growing evidence of implicit bias favoring White patients over Black and Hispanic patients in general (​Maina et al., 2018)​, most studies have used Implicit  

Association Tests with adult-based race vignettes. One evolution in implicit racial bias research in 2020 was the improved utilization of age-specific IAT measures. Results from all three 2020 pediatric studies in our dataset affirm that implicit racial bias affects medical trainees (medical students and residents) as well as established physicians with little to no impact of patient age on the level of healthcare provider bias. Together these findings together draw attention to the presence of biased racial attitudes among healthcare providers toward both adult and pediatric patients, underscoring the vulnerability of children within the healthcare system.   

In a review of implicit bias in pediatric physicians, Tiffani Johnson (2020) drew on two studies that compared measures by using both adult and child race vignettes to test heath care providers’ implicit bias and highlighted that children are also vulnerable to biased attitudes held by their health care providers (see also Thomas, 2018; Hymel et al., 2018). The two studies reviewed by Johnson were both conducted in the emergency department setting and examined implicit racial bias among physicians towards children. In one study, a majority of resident physicians from a pediatric emergency department showed pro-white/anti-black racial bias on both the Adult Race IAT (85%) and Child Race IAT (91%). Notably, there were no significant differences in the levels of implicit bias toward adults versus children. Prior research also affirms that levels of implicit bias did not vary based on resident demographic characteristics like gender or medical specialty when comparing pediatric resident physicians with emergency medicine residents (Johnson et al., 2016). In another study involving emergency department providers at five hospitals in the Midwest, a significant majority (84%) exhibited implicit bias favoring White children over American Indian children.  

Racial and ethnic disparities in the diagnosis of disruptive behaviors have similarly received ongoing scholarly attention ​(Cameron & Guterman, 2007; Feisthamel & Schwartz, 2009; Mizock & Harkins, 2011)​. Previous studies have identified a tendency for Black and Latino youth from urban and low-income backgrounds to be overrepresented in this type of diagnosis (Cameron & Guterman, 2007). On the contrary, these same groups of minority youth are less likely than their non-Hispanic White counterparts to receive a diagnosis of ADHD ​(Coker et al., 2016)​. Facing the same group of young vulnerable patients, Matthew Fadus and colleagues (2020) presented a review that focused on disparities in the diagnosis of psychiatric disorders, specifically pertaining to oppositional defiant disorder (ODD), conduct disorder (CD), and attention-deficit/hyperactivity disorder (ADHD). Fadus et al. (2020) linked this concerning pattern to implicit biases held by clinicians, which may lead them to over-pathologize behaviors of ethnic and minorities as more disobedient and dangerous ​(see also Clark, 2007).​ 

Third and finally pain-related disparities have been a longstanding concern in healthcare studies​ (e.g., Anderson et al., 2009; Campbell & Edwards, 2012; Green et al., 2003; Tait & Chibnall, 2014)​. Previous research found that providers are influenced by implicit racial bias, perceiving African Americans as having a higher pain tolerance than other racial and ethnic groups (Anderson et al., 2009). Megan Miller and colleagues (2020) assessed providers’ implicit bias in pain-related decision making in a pediatric medical setting.  Their study sampled 129 medical students, and the results were consistent with the findings of previous studies, showing that the medical students assumed that African Americans had stronger pain tolerance than their White counterparts. It was also found that providers were more likely to recommend opioids for African American pediatric patients than White patients, which could lead to potentially harmful over-prescription for Black pediatric patients. These findings extended the scope of pain tolerance bias research that previously focused on acute pains, showing that providers held the same type of implicit racial bias applies when it pertains to chronic pain tolerance. These findings underscore the alarming reality that implicit racial biases in pain perception cross age ranges to include children under the age of 18.  

COVID19  

The COVID-19 pandemic disrupted and transformed the field of healthcare in ways that few could have imagined. As the pandemic spread across the world in 2020, it became increasingly evident that different racial and/or ethnic groups had varying COVID-related incidence and mortality rates ​(Wadhera et al., 2020)​. Our dataset included five (5) early studies that explored the impact of implicit racial bias on COVID-19 as a public health issue. Moreover, a sudden increase of stigmatizing media reports about the virus’ origins also fueled implicit and explicit biases against Asian Americans ​(Darling-Hammond et al., 2020)​. The racial implications of the murder of George Floyd ​(Davis-Martin, 2020)​ and media linkages of COVID-19 to people of Asian descent alike posed significant challenges in health care outcomes. As we illustrated in the general introduction to the State of the Science, patient-provider trust plays a direct role in health outcomes. Moreover, studies by both Burgess et al. (2019) and Bonica et al. ​(2020)​ suggest provider ideology is increasingly playing a role in both provider location (Bonica et al.) and patient care (Burgess et al.). 

Using data from Project Implicit, a non-profit platform that provides IATs as a “virtual laboratory” for studying implicit social cognition on the Internet,  Thomas and colleagues (2020) analyzed anti-Black bias among non-Hispanic Whites across 2,994 counties. It was estimated that high levels of  implicit bias at the county-level was positively associated with high COVID-19 incidence rate among Black Americans. This study was the first to test and report the associations between county-level implicit racial bias and COVID-19 incidence and mortality rates in the United States (Thomas et al., 2020). The authors noted that further research is needed to better understand the relationship between implicit racial bias and disparities caused  by an infectious rather than chronic disease.   

