In September 2020, 37-year-old Atikamekw[1] woman and mother of seven, Joyce Echaquan, was admitted to Joliette Hospital in Quebec[2] with severe pain. She was clearly telling the medical staff that she was experiencing heart problems. Yet, from the very first minutes in the clinic, she was labeled as a “drug-dependent patient.”
As her pain intensified and staff continued to dismiss her suffering, Joyce began live-streaming her condition on Facebook. In the video, she appears exhausted from the pain, while nurses mock her: “You’re so weak,” one says. “Stop acting stupid, enough! What will your children think when they see you like this?” says another.
Joyce died from pulmonary edema caused by a rare heart condition. The subsequent investigation found that hospital staff tried to explain away her pain as drug dependence, disregarded her symptoms, and misdiagnosed her. She was transferred to intensive care too late, when her condition was already impossible to improve.
Ask any doctor, in any country, whether they always treat all patients equally, and almost invariably you will get the same confident answer: “Of course!”
That is precisely the response Tina[3], and I received when, as part of the Institute for War & Peace Reporting (IWPR) project, supported by the UK government, we conducted training for doctors across different regions of Georgia on inclusive healthcare, with a focus on patient communication.
But the picture changed radically when these same doctors took an implicit bias test. The test measured their predispositions across five criteria: ethnicity, gender, sexuality, disability, and age. To encourage honesty, physicians were informed from the outset that the test was solely for self-reflection, and they were not required to disclose their results. The outcomes surprised and unsettled them. They discovered that beneath the white coats and clinical routines, they often held unconscious assumptions about a patient’s gender, ethnicity, disability, poverty, appearance, or even body weight.
“In Georgia, nobody admits it, especially doctors, that certain groups of the population may be treated inappropriately, including ethnic minorities. I’m happy that even the skeptical doctors uncovered flaws in our profession. I believe this training will open many eyes, which is why everyone who interacts with patients should attend it,” said a physician from Batumi City Hospital (VIAN).
The moment a doctor realizes that they may feel differently toward a tattooed woman, or an older man who struggles to explain their problem, the illusion of objectivity cracks, and the door to potential change opens.
It’s imperative to mention that our goal wasn’t to show that these doctors are bad people. It means they are human, like all of us, and therefore can be subject to unconscious biases and cognitive shortcuts, which behavioral science calls heuristics.
Thanks to decades of research by Daniel Kahneman, the founding figure of behavioral science, and his colleague Amos Tversky, today we know that human decisions rarely rely on purely rational analysis. According to Kahneman’s famous theory, our thinking is governed by two systems: System 1, which is fast, automatic, intuitive, and emotional, and System 2, slow, deliberate, and rational. Given that our brain makes up to 35,000 decisions per day, most are handled quickly and efficiently by System 1. This system relies heavily on cognitive shortcuts, heuristics, which, while efficient, systematically lead to errors and biased perceptions. This is often how inaccurate assumptions form, even in medical decision-making.
You may recognize the influence of heuristics yourself. Sometimes we call it intuition, gut feeling, or common sense. “I feel it in my bones.” Familiar, isn’t it? But the downside of heuristics is that we are all “afflicted” by cognitive biases. Cognitive bias refers to a systematic pattern of deviation from rational judgment, leading to illogical conclusions about people or situations. By filtering perceptions in unique ways, people construct their own “subjective social reality.” And it is this subjective reality, contrary to objective facts, that often determines behavior in social contexts.
In healthcare, numerous studies have highlighted implicit bias[4], also known as “unconscious bias,” as a significant challenge. Though invisible, it has real, measurable impacts on patient care and services.
Implicit bias is defined as an attitude or internalized stereotype that subconsciously influences perception, actions, or decisions, often contributing to unequal treatment of people based on race, ethnicity, nationality, sex, gender identity, sexual orientation, religion, socioeconomic status, age, disability, or other traits. Such biases can cause significant harm to patients.
Bias knows no borders. It thrives not only in underfunded clinics or among underqualified staff, but just as easily within advanced, well-resourced health systems.
The World Health Organization (WHO) emphasizes that discrimination, stigma, and bias are key drivers of unequal health outcomes[5], especially for vulnerable groups such as women, people with disabilities, ethnic minorities, and LGBTQ+ individuals.
