Journal Logo


The Other Global Pandemic

Scientific Racism and the Normality Bias

Chang, David C. PhD, MPH, MBA; Oseni, Tawakalitu O. MD; Strong, Bethany L. MD, MS; Molina, George MD, MPH; Ortega, Gezzer MD, MPH; Chen, Herbert MD§; Rogers, Selwyn O. Jr MD, MPH

Author Information
doi: 10.1097/SLA.0000000000005168
  • Free

The year 2020 witnessed the emergence and persistence of an invisible virus—COVID-19—that changed our world and has killed nearly 4 million people.1 Human to human airborne transmission was the mechanism of death. Science led to the pathogen identification, sequencing, and creation of mRNA vaccines, and complemented by public health measures including social distancing, mask wearing, and handwashing decreased transmission.

A more pernicious and longer lasting virus has lived among us for centuries and has led to countless lives lost. That virus is racism. The vector is the hate of “Other.” Although we are one human race, differential treatment by societal racial/ethnic categories holding one group superior to another has led to untold suffering. In the medical literature, we have identified race as a factor in disparate health outcomes. In fact, race in and of itself does not lead to disparities, but racism does.

In this editorial, we aim to highlight the destructive impact of racism in science and the limits of using race as an explanatory variable. We then will discuss how best to incorporate race/ethnicity in surgical research. Finally, we suggest remedies to the race conundrum.

Raise Awareness of Racism by Owning the Problem

How do we eliminate racism from medicine and medical science? We start by raising awareness of racism. Racism in the US has often been emphasized in a binary fashion of Whites versus Blacks, and other interactions are often omitted or deemed inconsequential. But we must acknowledge that it is really much more, and the language and terms used to describe individuals and groups continue to evolve. And although race is often used in health disparities research, it is really a proxy for the effects and impact of systemic racism. Racism is “a system of advantage based on race that involves systems and institutions, not just individual mindsets and actions,” and that “the critical variable in racism is the impact (outcomes) not the intent.”2 For that reason, while open hatred is a commonly recognized form of racism, indifference is perhaps more common and an important part of how racism works. The Nobel Peace Prize laureate Elie Wiesel once said, “The opposite of love is not hate, it's indifference.”3 And Dr. Martin Luther King Jr. expressed a similar sentiment when he addressed the indifference of White moderates, who were “more devoted to order than to justice.”4

It is this indifference that is endemic to medicine. A recent review found that of the more than 200,000 total articles published over the past 30 years in the New England Journal of Medicine, The Lancet, the Journal of the American Medical Association, and British Medical Journal, less than 1% included the word “racism” anywhere in the text.5 These top journals, as the key opinion leaders in medicine, make a statement of the relevance of racism by their conspicuous silence surrounding the topic in scientific discourse.

We need to collectively raise awareness of racism at all levels, at the societal, institutional, and personal levels. The “Black Lives Matter” movement of recent years has raised awareness at the societal level. At the health system level, we need to start with a different mindset and re-think all of the known problems in our healthcare system through a disparity lens. We need to own the problems, and stop attributing them—in essence, blaming them—on patients.6 And when we do so, we will begin to see that many so-called “patient problems” are really system problems. For example, instead of suggesting that patients have “access” barriers, we need to recognize that the real issue is mis-allocation, in where we choose to build our healthcare delivery sites. Instead of saying that patients have mistrust, we need to understand that trust needs to be earned. Instead of saying that patients are non-compliant, we need to re-examine our communication strategies. From interpreter services for non-English speakers to health literacy awareness for English speakers, there are a myriad of ways to provide effective communication. Trauma systems provide a good example of this type of thinking. Efforts to reduce deaths from motor vehicle injuries have focused on developing EMS and trauma networks that at least try to save patients irrespective of seat belt or helmet usage. We need to similarly build a healthcare system that strive to deliver equivalent outcomes regardless of patient's social class or demographic factors. Another example of “owning the problem” is the efforts surrounding reducing medical errors; for example, the operating room time-out protocols remind us that we all have a part to play in reducing errors. Journal editors could help by “taking the pledge” to help raise awareness, committing to diversify their editorial boards, and regularly dedicating journal themes to addressing systemic racism in medicine and not merely reporting on racial disparities.7,8 Journal editors should also monitor reviewer comments and scoring of manuscripts that discuss racism in medicine. Organizations should consider implicit bias training.9 And we should all evaluate our personal biases and prejudices. Comics can be an innovative tool in anti-racism education, by starting with “stories, not numbers” and creating parallel and contrasting visual narratives—for example, by contrasting the societal response to cystic fibrosis versus sickle cell disease. By balancing levity with seriousness, comics can help diffuse tension and create psychologically safe space for conversation about racism.10

