Table 3 Characteristics of studies

COGNITIVE BIAS

 Chew et al. [34]

(2016)

Malaysia

To test the hypothesis that the TWED checklist facilitates metacognition among medical students so that they can make better-quality clinical decisions.

Quasi-experimental

40 students (intervention: 21; control: 19)

Bachelor's/ Undergraduate

University provider and setting

Medicine

Cognitive bias:

- Anchoring bias

- Availability bias

- Confirmation bias

- Search satisficing

1 x 90-minute tutorial on dual process thinking, cognitive biases, debiasing strategies and the TWED checklist (T = threat, W = what else, E = evidence and D = dispositional factors)

20 true/false question quiz based on 5 x clinical case scenarios

The ability of students to generate a second more serious diagnosis, and their ability to decide on appropriate investigations and treatment plans

Overall mean scores in all 5 clinical cases

Intervention group:

18.50 +/- 4.45 marks

Control group:

12.50 +/- 2.84marks

p<0.001

 Chew et al. [35]

(2017)

Malaysia

To test the hypothesis that the application of a mnemonic checklist aids the script evaluation stage by stimulating the consideration of additional patient data, thus leading to better clinical decisions.

Quasi-experimental

88 final year students (intervention: 48; control: 40)

Bachelor's/Undergraduate. University provider and setting

Medicine

Cognitive bias:

- Premature closure

2-hour tutorial covering topics of factors that contribute to diagnostic errors and strategies to minimize them using case-based discussions. The TWED checklist (Threat, what else, Evidence, Dispositional influences) was also introduced and explained to students.

Script concordance test consisting of 10 cases with 3 items

Participant’s decisions on the likelihood of a given hypothesis when additional patient data is provided.

The median total scores for the intervention group and the control group were:

18.65 (IQR 16.96 - 20.34) and 18.15(IQR 16.79 - 19.37) out of 30, respectively.

U=792

z=−1.408

p=0.159.

 Deveau et al. [36]

(2021)

Canada

To discuss a pilot project using simulation-based learning to integrate exploration of thinking and identification of cognitive biases

Cross sectional

19 students

Bachelor's/Undergraduate University provider and setting

Nursing/ Midwifery

Cognitive bias:

- Anchoring bias

- Availability bias

- Confirmation bias

- Framing effect

- Search satisficing

- Premature closure

Simulation based learning. Simulation (case scenario - 21 year old single mother with her infant son)

Debriefing post simulation (using good judgement framework and a cognitive autopsy approach)

Exploration of clinical reasoning and identification of cognitive biases

Clinical reasoning:

42% (n=8) identified the correct clinical impression.

Identification of cognitive biases:

Anchoring bias 63% (n=12)

Confirmation bias 47% (n=9)

Search satisficing 42% (n=8)

Unpacking principle 42% (n=8)

Premature closure 32% (n=6)

Availability bias 21% (n=4)

Overconfidence 21% (n=4)

Framing 16% (n=3)

Diagnostic momentum 16% (n=3)

Hershberger et al. [37] (1996) 

USA 

To determine:

(1) how capable medical students and practicing physicians are in avoiding cognitive biases in medical decision-making, (2) whether susceptibility to cognitive bias varies by medical specialty, and (3) whether awareness of cognitive bias in medical decision-making can be taught to medical students.

Quasi-experimental

230 students (118 intervention group; 112 control group)

Bachelor's/ Undergraduate

University provider and setting

Medicine

Cognitive bias:

- Availability bias

- Confirmation bias

- Representative-ness heuristic

1 x Seminar to address predictable tendencies in information processing that can adversely impact decision making

The Inventory of Cognitive Biases in Medicine (ICBM) was used to determine the influence of bias, and the effectiveness of the seminar in teaching the principles of cognitive bias in medical decision-making.

1: The extent of cognitive bias in medical decision-making and whether experienced physicians differed from novices

2: If awareness of cognitive bias in medical decision-making could be taught.

1: Preclinical medical students' mean score: 40.7% (SD = 12.5%). Practicing physicians mean score: 49.0% (SD = 14.9%).

Influence of bias (49% vs 41%, t = 4.07, p < 0.001).

2: Intervention group: 40.7% (SD = 12.5%).