Research shows that healthcare professionals, including doctors, nurses, and medical students, exhibit mild to moderate levels of racial/ethnic bias, like the general population. These are not isolated cases but systemic tendencies, shaped by repeated cultural exposure.
A widely cited 2015 systematic review in the American Journal of Public Health (AJPH), which analyzed 15 studies[6], found that most health professionals subconsciously held more positive attitudes toward white patients and more negative attitudes toward minority groups, based on ethnic and religious backgrounds as well as their sexuality. This showcases an example of pro-white, anti-minority bias. These biases impact clinical decisions, patient interactions, treatment adherence, and ultimately, health outcomes. A 2021 UK review confirmed similar implicit biases globally regarding race, gender, body size, and age.[7]
Bias is not only individual, but also structural.
Implicit bias operates within broader institutional and social structures that perpetuate inequality. As Professor Janice A. Sabin, researcher at the Department of Biomedical Informatics and Medical Education at the University of Washington School of Medicine, writes:
“Bias-based discriminatory practices not only harm patient care and medical education environments—they also limit workforce diversity and the allocation of research funding.”
A 2023 study showed that the evaluation of medical articles was also influenced by the country where the research was conducted (Global South vs. Global North). This demonstrates that bias extends beyond clinical practice, affecting knowledge generation and policy priorities. [8]
Bias is widespread and often learned during practice.
A key study revealed that nearly 49% of U.S. medical students reported hearing negative comments about Black patients from senior physicians. Over time, by their fourth year, these students exhibited more substantial racial biases than in their first year[9]. In fact, within clinical culture, bias is not only tolerated but also transmitted from generation to generation.
For example, a 2016 study found that white patients were referred for stenting more often than Black patients. The justification was: “Black patients lack education and won’t engage in physical activity after surgery. So, it would be pointless.”
And this is not limited to race. Studies describe systemic bias against overweight individuals, women, people with disabilities, migrants, and patients living in poverty. [10].
A 2015 study found that in overweight patients, their complaints were often attributed primarily to obesity, resulting in frequent misdiagnosis.
In Georgia, too, for example, patients with disabilities or mental health conditions are often incorrectly labeled as “unmotivated” or “noncompliant,” not officially, at the policy level, but based on implicit assumptions that affect their accurate diagnosis, respectful treatment, and communication. [11].
Where does the Bias show most commonly according to scientific studies
- Ignoring pain by race and gender
Women’s complaints of chronic pain are more often ignored than men’s—driven by the stereotype that women are more emotional and sensitive, and thus their complaints are “baseless.”
After a heart attack, women’s in-hospital mortality is 15–20% higher, due to misdiagnosis influenced by bias toward anxiety, stress, or “atypical symptoms.”
Black patients—especially Black women—frequently report not being believed when they describe pain. Studies show clinicians systematically underrate Black patients’ pain, leading to undertreatment with analgesics, delayed interventions, or minimization of symptoms.
In the U.S., Black women are 3–5 times more likely to die from pregnancy-related complications than white women—not because of biology, but because of bias, stress, and systemic discrimination.
- Bias in communication and lack of listening
Health-care professionals often interrupt or fail to engage patients from vulnerable groups. For example, patients with limited English proficiency or a different accent are often perceived as less intelligent or less able to understand treatment instructions.
“Doctors interrupted non-white patients far more often and used patient-centered communication standards less frequently. As a result, patients were less likely to feel that they had been heard.” - Assumptions about ability to adhere to treatment
Clinicians may assume that certain groups—older adults, people with mental health conditions, or, for example, Roma patients in Eastern Europe—will be less likely to follow treatment plans. Such biased assumptions can reduce the quality and depth of care.
“Implicit bias in physicians often triggers doubts about a patient’s credibility and commitment to treatment, especially when the patient belongs to a vulnerable social group. Consequently, the patient may receive fewer diagnostic tests or less intensive therapy.” - Bias in medical documentation
Bias appears not only in care but also in records, where non-white patients are more often described with stigmatizing terms (e.g., “noncompliant,” “crazy,” “drug-seeking”), even when clinical indicators are identical to those of white patients.