Institutional racism and the normality bias in research

While some have argued that race should be eliminated from clinical practice and medical research completely—which has led to alarming decision by some health care institutions to remove race from GFR calculations which would ignore potential population differences without considering the clinical performance characteristics or consequences for Black patients11—a more prudent approach may be to keep race, but to recalibrate how race is used.11–13 The primary concern is that eliminating race/ethnicity may further solidify institutional racism. US and Europe have dominated the medical literature over the last 200 years, one of the unfortunate consequences of colonialism. This lack of representation of non-White patients has forced patients around the world to adopt standards developed in relatively small US/European White patient populations. Just as a lack of diversity in television and entertainment industry has distorted our sense of beauty, this lack of diversity in science distorts our sense of health and healthcare. Without race data in research, we will be blinded to the continued problem of lack of diversity and representation in medical science. This has led to the “normality bias”, or the normalization of data and concepts that are prevalent in White populations while pathologizing conditions in non-White populations. Simple and neutral differences have become distorted “evidence” of normality versus pathology, or worse, superiority versus inferiority.14 There are many examples of this bias. A recent analysis found that most cancer guidelines have been based on publications that included very few non-White patients.15 Also, our understanding of the shape of a “normal” pelvis is mostly based on anatomical studies of American and European populations, which has led to unnecessary “treatments” to “correct” fetal progression during birth.16,17 And what are considered to be “risk factors” of glaucoma are actually common ophthalmologic traits in Asian patients, leading to inappropriate prescription of medications to young healthy individuals to reduce intraocular pressure when no disease really exists.18 And white blood cell counts in normal healthy Black military recruits are lower than the White population, leading to the possibility that the current definition of leukocytosis and other indicators of infections may be inappropriately high for Black patients.19,20 Outside of clinical medicine, the normality bias has led to a pathologizing of non-European diets. The perception that Mediterranean diet is the “healthiest” form of diet,21,22 while criticisms abound of diets from other cultures, has nothing to do whether comparative studies have been performed, but because of the absence of study on diets of non-White cultures while NEJM published a flawed study touting the purported benefits of Mediterranean diet.23 IQ tests is another example of normalizing “knowledge” from White populations at the expense of non-White populations.24 Similar concerns about lack of diversity have been raised with artificial intelligence research, and has been dubbed “techno-racism.”25 One example involves self-driving cars that have been reported to avoid White pedestrians more than non-White pedestrians, because there were insufficient non-White subjects included in their development.26

How can we recalibrate race-based research so that it does not contribute to racism?

If we keep race in research, how, then, can we recalibrate race-based thinking so that it does not contribute to racism? Unfortunately, the healthcare disparity literature, to date, has made limited progress. They have mostly focused on documenting differences between groups; although it is an important first step, it has not addressed the deeper causes or potential solutions. We need to make some modifications in how we do disparity research.

A common approach in the disparity literature to date is to include race as a covariate. But that can easily lead to misinterpretation that race is a risk factor rather than other covariates that are associated with race. A more appropriate approach may be to consider the possibility of heterogeneity of treatment effects, and to examine for interaction effects with race, or to stratify the analysis on race. With this approach, race is not explicitly a variable, thus avoiding the misinterpretation that race is a risk factor. And it will show us that there could be different sets of risk factors than Whites, such as social determinants of health. The data from this approach have the potential to lead to a more correct message regarding race and health: That race is NOT a risk factor, but that patients may have different risk factors based on their race. Such an analytic approach has led to the recommendation that breast cancer screening guidelines should be different for different racial groups.27 And by developing different race-specific scales and guidelines, this approach could prevent the application of lung- and kidney-function equations derived in White patients to Black patients which would ignore long-recognized racial/ethnic differences in normal physiological function or biomarkers.11 In a way, this line of inquiry starts with race, and looks for additional risk factors that are associated with race, as opposed to stopping at race and labeling it as a risk factor.