The COVID-19 pandemic not only shed light on the detrimental effects of healthcare providers' implicit racial bias but also sparked discussions about the potential influence of implicit bias as perceived by patients in physicians.  Davis-Martin (2020) chronicled their experiences as a behavioral health provider to help their patients deal with anxiety about upcoming procedures. The patient in this case experienced heightened anxiety due to the combined impact of COVID-19 and racial tensions prevalent at the time. In his expression of concern to the author, the patient talked about how he feared that the riots featured in the news might negatively affect perceptions of his racial group. Consequently, the patient felt uncomfortable disclosing this fear to his White primary care provider, worrying the quality of care he receives may be negatively impacted, stating, “What if my surgeon does a bad job because they think I hate the police or America?” This narrative offered a unique perspective from a Black patient’s point of view, showing that even the mere suspicion that one’s physician may have implicit bias is enough to alter a patient’s decision-making process, leading to hesitation in scheduling a surgery appointment with a White healthcare provider.  

While African American communities were disproportionately burdened by higher COVID-19 incidence and mortality rates, the pandemic also sounded an alarm about the effects of stigmatizing media on the health and welfare of Asian Americans. On March 8, 2020, a 650% increase in Twitter retweets using the term “Chinese virus” and affiliated terms was observed.  

The use of these terms had an 800% increase in conservative news media articles the following day. Darling-Hammond and colleagues (2020) analyzed data from non-Asian participants of the Project Implicit “Asian Implicit Association Test” from 2007-2020. The results showed that Implicit Americanness bias (i.e., the subconscious belief that individuals of European origins are more “American” than those of Asian origins) declined from 2007 through early 2020. The trend was reversed, however, following the explosion of use of racially stigmatizing language on media outlets. This trend reversal is consistent with the longstanding stereotype of Asian  

Americans as “perpetual foreigners;" the researchers argued that it may also lead to increases in discriminatory behaviors, with adverse mental and physical outcomes for those who experience discrimination. While these articles were published early in the pandemic cycle, numerous articles have sadly documented tremendous increases in violence and discrimination against Asian Americans well beyond the healthcare sphere in the shadow of the pandemic ​(Chan et al., 2024)​.  

Implicit Racial Bias Across Specialties  

In 2020, new studies on differential treatment in healthcare focused more on specific types of providers and specialties. A systematic review showed that health care providers from various backgrounds and levels of training hold implicit biases against non-White individuals (Maina et al., 2018). Similar results were also found in emergency medicine and radiology in more up-to-date studies published in 2020. Examining 40 emergency medicine faculty, Cleveland Manchanda and Macias-Konstantopoulos ​(2020) ​found that 45% of their sample reported IAT results with an implicit preference for Whites. Moreover, Maxfield and colleagues (2020) examined the IAT results reported by 31 faculty participants at three academic radiology departments. The results of this study revealed that 71% of participants exhibited implicit anti-Black bias. Adding to these studies that primarily focused on physician implicit bias, Lynn Crandlemire ​(2020)​ underscored in her review of implicit bias’s impact on caring, that the majority of nurses are susceptible to implicit racial bias as well (see also ​Haider et al., 2015)​.  

Previous research has also shown that implicit racial bias held by clinicians can negatively impact communication with patients, which can foster decreases in patient trust, engagement, and adherence to treatment (Zestcott et al., 2016). Chenery Lowe and colleagues (2020a, 2020b) extended this line of research to the field of genetic counseling to examine counselors’ implicit bias with mixed findings. Using Linguistic Inquiry Word Count (LIWC) and Roster Interaction Analysis System (RIAS), they examined the influence of implicit racial bias in genetic counselors’ facilitation of emotional and cognitive processing (Lowe et al, 2020a). Contradictory to the conventional findings in studies with clinicians, they did not find an association between genetic counselors’ pro-White implicit bias and their communicative behavior during sessions. The researchers suggested that it is possible that the LIWC measure was not sensitive enough to capture an implicit bias effect in counseling sessions, or the small sample size used in the study could have limited their ability to detect associations between cognitive and emotional processing and IAT scores.   

Adopting a different approach to focus on counselors’ framing of information, the researchers conducted another study to investigate whether genetic counselors’ implicit racial bias affected their communication style with simulated minority clients (Lowe et al., 2020b). The study involved a sample of sixty genetic counselors, primarily consisting of females (92%) and individuals identifying as non-Hispanic/Latino White (90%). Participants completed the Race IAT assessment and engaged in simulated counseling sessions with clients (17 sessions with simulated White clients and 43 sessions with simulated African American or Hispanic Latino clients). The information framing employed by counselors in the recorded sessions was analyzed using a genetic counseling-specific adaptation of RIAS coding. The results showed that implicit racial bias may negatively impact clinicians’ communication style. Genetic counselors with higher pro-White implicit bias provided less informational individuation and made fewer facilitation and activation statements to minority clients compared to their White counterparts. It was also found that with counselors who scored higher on the IAT, minority clients asked fewer medical questions and disclosed less information (Lowe et al., 2020b). These results highlighted that the implicit racial bias held by a genetic counselor likely affects counselor-patient communication in both directions.   

Conclusion 

While much has been written about COVID19 and explicit bias, discrimination, and even violence, particularly against African Americans in light of George Floyd’s murder and against Asian Americans in light of the pandemic, we kept a strict focus on implicit racial bias for this 2020 analysis. For example, Thomas’ 2020 results noted that implicit bias was significant under a single circumstance, while explicit bias was consistently statistically significant. As we’ve already noted, the negative effects of implicit racial bias are real, and that is why we have kept our attention focused on them. We take up lingering questions about research impact along with those from Maina et al., 2018 and other articles in the Conclusion and Directions for Future Research section of this report.