Negative descriptors appear more frequently in medical records about Black patients, which then influence other clinicians and can affect long-term treatment decisions.” - Inequity in mental-health assessments
A white adolescent with anxiety may be quickly referred to therapy, whereas a non-white or low-socioeconomic-status adolescent with identical symptoms may be labeled “defiant” or as having a behavioral disorder. Symptom interpretation often shifts with the patient’s race or socioeconomic status.
These real-world examples remind us that implicit bias is not ill intent; it arises from automatic associations and systemic tendencies that affect patient experience, access to care, and health outcomes.
What can we do to reduce implicit bias?
Although no single effort can eliminate bias permanently, intentional training raises awareness, builds skills, and promotes patient-centered behavior.
A 2025 meta-analysis[12]. prepared through a collaboration between Massachusetts General Hospital and the Urban Institute, synthesizing 55 studies, found that:
- 96% of programs improved clinicians’ knowledge, skills, and attitudes toward recognizing bias;
- Gains included better communication, increased self-reflection, and greater comfort managing bias;
- These results were consistent across physicians, nurses, students, researchers, and clinical faculty, using formats such as online modules, workshops, and reflective sessions.
However, most programs did not measure long-term impacts on patient health outcomes, underscoring the need for deeper investment in impact evaluation.
In the end
Bias is human. Healthcare must be humane.
A doctor is a human being working in a system that often rewards fast thinking, stereotype-driven impulses, and hierarchical norms. Yet healing is impossible when trust is broken. Trust—especially in healthcare—rests on empathy, respect, and seeing the patient as a whole person. Bias is not a flaw to be ashamed of; it is a reality to face—and to reduce wherever possible.
[1]Cree – an Indigenous people of the Algonquian group in Canada and the U.S. Their population in Canada exceeds 200,000, making them the largest Aboriginal nation
[2] https://www.theguardian.com/world/2021/oct/06/joyce-echaquan-coroner-indigenous-systemic-racism-death
[3] Tinatin Stambolishvili, Strategic Communications Expert, Behavioural Designer, Executive Director of BISC Partners.
[4] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5333436/
[5] World Health Organization (2023). Addressing stigma and discrimination in health systems. https://www.who.int/news-room/fact-sheets/detail/stigma-and-discrimination
[6] Hall, W. J., et al. (2015). Implicit Racial/Ethnic Bias Among Health Care Professionals and Its Influence on Health Care Outcomes: A Systematic Review. American Journal of Public Health, 105(12), e60–e76. https://doi.org/10.2105/AJPH.2015.302903
[7] Gopal, D. P., et al. (2021). Implicit bias in healthcare: clinical practice, research and decision making. Future Healthcare Journal, 8(1): 40–48. https://doi.org/10.7861/fhj.2020-0233
[8]https://pmc.ncbi.nlm.nih.gov/articles/PMC10749034/#:~:text=Citations%20over%20a%202%2Dyear,often%20perceived%20as%20lower%20quality.&text=The%20preference%20even%20among%20Global,North%20journals%20is%20thus%20unsurprising.&text=The%20numerical%20advantage%20of%20Global,must%20claim%20centrality%20by%20itself’.
[9] https://pmc.ncbi.nlm.nih.gov/articles/PMC4907959/#:~:text=Our%20pedagogy%20may%20be%20undermining,and%20risks%20of%20the%20augenblick.
[10] Gopal, D. P., et al. (2021). Implicit bias in healthcare: clinical practice, research and decision making. Future Healthcare Journal, 8(1): 40–48. https://doi.org/10.7861/fhj.2020-0233; Lokugamage, A. (2020). Cultural safety in maternity care: A model to reduce disparities. BMJ Global Health.; Boge et al. (2020). Gender recognition bias in clinical settings. BMJ Open.
[11] UNICEF Georgia. (2023). Assessment of access to health services for persons with disabilities. https://www.unicef.org/georgia/reports/access-healthcare-persons-disabilities
[12] Walker, T., et al. (2025). Implicit Bias Training Has Positive Impacts on Health Care Workers’ Knowledge, Skills, and Attitudes. Urban Institute / Massachusetts General Hospital.
