Race, Bias, and Disparities in Medicine

racial bias in medicine

In the first of our diversity and inclusion series, our Executive Medical Director explores the history of racism in medicine, social determinants of health, implicit bias in healthcare, and what we can do to overcome it all.

Racial bias impacts our daily lives, and when that bias seeps into medical care, it can negatively affect patients in physical, mental, and emotional ways.

Racism is entrenched in many systems in the United States–it may be covertly taught by teachers, parents may pass on outdated beliefs, and news and media can perpetuate harmful stereotypes. In order to understand how we as a clinical community ended up with a healthcare system that produces poor outcomes for racial and ethnic minorities, it’s important to first study the historical context of racism and healthcare. In this blog, I’ll take us through a timeline of racism in medicine, discuss how it unveils itself in the modern era, explain the social determinants of health, and discuss implicit and explicit bias and how to tackle it as a clinician.

History of racism in medicine

The Tuskegee Study

One of the most infamous medical studies ever to be conducted was the Tuskegee Study—a syphilis study that began in 1932 in conjunction with the U.S. Public Health Service. There were 600 male participants, 399 with syphilis and 201 without. The participants were told they were being treated for “bad blood,” a vague term used locally to refer to anemia, syphilis, and other diseases. However, the participants were actually given nothing for the condition.

None of the participants gave informed consent, meaning none of them knew exactly what they were being “treated” for or that they were simply being observed for syphilis throughout their lives. The study was supposed to last just 6 months, but it continued for 27 years after penicillin was declared the standard of care for syphilis. The men in this study lived their lives with syphilis even after a drug was readily available to treat the condition. The only compensation participants received were free meals and burial insurance.

La Operacion: forced sterilization

In 1898, the United States assumed governance of Puerto Rico and began an emphasis on population control to avoid a proclaimed social and economic decline. In 1935, Puerto Rico enacted Law 116, a eugenics sterilization law. By the 1970s, about 37% of Puerto Rican women of child-bearing age had been sterilized. The forced sterilization became so common, it was colloquially referred to as “la operacion.”

Informed consent was never obtained by these women and they were not presented with alternative forms of contraception either. Women were coerced into sterilization by industrial employers and door-to-door visits. The sterilization policy was heavily funded by the United States.

Law 116 was finally repealed in 1960. In the contiguous U. S., 30 states had laws similar to Law 116, all allowing government-mandated sterilizations. Similar to the Tuskegee Study, La Operacion in Puerto Rico and similar movements in the States set the tone for systemic mistrust of healthcare providers.

Racism in medicine today

Are there still racial disparities in healthcare?

The latest statistics on racism in healthcare give us a clear picture that we still have work to do.

According to the Centers for Disease Control (CDC)’s Health Disparities and Inequalities Report from 2013:

  • Rates of premature death (death before age 75 years) from stroke and coronary heart disease were higher among non-Hispanic Blacks than among Whites
  • Infant mortality rate for non-Hispanic Black women was more than double that for non-Hispanic White women
  • Rates for drug-induced deaths were highest among American Indians/Alaska Natives
  • DM prevalence was highest among males, non-Hispanic Blacks, those with mixed race, Hispanics, the disabled, and the poor
  • Women, minority racial/ethnic groups (except Asians/Pacific Islanders), the less educated, those who spoke a language besides English at home, and those with a disability were more likely to report fair or poor self-rated health, more physically unhealthy days, and more mentally unhealthy days than others

A 2020 American Cancer Society Report details the incidence rates per race for various cancers and the corresponding mortality rates.

  • Although incidence rates were the same between White and non-Hispanic Blacks, non-Hispanic Blacks died at a much higher rate for almost all cancers
  • Select cancers had a higher mortality rate for Hispanics and American Indians/Alaskan Natives

Social determinants of health

The six social determinants of health influence health outcomes including mortality, morbidity, life expectancy, healthcare expenditures, health status, and functional limitations.

  1. Economic stability: includes employment status, income amount, expenses, debt, medical bills, and support
  2. Physical environment: includes housing type and stability, transportation, safety, parks, playgrounds, walkability, and zip code/geography
  3. Education: includes literacy, language(s) spoken, early childhood education, vocational training, and higher education
  4. Food: includes hunger and access to healthy food options/food deserts
  5. Community and social context: includes social integration levels, support systems available to an individual, community engagement, discrimination, and stress
  6. Healthcare system: includes health coverage, provider availability, provider linguistic and cultural competency, and quality of care

    Each of the above determinants can skew your health outcomes.

    For example:

    • If you live in a food desert and work hectic hours, when will you make time to go out of town to a grocery store for fresh fruits and vegetables?
    • Do you have access to reliable public transportation that takes you to the grocery store?
    • Do you own a vehicle?
    • Do you have time between when your shift at work ends and when dinner needs to be on the table for your family?

    Due to the convenience and cheapness of fast food, many residents in a food desert may opt for unhealthy restaurant meals.

    Implicit vs. explicit bias in healthcare

    Is there racial bias in medicine today?

    Racial bias still exists in two forms: explicit and implicit.

    • Explicit bias: requires that a person be aware of his or her evaluation of a group, believes it to be true and correct, and then acts upon that belief intentionally. Explicit bias has decreased significantly over the past 50 years.
    • Implicit bias: operates unintentionally. It does not require a person to perceive it, endorse, it, or devote attention to its expression. Implicit bias is very common and persistent in society today.

    Studies of unconscious bias in medicine

    There have been multiple studies measuring the implicit biases of clinicians.

