About the SA Blog Network

Unofficial Prognosis

Unofficial Prognosis

Perceptions and prescriptions of a medical student
Unofficial Prognosis Home

When should medicine talk about race?

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

Race is everywhere in medicine. Most health statistics are broken down by race. We routinely characterize diseases by which populations they affect more and less and medications by which ethnicities respond better or worse.

It’s so ubiquitous that it’s easy to take for granted as justified. But the use of race in medicine is a subject that is vigorously debated. Whenever a new study comes out stratifying results by race, there are inevitably supporters and critics.

The question under debate: is there a place for race in medicine?

There’s a growing number who say we should toss this way of thinking entirely. Many scholars now contend that race is closer to a social construct than a biological category, and there’s the legitimate fear that pointing out differences between races sends the message that the difference is biological. Even if there are certain genetic differences among populations, we know that self-reported race is at best a crude proxy for indicating them. Moreover, studies often do not adjust for all other variables besides genetics, such as socioeconomic status, culture, and discrimination – meaning if differences are shown, the knee-jerk tendency to think biology might overshadow important environmental disparities that deserve our attention. There are social concerns too, in that historically ethnicity in research has been abused by pseudoscientists with racist agendas of demonstrating the superiority of certain people over others. In light of that history, profound sensitivity toward using race as a variable in medicine is understandable and warranted.

Part of the problem may be that some simply do not give it enough thought. There are some who stratify any data they collect on any health-related subject by race because that’s what others did before them, along with others before that. But when you do any data analysis, you need to justify its being done. There’s no such thing as just “laying out the facts” because there is no such thing as a predetermined set of facts that we either expose or hide. We make choices with everything. Collecting, breaking down, and representing data all involve choices. When comparing groups, we can draw the lines wherever we want. Telling of this point is that many studies that talk about race still only compare blacks to whites, ignoring all other groups along with cases of mixed ancestry.

When the choice lies with the researcher, she has an obligation to use it responsibly. As such, it’s not enough to enough to justify a project with some ambiguous version of: “this will contribute to the literature by showing something we do not know.” We don’t know infinite numbers of things. Research has to have value. At the forefront of every decision should be the questions: What’s the point? Are the differences I’m trying to show relevant to anything? Are there implications for disease prevention, diagnosis, management, or treatment?

Sometimes, indeed the answer is yes. There have been cases where thinking about race, even as a rough guide, have led to benefits for patients. Knowing that sickle cell anemia is more prevalent among populations of sub-Saharan African ancestry can tip physicians off for earlier and thereby more effective diagnosis and management. Since Tay-Sachs is a genetic disease with increased prevalence among Ashkenazi Jews, Jewish communities early on welcomed genetic testing for prospective parents and by doing so dramatically reduced the incidence of the disease. Individuals of Asian descent are more likely to carry certain genetic polymorphisms resulting in slower drug metabolism – meaning patients need lower doses to achieve the desired effects and avoid toxicity. There are many more examples. While it is such an important point that I’ll say it again – that race is only a very imperfect proxy for genetics – there has been demonstrated medical value in being aware of these trends.

The reason is that medicine is a field that uses heuristics – simple “rules of thumb” that help home in on best guesses when comprehensive searches are not feasible. These shortcuts are so frequently employed because medicine is the perfect storm of information overload combined with limited time. Best guesses in medicine are probabilistic; doctors collect clues from various sources to select more likely and less likely options. Every test, every new piece of information contributes to that ranking. Thus, some argue that just as doctors clue into best guesses based on a patient’s constellation of symptoms and test results, so too can race be used as an approximate guide. With the recognition that heuristics can lead to biases, the solution is not to discard them but rather to make doctors more cognizant of biases so they can work to eliminate them and use heuristics more effectively.

The use of race in medicine is a deeply sensitive issue and should be treated as such. One thing to note is that in contrast to shameful periods in history that focused on race with unethical agendas, the vast majority of current research is completely well-intentioned, toward the goal of optimally tailoring medical care to a diverse patient population. Those on both extremes of the debate are looking out for patients. So where does that leave us? While there is a place for race in medicine, the literature also remains rife with studies with seem to point out differences with no valid reason for pointing out differences, and my sense is that there’s a greater tendency to overuse race when it’s not appropriate than to neglect it when it is. The burden should be on every medical researcher who wants to talk about race to be explicit as to what contribution this data would make to the world. And, if those measures fail, it would behoove readers and patients to apply just as critical an eye.

Ilana Yurkiewicz About the Author: Ilana Yurkiewicz is a fourth-year student at Harvard Medical School who graduated from Yale University with a B.S. in biology. She was an AAAS Mass Media Fellow, and her work has appeared in the New England Journal of Medicine, Aeon Magazine, Science Progress, The News & Observer, and The Best Science Writing Online 2013. She has an academic interest in bioethics, currently conducting ethics research at Harvard after previously interning at the Presidential Commission for the Study of Bioethical Issues. She is going into internal medicine and is also interested in quality and systems improvement. Follow on Twitter @ilanayurkiewicz.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 11 Comments

Add Comment
  1. 1. priddseren 9:23 pm 08/25/2012

    Race is only relevant if there is some sort of medical condition or disease that is in fact specific to a race. There are some of those out there. There are also certain effects certain races of people have that is of interest. Such as reviewing why certain peoples live longer or something like that. This is the same reasoning behind why we need to consider women and men separately. Since both genders have different parts, different levels of hormones and etc… we should look at genders and race, where it makes sense to do so. Age is another legitimate classification.