Control group: 55% (no SD provided)

t = 7.83, p < 0.001

Sherbino et al. [18]

(2014)

Canada

To determine the effect of cognitive forcing strategies (CFS) training on diagnostic error in senior medical students

Quasi-experimental

198 final year students of a 3-year curriculum

(intervention: 145, control: 46)

Bachelor/Under-graduate

University provider

Hospital/Clinic setting

Medicine

Cognitive bias:

- Availability bias

- Search satisficing

1 x 90 minute interactive case-based seminar facilitated by expert EM clinicians experienced with teaching Cognitive Forcing Strategies based on Croskerry’s model.

A 2-hour, 48 question, computer-based examination using case studies

Search Satisficing Bias: Compared the proportion of students in the intervention group and the control group who searched for a second diagnosis and the proportion whose second diagnosis was correct, and the proportion that identified a second diagnosis was compared for false positive cases (for which there was no second diagnosis).

Availability Bias:

Compared the proportion that identified the uncommon diagnosis.

Search Satisficing Bias:

52% of intervention group and 48% in the control group initiated a search for the second diagnosis (x²= 0.13, df = 1, p = 0.91).

Of these students, 54% of the control group correctly identified the second diagnosis compared to 48% in the control group (x² = 2.25, df = 1, p = 0.13).

Second diagnosis/false positive cases, 64% of the intervention group and 77% of the control group incorrectly identified it (x2 = 2.38, df = 1, p = 0.12).

Availability Bias:

45% in each group identified the uncommon correct diagnosis (x² = 0.001, df = 1, p = 0.98).

IMPLICIT BIAS

Avant et al. [38]

(2019)

USA

To evaluate an elective course focused on exposing students to the root causes of health disparities, con-temporary factors that perpetuate disparities, and evidence-based policies to reduce health disparities.

Qualitative

9 second- and third-year students

Undergraduate/ Bachelor’s

University provider and setting

Pharmacy

Implicit bias:

- Race

Course on bias and structural inequalities as drivers of inequality.

Students learned through self-reflection; perspective taking; and group activities.

Active learning used throughout the course to promote self-discovery (e.g., social identity mapping).

- Short answer exam

- learning management system discussion threads

- weekly reflections

- photo presentation on social determinants of health and equity,

- comparing two racially segregated neighboring suburbs

Identified 5 strategies to facilitate this course:

Knowledge and understanding of drivers of health disparities, bias and structural inequalities. Regarding racial/ethnic health disparities; and encouraging personal awareness of privileged and marginalized identities.

Five themes emerged from student reflections underlying strategies to facilitate this course:

create and maintain a welcoming and inclusive learning environment.

(2) utilize experiential learning for personal awareness development and knowledge expansion.

incorporate intergroup diversity to empower learners to create change.

anticipate and acknowledge emotions to facilitate learning and:

5) provide students with an opportunity to complete a final self-reflection paper.

Gatewood et al. [39]

(2019)

USA

To raise awareness of implicit bias and its influence on health outcomes and support a discussion on ways to mitigate the impact of implicit bias.

Cross sectional

110 nursing students from 4 universities

Bachelor's/ Undergraduate 13 students); Masters (33 students); Doctorate (64 students)

University provider and setting

Nursing/ Midwifery

Implicit bias:

- Race

3-part educational activity:

1)preparatory activities (video and articles on implicit bias in healthcare and the IAT).

2)IAT.

3)Discussion activity regarding results of IAT and potential actions to mitigate effects of implicit bias.

- Facilitated discussion about activity

- Online evaluation of usefulness of the assignment

Students level of agreement preparatory material provide to increase awareness of the effects of implicit bias (IB) on quality in healthcare, identify a resource for self-assessment of IB using the IAT, and integrate knowledge of students' own IBs into their nursing care to improve the quality of their care.

Preparatory learning activities were helpful:

Video: 89% strongly agree or agree

Readings: 86% strongly agree or agree.

Increased awareness of personal biases:

83% strongly agree or agree.

Recognition of their IB (via the IAT) would be helpful in managing their nursing care:

90% strongly agree or agree.

Usefulness of discussion and overall impression of activity to manage IB:

Discussion - 70% strongly agree or agree.

Overall - 69% strongly agree and agree.

