The language of race and ethnicity in academic medical publishing

If your research has been published in the Journal over the last 6–12 months, you might have received queries from us during the copyediting process about race and ethnicity in your manuscript: Did your participants self-identify their ethnicity, or was patient race designated by the research team? Did patients choose amongst a predetermined set of traditionally narrow racial categories, or were they permitted to answer in free text? Did the racial and ethnic categories into which patients were stratified impact your data or truly inform your conclusions? Was patient race and ethnicity a meaningful data point in your study, knowing that it is in many ways a social construct? Were you evaluating a treatment or outcome that was somehow race-dependent? While these questions go beyond the scope of traditional copyediting, they speak to a commitment on the part of the Journal of Osteopathic Medicine’s Editorial Board, reviewers, and staff to ensure that we are explicit about our obligation to ethical and equitable reporting in science, as well as to specificity and accuracy in the language we collectively use to describe research results.

Last year brought new phases and levels of reckoning around a variety of issues related to race and ethnicity, particularly in the American social sphere. It also, of course, brought us into a new era in the history of your Journal.¹ As we move forward in both arenas, it has become increasingly important that we speak openly and directly about our intentions, our ethos, and our priorities when it comes to race and ethnicity. To that end, I write this Editorial to document the JOM’s policies about patient ethnicity in osteopathic medical research. I fervently hope it will serve as a touchstone for future authors submitting their work to us, but also for the many osteopathic physicians who read JOM, whether they are involved in research or not. The old ways of conceptualizing, discussing, and writing about patient race and ethnicity – where analyzing and reporting race was a de facto portion of most scientific research (as well as patient treatment, in many cases²) – have been examined and found wanting.^3,4,5,6,7

Academic and scientific journals, particularly academic medical journals like JOM, are of course meant to be an enduring record of evidence-based research. But in the Platonic ideal (so to speak), a journal can exist in conversation with its reader-practitioners, acting not just as a megaphone that amplifies the research we receive, but as a telephone we can use to communicate changing ideas and standards to those whose research (and patient care) might benefit from new information. In 2016, Richard Smith, former Editor-in-Chief of British Medical Journal, published a blog post entitled, “What Are Medical Journals For and How Do They Fulfill That Function?”⁸ In later describing and evaluating Dr. Smith’s grading system, Dr. Nicholas Pimlott commented, “In what most of us consider to be their primary functions—selecting and publishing relevant, high-quality research and supporting continuing professional development—[Dr. Smith] rated journals poorly. Paradoxically, he rated medical journals highly for the very thing we least expect them to do: putting issues on the agenda—like, for example, the social determinants of health, climate change, and the effects of colonization on the health of Indigenous people.”⁹ As the journal of record for the osteopathic medical profession, we’d like to do exceptionally well at both.

First, I’d like to outline our guidelines and requirements around the language of race and (more appropriately in many cases) ethnicity. The American Medical Association’s Manual of Style¹⁰ is widely used in academic medical writing, and we rely heavily on their style recommendations as we prepare accepted manuscripts for publication. The most recent edition, released in 2020,¹⁰ contains rules specifically related to race/ethnicity, appropriately couched in the section on Inclusive Language. In a brief article on the AMA Style Insider blog, Iris Y. Lo – incidentally, a former manuscript editor at JAOA – described some of these rules: “AMA Style instructs authors to indicate who classified the race/ethnicity of study participants (i.e., the investigator or the participant)… The AMA Manual of Style [also] notes that we should avoid using “non-” (e.g., White and nonWhite participants) because it is a nonspecific ‘convenience’ grouping. Instead, editors can query the author about using a specific race/ethnicity or using multiracial or people of color to address the heterogeneous ethnic background of many people.” (Of note, Iris’ brief blog entry provides personal insight into how the language used to describe race and ethnicity impacts someone with an interesting combination of roles, as she is both editor and patient).

Building on those concepts, here is a list of AMA guidelines to which we subscribe – and which we have attempted to enforce with increasing consistency.

1. “Because many individuals may have mixed heritage, a racial or ethnic distinction should not be considered absolute, and ideally it should be based on a person’s self-designation.”¹¹

2. Authors must disclose who classified participants’ race/ethnicity, which categories or classifications were used, and whether the options were predetermined (and by whom).

3. In their Results, authors must use an equitable approach in reporting data. It is no longer sufficient to report only majorities and leave readers to calculate minority data, since this presupposes the majority (whether race or sex) as the default. Where the number of White participants is reported, so must the number of other participants be reported with equal clarity.

4. As noted above, “non-” convenience categories will not be used when discussing race or reporting results, in both text and Tables/Figures.

5. Racial and ethnic classifications are capitalized. (Interestingly, this is a relatively recent update to the AMA guidelines, one considered carefully by the Manual of Style committee and based on advice from a number of sources, including the National Institutes of Health.¹¹)

6. Black and African American are not equivalent, nor are Hispanic and Latino or similar racial categories. African American and Hispanic designations relate to countries of origin, and unless patients are specifically queried about country of birth, these categorizations are potentially inaccurate.

7. Perhaps most importantly, “the reasons that race/ethnicity were assessed in the study also should be described (e.g., in the Methods section and/or in table footnotes).”^11,,¹² It is no longer adequate to assess or report participant race and ethnicity without a burden of proof as to its necessity.

These guidelines are all given with one aim: to free medical research, and the language around it, from bias. While this list should not be considered comprehensive and is subject to additions, we are proud to document this set of guidelines. Our Instructions for Authors will also be updated accordingly to make more explicit what has been previously implicit in our copyediting protocols.

We encourage future authors to consider these rules carefully before they undertake research, evaluating how these language and reporting guidelines might ultimately inform their methods. If authors submitting to JOM have not considered these protocols prior to data collection, we ask that they make every conceivable effort to follow the “letter” of these rules by both updating the language in the final text of their manuscript before submitting, and also follow the “spirit” of the list by recalculating and reconfiguring data pertaining to racial and ethnic categories as needed. We also encourage all readers to consider how these rules, although specific to the publication of medical research, might also apply to patient care and any implicit assumptions or biases therein.

In closing, I draw your attention to two articles in this same issue.^13,,¹⁴ In a Special Communication in our Medical Education section,¹³ Dr. Nadege Dady and co-authors describe a program to recruit and support underrepresented minority medical students at Touro College of Osteopathic Medicine-New York. The program, undertaken in 2015, is called Creating Osteopathic Minority Physicians who Achieve Scholastic Success (COMPASS) and was designed to address the widening gap in representation amongst medical school candidates, especially Black or African American and Hispanic or Latino applicants. In a passionate Commentary in the same section, Dr. Kendra Gray and colleagues present a narrative history of racism and segregation in the medical education system and medical society membership, contending that we can only change the future by understanding the past. While these articles and the present Editorial do make this issue of JOM a topic-based one, they do not represent the full scope of the conversation on this crucial topic. In fact, we hope they won’t be the “last word”; rather, we call for more papers on this topic, especially original research and quantitative analyses.²