    Additionally, culture, not necessarily race is of interest because groups of people sometimes do activities or ingest foods or other substances unique to the group and may cause problems or benefits the world may want to know about.

    However, to just separate everything by race because of idiot government requirements, moronic concepts of diversity and simply just doing it for no reason other than coming up with excuses to form what are otherwise ridiculous opinions, such as X race is somehow deficient because they have 1000th of a percent more occurances of something like the Flu or for example, the government basically separating everything by race and because of government intrusion in the healthcare industry, the government simply forces everyone to list everything by race.

    None of that is reasonable and should be discouraged.

    Link to this
  2. 2. way2ec 10:59 pm 08/25/2012

    Thank you for using examples that were in fact not “racial”, sub-Saharan African ancestry and Jewish communities. The third group mentioned was Asian and given that Asia is a geo-cultural invention, no wonder so many scientists used only the “black and white” “racial” comparisons. Thank you for also making clear that mixed ancestry defies racial profiling as well. Ancestry, ethnicity, the genetic pools of our origins can and should be important in scientific and medical research. But we have far to go to overcome the “science” of red, black, white, yellow, brown “races”; the Eurasian continent being divided at the edge of the “white” Russians; Caucasoid, Caucasian, and Mongoloid; and being racially defined as “black” with only one sixteenth “black” ancestry. Pacific Islanders but not the Pacific Islands of Japan, Native American but not Alaska Natives, Hispanics white or non-white, each “racial” category as unscientific as the next. Will the use of genetic markers rescue our medical researchers from the shameless practices of the past?

    Link to this
  3. 3. Fanandala 7:35 am 08/26/2012

    “Origin” should be a more acceptable classification than “race”. Not all Africans are black, not all Asians are mongoloid. In a different landscapes similar looking people had to face different environmental pressures. As our statistical tools improve we might be able to pinpoint the effects of treatments much more than at present to people of certain origins. We might even be able to tailor make a treatment to people who come from one side of the mountain or from the other, even if they look quite similar.
    Another factor that weighs heavily on health is wealth and education. One day we might figure it all out.
    But being politically correct will not get us anywhere.

    Link to this
  4. 4. AllanRBrewer 7:38 am 08/26/2012

    Statistical analysis produces a number describing the level of significance of the proposed difference – should we not let that speak for itself?

    Link to this
  5. 5. mihondo 8:35 am 08/26/2012

    Heuristics for a medical diagnosis (or any other problem) are selected because a) they are easy to observe, b) they have had some success in the past, and c) there isn’t any better way of doing it.

    Race can be a marker for both genetics and social behaviors. As genetic analysis becomes available (cheaper, faster, ubiquitous), it can / should /will be used. Using race as a marker for social behaviors may be less justifiable.

    The real issue is confusing attributes that are correlated with race (due to some other correlating factor such as socioeconomic status), rather than race being a causitive factor.

    In diagnosis, even correlating factors are useful.

    On the other hand, if you are trying to find a cure, you need to eliminate the correlated factors to find a ‘root cause’.

    Link to this
  6. 6. GGlynn2 11:14 am 08/27/2012

    God does not compete with man; therefore race has no place in medicine.

    Link to this
  7. 7. Optimtech 3:20 pm 08/27/2012

    I don’t think race is everywhere in medicine. Because race is only for adventure.

    Link to this
  8. 8. HubertB 6:20 pm 08/28/2012

    A much lower percentage of African Americans respond to Hepatitis C treatment than do White Americans. Perhaps some other drug or combination would do better.
    Some Americans have had a severe reaction when transfused with blood in Russia. Perhaps people from that part of Russia should not be blood donors.

    Link to this
  9. 9. tucanofulano 8:43 pm 08/28/2012

    Race? Well, maybe it is.

    Link to this
  10. 10. Padgie 11:43 pm 08/28/2012

    Race would seem to be one of the least precise measuring tools available. Ethnic groups may have some validity but endevouring to slot someone into one of a few groups is dodgy. Even assuming that Afro and Caucasian are singularly well identified groups, where is the line between Afro and Afro/American and American?

    Link to this
  11. 11. VivaLaEvolucion 1:39 am 08/31/2012

    One thing that should not be overlooked when doing a study on race and disease risk is the diet of each race being studied. For instance, poor people of whatever race probably have a different diet, for better for for worse, than wealthy people of whatever race. Diet and lifestyle have a big influence on the disease risk in a population no matter what the race. So, if a study is done comparing disease risk of different races, and those races had a different diet and lifestyle, then one couldn’t say it was necessarily race that put a population at higher risk for certain disease, as it may have been diet or lifestyle that put them at the higher disease risk. I got my genetic test at and it is pretty cool to see what disease risk you have, even if it isn’t 100% or anything. I think that as humans continue to inter-racially breed “race” will become less meaningful, and specific genetic mutations, better or for worse, will be used more to identify people. (like the Sickle Cell Gene vs. beneficial Italian Al-Milano Gene mutation which is more effective at removing cholesterol than average person.)

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article