Gonzalez et al. [40]

(2014)

USA

To describe an educational intervention addressing both health disparities and physician implicit bias and the results of a subsequent survey exploring medical students’ attitudes and beliefs toward subconscious bias and health disparities

Cross sectional

218 3^rd-year students over a 2-year period (2 cohorts)

Bachelor's/ Undergraduate

University provider and setting

Medicine

Implicit bias:

- Race

- Religion

Single session within a course. Preparatory activities: pre-reading on health disparities, physician behaviour and implicit bias; written reflection; IAT

Session: Faculty-led discussion on health disparities; students’ personal experiences; and effect of bias on decisions.

- A 15-question anonymous 4-point Likert scale survey

Students’ attitudes and beliefs toward subconscious bias and health disparities and to evaluate the association between students’ self-assessment and what they believed or observed about health disparities following an educational intervention and completion of the IAT.

Survey response rate: 69%. 22% of responders were “strongly disagreed or disagreed (deniers) to the statement Unconscious bias might affect some of my clinical decisions or behaviors compared to 77% of responders who strongly agreed or agreed (acceptors) with the statement.

9% of deniers and 1% of acceptors disagreed with the statement Health -Disparities Do Not Exist in the United States (p = 0.02)

Gonzalez et al. [41]

(2020)

USA

To provide an overview of a skills based elective course and its evaluation in implicit bias recognition and management

Mixed method

First year students over 3 years (3 cohorts)

Bachelor's/ Undergraduate

University provider and setting

Medicine

Implicit bias:

Not specified

Elective subject consisting of 9 x 1.5hour sessions. The conceptual framework developed by Teal et al. was used to guide instructional design. The course consisted of 2 sections.

Section 1: reflections on own biases and experience of completing the IAT; perspective-taking exercises.

Section 2: roleplays and brainstorming session on addressing perceived bias during witnessed encounters.

2 x Focus groups during and post-completion of course, exploring perceptions of course.

The student’s ability to:

Recognize when implicit bias may influence one’s own communication with a patient or peer.

Advocate on behalf of patients when perceiving bias in a witnessed encounter; and

3. Address biased comments made within the learning environment.

3 themes emerged from analysis of focus groups:

1: Student engagement can be enhanced.

2: Instruction is empowering, and

3: It (addressing bias in one’s own and witnessed encounters) can be done!

Matharu et al. [42]

(2014)

USA

To determine whether an innovative educational intervention (reading a play about obesity) could diminish obesity prejudice relative to a standard medical lecture

Randomized controlled trial

129 1st-4th year students from four universities

Bachelor's/ Undergraduate

University provider and setting

Medicine

Implicit bias:

- Weight bias

Intervention: 1 hour dramatic reading of a script incorporating narratives about their weight in the context of social discrimination.

Control group: 1 hour lecture on the medical management of obesity.

- The obesity specific IAT

- The anti-fat attitudes questionnaire.

- Jefferson Scale of Physician Empathy (JSPE)

- Two open-ended questions:

- A) Discuss whether they viewed obesity as a civil rights issue or a medical/public health issue.

- B) Formulate a treatment plan for an overweight but otherwise healthy older woman.

Explicit attitudes and implicit bias toward obese individuals were assessed prior to intervention and after four months.

Explicit fat bias difference change in post measures for the intervention group -4.5 +/- 11.5 (P = 0.002), compared to the control group with 0.76 +/- 12.0 (P = 0.61), representing a difference in change from baseline of -5.5 (P = 0.05).

Implicit fat bias difference change in post measures for the intervention group -0.03 +/- 0.44, compared to the control group with 0 +/- 0.48 representing a difference in change from baseline of -0.04.

Empathy score differences change in post measures for the intervention group -4.7 +/- 13.4 (P = 0.007), compared to the control group with 2 +/- 9.3 (P = 0.02), representing a difference in change from baseline of +2.2.

Students in the control group were more likely to endorse a prescriptive model of patient care (P = 0.03).

Motzkus et al. [43]

(2019)

USA

To explore whether three-hours of focused content and discussion regarding bias and stereotypes could produce student reflections demonstrating insight into these issues.