Another simple but powerful approach is to consider a different reference group in the reporting of research findings. Since the White population is still dominant in US and Europe, most research studies would default to White patients as the reference group, and report findings in other populations relative to White patients. While this is done mainly for statistical reason, it very subtly promotes the White population as the “normal” group. Although mathematically identical, switching the reference group to one of the non-White populations could help introduce and reinforce the message that “reference” does not mean “normal”, and that non-White patients could, in fact, be the normal group. When differences in prevalence rates of ADHD were observed among school-age children from different racial groups, a popular interpretation was to consider the rates in White children to be normal, and thus more must need to be done to increase detection among non-White patients; however, an equally plausible explanation could be that the rates observed among non-White patients were normal, while White patients were simply being over-diagnosed.28 Similar concerns have been raised regarding opioid use for pain control between races, about whether the problem is an overuse among Whites, or under-use among non-Whites, or both.29

In implementation research, we need to be careful with a “one size fits all” approach. We need to focus on developing new interventions that are culturally appropriate for different racial groups. The literature has often focused on expanded implementation of current interventions to serve non-White patients. However, in some cases, de-implementation of treatments and interventions that were tested in White patients but inadequately tested in non-White patients may be as important as implementation. For example, uniform DVT prophylaxis guidelines after major surgery may need to be revisited for different racial groups.30

A long-term solution may only come when the scientific community borrows a lesson from civil society. Identifying racial differences by itself does not lead to biases, although there have been many unfortunate examples in the past. What ultimately leads to biases is flawed interpretation of objective facts. And these flawed interpretations are due, ultimately, to the fact that the scientific fields have not had a diverse workforce. It is hard to imagine a scenario in which phrenology and craniometry, and other studies of racial differences, would have been able to uphold such unfounded racial conclusions for so long, if non-White scientists had been involved in the research. Likewise, there were probably few LGBTQ scientists when homosexuality was deemed a disease. The human experiment with democracy has demonstrated that broad-based participation in the legislative process is important to ensuring justice. We need representation from all segments of society that represent diverse interests—as well as prejudices. The same needs to happen with science. Broad-based participation in the scientific process is the best way to safeguard against interpreting scientific evidence to support racist or prejudiced biases. We need to dispel the myth that the scientific process is objective, and we need to encourage physician scientists from diverse backgrounds, including under-represented minority groups, to be interested in becoming physician-scientists. The democratization of society with regards to historically marginalized populations led to the revision of discriminatory laws, albeit with a lag. It is only through a parallel democratization in the production and interpretation of science that we can ensure healthcare professionals do not unknowingly become the supporter for social biases. However, non-White scientists, within and outside of medicine, often face challenges in finding acceptance.31 We must redouble our efforts at recruiting not only non-white physician to deliver culturally competent care, but also more non-white physician-scientists to produce culturally competent medical science.13


The United States, despite our challenge with racism, is actually on the vanguard of this new frontier because we are starting the difficult conversations surrounding race. Other countries with various ethnic groups should consider our example and use this opportunity to examine their own challenges in optimizing healthcare delivery for all within their borders.32 Although we have made some progress in eliminating racism, it has become a sort of fever of unknown origin in our society, and continues to rear its ugly head often from unknown nidi of infection. Science, with its aura of objectivity, has become one of these nidi of infection. With the recent momentum in the US, we have a real opportunity to lead the world in eliminating this safe harbor for the racism thought virus.


