Association of Racial and Socioeconomic Diversity With Implicit Bias in Acute Care Surgery

Abstract

This study assesses whether racial and socioeconomic diversity mitigates the existence of implicit bias in the field of trauma care among surgical health care professionals.

Our perceptions of others are created by societal influences that can develop into explicit (conscious) and implicit (unconscious) biases. These biases can be based on race, social class, sex, sexual orientation, or other factors, in isolation or in combination. Racial and social class disparities in health outcomes have been shown to exist for many years and may, in part, be related to biases.¹ Studies of health care professionals as early in their careers as medical students² identified that 69% had implicit bias in favor of white people and 86% in favor of those in the upper class. These biases have also been demonstrated in other medical professionals throughout the United States, reflective of the overall population.³ A recent survey of trauma surgeons showed that 74% demonstrated an unconscious preference toward white people and 92% toward the upper class.⁴ This study explores whether the racial and socioeconomic diversity in a city such as Miami, Florida, which is 60% Hispanic and 20% black and has a poverty level of 18%, would mitigate the existence of implicit bias in the field of trauma care.

Methods

Participants were recruited from varied health care staff, including surgeons, anesthesiologists, registered nurses, social workers, emergency medical technicians, and respiratory therapists from the Ryder Trauma Center (American College of Surgeons level I), in Miami, Florida. Data on race/ethnicity, sex, professional and socioeconomic demographics were collected. Individuals were tested for implicit race and social class biases during a 2-month period using the validated, web-based Implicit Association Test.⁵ Implicit Association Test D scores compare the time in which an individual sorts associated concepts with preferences determined by shorter reaction time as a surrogate for unconscious bias.⁵ This study was approved by the institutional review board of the University of Miami and Jackson Memorial Hospital, Miami, Florida. Oral informed consent was obtained from each study participant before initiating the Implicit Association Test. Analysis included multivariable logistic regression (P < .05) with the D score discarding criteria of greater than 30% error or greater than 10% faster response (<300 milliseconds). We performed χ² testing with significance at 1-sided P < .05.

Results

Of the 91 participants, 54 (59%) were male and 37 (41%) were female. The largest age quintile was 31 to 40 years of age (40 [44%]). In terms of self-identified race/ethnicity, 38 participants were Hispanic (42%), 30 were white (33%), 12 were black (13%), 8 were another race/ethnicity (9%), and 3 were Asian (3%). The most represented occupations were registered nurses (32 [36%]), surgeons (19 [22%]), emergency medical technicians (17 [19%]), and anesthesiologists (7 [8%]). Assessment of financial security of our participants determined that 40 (56%) worried about money 1 to 10 times each week, whereas 6 (8%) worried about money more than 10 times per week. (Table). Eighty participants (88%) stated that they had no explicit preference for white compared with black or Hispanic people. Nonetheless, there was implicit bias by Implicit Association Test D score in favor of white compared with black people and white compared with Hispanic people at 45 (49%) and 57 (63%) participants, respectively. Forty-six participants (51%) stated that they had no explicit preference for upper class compared with poor people, yet 75 (82%) were implicitly in favor of the upper class. Implicit bias among health care professionals in our study was overall lower for race (51 of 91 [56%]; P = .002) and social class (75 of 91 [82%]; P = .02) when compared by χ² analysis to previous studies of trauma surgeons across the United States.⁴

Table. Participant Demographic Characteristics.

Characteristic	Study Sample, No. (%) (n = 91)
Racial/ethnic group
Asian	3 (3)
Black Hispanic	2 (2)
Black non-Hispanic	12 (13)
White Hispanic	36 (40)
White non-Hispanic	30 (33)
Other	8 (9)
Occupationa
Environmental service assistant	1 (1)
Social worker	3 (3)
Patient transport services	1 (1)
Registered nurse	32 (36)
Nurse practitioner	1 (1)
Respiratory therapist	5 (6)
Radiographic technician	2 (2)
Anesthesiologist	7 (8)
Surgeon	19 (22)
Emergency medical technician	17 (19)
Financial concerns (weekly money worry)
Never	41 (45)
1-3 times	25 (27)
4-6 times	12 (13)
7-10 times	3 (3)
>10 times	6 (7)
Multiple times per day	4 (4)
Sex
Male	54 (59)
Female	37 (41)

^aAvailable for 88 participants.

Discussion

Although trauma surgeons in the United States have an implicit bias in favor of the upper class and white people,⁴ trauma health care professionals in a multicultural city had lower levels of racial and social class bias. Although effects on clinical outcomes have been difficult to show, implicit bias affects the overall patient environment.⁶ These results indicate a need for further research to explore the concepts of diversity and even exposure through cultural dexterity training as possible protective factors to mitigate implicit bias in health care professionals. Limitations include the lack of robust validation of the social class component of the Implicit Association Test, which is relatively novel, and prior knowledge of study intent and attempt at social desirability responses. The lack of associated clinical vignettes in this study make it difficult to understand the clinical repercussions of such bias. Trauma health care professionals are often on the front lines of health care; thus, addressing such biases is important to appropriately advocate for our disadvantaged patient populations.