    A systematic review of 15 studies analyzing implicit bias compared the following groups:

    • Black and White (9)
    • Black/Hispanic and White (3)
    • Darker and lighter tones (1)
    • Hispanic and White (1)
    • Black/Hispanic/darker tone and White/lighter tone (1)

    The studies used double independent data extraction and assessed for the presence of bias as determined by the Implicit Association Test (IAT). Primary outcomes were an assessment of potential implicit biases held by health care providers, and if present, their impact on health outcomes.

    Implicit bias findings:

    • Low to moderate levels of bias among health care professionals in 14 of 15 studies
    • Low to moderate levels of negative association towards Black people in 13 of 14 studies
    • Similar levels of implicit bias against Black, Hispanic, and dark-skinned minorities
    • Significant relation found with the patient-provider relationship, patient health outcomes, treatment decisions, and adherence

    Race and medical treatment

    Those with more implicit ethnic/racial bias have poorer interpersonal interactions with minority individuals, often in very subtle ways. Here are a few specific differences:

    • Hispanic/Latinx patients were viewed as unlikely to accept responsibility for their own care and more likely to be non-compliant with the treatment recommendations.
    • African Americans receive worse care for heart failure and pneumonia.
    • African Americans get more preventive medicine from African American doctors, which could reduce the mortality rate if matched by 19%.
    • Meta-analyses covering 20 years shows African Americans are 22% less likely than control populations to receive pain medications for the same conditions.

    Cognitive bias in medical decision making

    Cognitive bias is the evaluation process by which one deviates from rational perception. There are two ends to the spectrum: inaccurate judgment and irrationality. Type 1 (“Intuitive”) Processing is reflexive and happens autonomously. There are over 100 well-defined biases impacting clinical decision making.

    Common cognitive biases in healthcare

    • Framing effect
    • Confirmation bias
    • Anchoring effect
    • Availability heuristic (recall bias)
    • Gambler’s fallacy
    • Observer-expectancy bias
    • Observer-selection bias
    • Fundamental attribution error

    Framing effect

    Drawing different conclusions from the same data depending on how it was presented.

    • An obese, diabetic, 40-year-old patient presents with substernal chest pain occurring after meals
    • A 40-year-old male with obesity, diabetes, and GERD presents with burning, substernal chest pain occurring after spicy meals

    Confirmation bias

    Looking for or interpreting data points that support one’s theory and glossing over those that don’t.

    • A 40-year-old patient with obesity, DM, and GERD presents with burning and substernal chest pain occurring after spicy meals
    • The patient improves after getting the GI cocktail; mild bump in Troponin and non-specific T wave inversions are written off

    Anchoring effect

    Relying disproportionately heavily on a data point when making interpretations/decisions.

    • An obese, diabetic, 40-year-old patient presents with sub-sternal chest pain occurring after meals
    • After a shift change in the ER, the patient is signed out to the oncoming physician as having GERD/reflux/esophagitis

    Availability heuristic (recall bias)

    The tendency to overestimate the likelihood of an event based on its recent occurrence.

    • A recent patient with a pulmonary embolism prompts you to think of PE in the next patient presenting with chest pain

    Gambler’s fallacy

    The belief that future probabilities are influenced by past events.

    • The patient is less likely to have pneumonia because of a recent spate of patients with pneumonia

    Observer-expectancy bias

    When one expects a particular outcome and subconsciously manipulates or misinterprets data in order to find it.

    Observer-selection bias

    The tendency to notice something more when something causes us to be aware of it.

    Fundamental attribution error

    The tendency for people to under-emphasize situational explanations for an individual’s observed behavior while overemphasizing dispositional and personality-based explanations for their behaviors.

    • Attributing good patient outcomes primarily to care delivered and bad patient outcomes to the patient’s underlying comorbidities or presumed non-adherence

    The financial cost of racial disparities

    14% of all elderly adults are uninsured as of 2014. 11% of White elders are uninsured compared to 20% of all elderly people of color. Such racial disparities result in $35 billion in excess healthcare expenditures, $10 billion in illness-related lost productivity, and $200 billion in premature deaths according to the Harvard Business Review.

    Telehealth & healthcare disparities

    The advantages of telehealth in combating healthcare disparities are many.

    • Telehealth can help overcome geographic/transportation limitations and is convenient–many patients don’t need to miss work or school for telehealth visits.
    • Telehealth often circumvents insurance and is less expensive overall than in-person visits.
    • Telehealth addresses regional clinician shortages, lack of available clinicians, and lack of quality care.
    • Telehealth can be a powerful tool in rural areas.
    • Telehealth can mitigate bias when clinicians are blind to race/ethnicity of a patient (depending on the modality: face-to-face video, phone call, or chat)

    Combating racial bias and disparities in medicine

    First and foremost, be thoughtful and consider the perspectives of others, especially patients. Understand your own biases. You can start by taking an implicit bias test. You can help establish an education program and rely on objective measures to remove subjectivity, too.

    Just because racial bias has existed in the fabric of our healthcare system—from labs and observational studies to patient rooms—does not mean it must continue. Years of collected data points toward a trend of declining health for various racial and ethnic groups. As clinicians, we are often the first point of contact and care for a patient. By understanding and improving upon our own biases, we can begin to transform the healthcare outcomes for people of color over time.

    - Rafid Fadul, MD, MBA, Executive Medical Director at Wheel


      At Wheel, our cultural value is to show empathy every day. Follow along with our series on diversity and inclusion in healthcare to learn more.