1. World Health Organization. WHO Coronavirus (COVID-19) Dashboard. Available at: June 12, 2021.
2. National Museum of African American History and Culture. Talking about race. Available at: Downloaded May 8, 2021.
3. Sanoff AP. One Must Not Forget. U.S. News and World Report, 27 October 1986.
4. King ML Jr. Letter from the Birmingham Jail. San Francisco: Harper San Francisco; 1994.
5. Krieger N, Boyd RW, De Maio F, Maybank A. Medicine's Privileged Gatekeepers: Producing Harmful Ignorance About Racism And Health. Health Affairs Blog, April 20, 2021; Available at: April 22, 2021.
6. Calhoun A. The pathophysiology of racial disparities. N Engl J Med 2021; 384:e78.
7. American College of Surgeons. Journals That Have Agreed to the Pledge. Available at: Downloaded May 8, 2021.
8. Lillemoe KD. Annals of surgery takes the challenge. Ann Surg 2021; 273:1–2.
9. Hall WJ, Chapman MV, Lee KM, et al. Implicit racial/ethnic bias among health care professionals and its influence on health care outcomes: a systematic review. Am J Public Health 2015; 105:e60–e76.
10. Obuobi S, Vela MB, Callender B. Comics as anti-racist education and advocacy. Lancet 2021; 397:1615–1617.
11. Borrell LN, Elhawary JR, Fuentes-Afflick E, et al. Race and genetic ancestry in medicine - a time for reckoning with racism. N Engl J Med 2021; 384:474–480.
12. Ioannidis JPA, Powe NR, Yancy C. Recalibrating the use of race in medical research. JAMA 2021; 325:623–624.
13. Chang DC, López DB, Molina G. Culturally Competent Science. JAMA Surg 2018; 153:699–700.
14. Harmon A. Why White Supremacists Are Chugging Milk (and Why Geneticists Are Alarmed). New York Times, October 17, 2018. Available at: Downloaded May 8, 2021.
15. Perez NP, Baez YA, Stapleton SM, et al. Racially conscious cancer screening guidelines: a path towards culturally competent science. Ann Surg 2020.
16. Betti L, Manica A. Human variation in the shape of the birth canal is significant and geographically structured. Proc Biol Sci 2018; 285:20181807.
17. Yin S. Why Textbooks May Need to Update What They Say About Birth Canals. New York Times, October 27, 2018. Available at: Downloaded May 8, 2021.
18. Doss EL, Doss L, Han Y, et al. Risk factors for glaucoma suspicion in healthy young Asian and Caucasian Americans. J Ophthalmol 2014; 2014:726760.
19. Reed WW, Diehl LF. Leukopenia, neutropenia, and reduced hemoglobin levels in healthy American blacks. Arch Intern Med 1991; 151:501–505.
20. Chang DC, Cornwell EE 3rd, Phillips J, et al. Early leukocytosis in trauma patients: what difference does it make? Curr Surg 2003; 60:632–635.
21. U.S. News and World Report. Mediterranean Diet: #1 in Best Diets Overall. U.S. News and World Report, Available at: Downloaded June 14, 2021.
22. Greenberg P. How to Get America on the Mediterranean Diet. New York Times, July 19, 2018. Available at: Downloaded June 14, 2021.
23. Estruch R, Ros E, Salas-Salvadó J, et al. PREDIMED study investigators. primary prevention of cardiovascular disease with a mediterranean diet supplemented with extra-virgin olive oil or nuts. N Engl J Med 2018; 378:e34.
24. British Broadcasting Corporation. NFL agrees to drop race bias in concussion claims. BBC News, June 3, 2021. Available at: Downloaded June 14, 2021.
25. Karimi F. People of color have a new enemy: Techno-racism. CNN, May 9, 2021. Available at: Downloaded May 10, 2021.
26. Samuel S. A new study finds a potential risk with self-driving cars: failure to detect dark-skinned pedestrians. Vox, March 6, 2019. Available at: Downloaded May 10, 2021.
27. Stapleton SM, Oseni TO, Bababekov YJ, et al. Race/ethnicity and age distribution of breast cancer diagnosis in the United States. JAMA Surg 2018; 153:594–595.
28. Morgan PL, Staff J, Hillemeier MM, et al. Racial and ethnic disparities in ADHD diagnosis from kindergarten to eighth grade. Pediatrics 2013; 132:85–93.
29. Chen I, Kurz J, Pasanen M, et al. Racial differences in opioid use for chronic nonmalignant pain. J Gen Intern Med 2005; 20:593–598.
30. Chen YW, Fong ZV, Qadan M, Kunitake H, Mullen JT, Chang DC. Should all patients receive the same prophylaxis? Racial variation in the risk of venous thromboembolism after major abdominal operations. Am J Surg. 2021 Jun 10.
31. Tiku N. Google hired Timnit Gebru to be an outspoken critic of unethical AI. Then she was fired for it. Washington Post, Dec. 23, 2020. Available at: May 10, 2021.
32. Hagiwara N, Kron FW, Scerbo MW, et al. A call for grounding implicit bias training in clinical and translational frameworks. Lancet 2020; 395:1457–1460.

normality bias; scientific racism

Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.