Qualitative study using grounded theory methodology

250 1^st-2^nd year students

Bachelor's/ Undergraduate and Doctorate

University setting and provider

Medicine

Implicit bias:

Not specified

1 x 3 hour lecture / discussion session on implicit bias and completion of IAT on any topic of students choosing

Written reflective essay

Understanding students’ readiness to accept their implicit biases and to address unconscious bias.

25 themes were identified that could be categorized into 3 areas:

1: Experiencing taking IAT.

2: Bias in medicine; and

3: Prescriptive comments.

84% of students acknowledged the importance of recognizing implicit bias.

60% of students noted that bias affects clinical decision-making.

19% stated they believe it is the physician’s responsibility for dismantling bias.

56% of students felt that the IAT promoted self-reflection.

56% acknowledged that bias is a product of society/culture/ upbringing.

Schwartz et al. [44]

(2020) USA

To investigate whether increases in students’ reflective capacity are associated with improved cognitive empathy scores; and whether an increased emphasis on understanding the nature of implicit bias in a medical humanities course, and the resultant pedagogy of discomfort, would cause students to accept the possibility that unconscious bias might affect some of their clinical decisions or behaviors as healthcare professionals.

Mixed methods

34 prospective medical students

Masters

University setting and provider

Biomedicine

Implicit bias:

Not specified

A humanities course designed to promote development of personal and professional skills and identities through readings, quizzes, discussions about communication and implicit bias, and written reflections on service to the community

- Reflective Practice Questionnaire

- The Jefferson Scale of Empathy.

- Custom designed survey of student’s attitudes toward team service-learning projects and unconscious bias post course completion

- Team minutes and reflections

If performing service-learning projects would foster students’ compassion by raising their reflective capacity, empathy, and unconscious bias mitigation.

An increase in reflective capacity scores of prospective medical (MSBS) students in association with written reflections on service learning between August and December 2019 (r=0.26, p=0.02).

An increase in the cognitive empathy (JSE) scores of prospective medical (MSBS) students occurred in association with the rise in their reflective capacity (RC) scores between August and December 2019 (r=0.27, p=0.015).

Prejudices expressed by students in a survey (and number of times expressed) regarding the biases of which they became aware in their team service-learning experiences:

- Age (10)

- Homeless people (7)

- Culture/Race (6)

- Socioeconomic status (4)

- Obesity (3)

- Gender (3)

- Hygiene (1)

- Mental health patients (1)

- Men I do not know (1)

- Environmental (1)

Van Winkle et al. [45]

(2021).

USA

To determine if remote learning (compared to in class learning) diminished students’ capacity for reflective capacity and cognitive empathy to foster positive attitudes and behaviors for critical reflection, implicit bias mitigation, empathy, and compassion.

Cross sectional

61 prospective students compared to 81 students’ results from three previous cohorts

Masters

University setting and provider

Biomedicine

Implicit bias:

- Age bias

- Racial bias

- Sex/sex-role bias

- Socio-economic status bias

- Weight bias.

- Substance abuse bias

- Disability bias

- Mental health bias

Readings, IAT. Classes held by Zoom, including quizzes and application exercises concerning communication skills and implicit bias. Service-learning projects. Team-based activities to write reflections, minutes from team meetings.

- Written assignments.

- Reflective Practice Questionnaire

- Jefferson Scale of Empathy.

- Online survey about opinions concerning implicit bias, team-based learning, and service to the community

Students’ capacity for reflective capacity, empathy, and unconscious bias mitigation.

Reflective capacity scores increased during Medical Humanities course between August and December 2020 (r=0.56, p=0.000).

Cognitive empathy (JSE) scores increased during Medical Humanities course between August and December 2020 (r=0.37, p=0.003).

Prejudices expressed by students in a survey (and number of times expressed) regarding the biases of which they became aware in their team service-learning experiences:

- Ageism (10)

- Economic class/homelessness (8)

- Obesity (7)

- Sexual orientation (7)

- Race (3)

- Substance abuse/addiction (3)

- Strong political opinions (3)

- Men (3)

- None/no interaction (3)

- Disabled (2)

- Mental health issues (2)

- Veterans (2)

- Favor same as me (2)

- Appearance/dress (1)

- Smokers (1)

- Women (1)

- Favor CNAs (1)

- End of life care (1)