Academic medicine is increasing recent efforts to unequivocally identify systemic racism as a public health crisis because it drives health inequity to racially minoritized groups [1,2,3]. We write today as three medical professionals across the spectrum of social justice education — Drs. Black and Spearman-McCarthy as Director of Social Justice and Health Equity (SJHE) education and Associate Dean for Learner Diversity, Equity, and Inclusion, respectively, and Nishita Pondugula as a Yale School of Medicine (YSM) student leader for medical school reform centered on social justice and diversity under our tutelage. In our personal experience and expertise, a health equity emphasis in clinical education, practice, and research differs from a disparities approach because it seeks to dismantle the systems of racism that create inequitable health outcomes in the first place. Therefore, medical education, practice, and research are slowly transitioning from a lens of health disparities to one of health equity. In order to support this transition, authors and journals must restructure the depiction of health inequities caused by racism.

Nearly 200 years ago, Dr. Rudolph Virchow was a pioneer of social medicine who believed that poor health originates in social inequality. He once said:

Medicine is a social science, and politics nothing but medicine at a larger scale [4].

Science for its own sake usually means nothing more than science for the sake of the people who happen to be pursuing it. Knowledge which is unable to support action is not genuine – and how unsure is activity without understanding… [5].

Race and racialization are social and political constructs [6], and Virchow taught that wellness and healthcare are also sociopolitical processes [4]. Virchow argued that scientific and medical knowledge must name the drivers of disease to support targeted action capable of ending that disease. Otherwise, knowledge that cannot inform solutions is not worth knowing and poorly informed actions are not worth doing [5]. In the American context, healthcare privilege is conferred to those Americans racialized as white by depriving resources and creating health inequity to all others [6]. In that light, one may apply Virchow’s lessons of social inequalities as drivers of disease to the American landscape of racialized health inequity. Many racialized health inequities in the USA have been known for decades. However, academic medicine, individual clinicians, and larger healthcare systems have not yet supported action towards sufficient and meaningful solutions, as evidenced by the persistence of racialized health inequities over time [7]. In applying Virchow’s framework to this stagnation of racial justice progress, one could argue that the knowledge disseminated by academic medical literature is efficacious towards solutions only if societal and healthcare racism are clearly named as the driver of each inequitable outcome.

Yet, health disparities literature stylistically omits the driver of inequity quite frequently, which perpetuates colorblind and aversive racism by rendering such knowledge unable to support action. More specifically, colorblind racism occurs when the observed consequences of systemic racism are described using nonracial language, justifications, and explanations [8]. Aversive racism occurs when healthcare professionals explicitly denounce racism even while performing behaviors or upholding policies and practices that manifest racism [9]. We argue that the stylistic omission of racism in academic writing sustains the disconnect we have between the extensive knowledge about disparities and true progress towards equity.

Stylistic Precedents in Academic Medical Writing

Many academic medical journals and major style manuals already uphold stylistically writing in the active voice as the benchmark for quality authorship [10], but these benchmarks appear unevenly applied as it relates to racial health equity. This notably includes style manuals like the American Medical Association (AMA) Manual of Style and Publication Manual for the American Psychological Association as well as prominent journals like Nature, Science, and British Medical Journal [10]. To begin, the AMA Manual of Style informs, “In the active voice, the subject does the acting; in the passive voice, the subject is acted on [11].” Let us explore further (Fig. 1).

Fig. 1
figure 1

Stylistic strategies for racial health equity

Full size image

Active Voice: The cat caught the mouse.

The active voice clearly identifies the subject of the sentence, the cat, as the agent performing the verb, caught. The reader’s attention focuses directly on the subject performing the action.

Passive Voice: The mouse was caught by the cat.

The passive voice redirects the reader’s attention to the recipient of the agent’s action as the subject — ironically named “the semantic patient” — being the mouse [12]. The prepositional phrase starting with “by” contains the agent.

Passive Voice Without Agent: The mouse was caught.

Here, attention remains on the patient (mouse), but the action is driverless because the sentence drops the prepositional phrase containing the agent (cat). This style may be appropriate if the agent is unimportant, obvious, and/or well-known [13]. Yet, the agent is not obvious in this example. Was the mouse caught by a mouse trap, a cat, a bird, a snake, an exterminator? Or, did the mouse get stuck beneath a bowl? Neither is the agent unimportant, as omitting the agent requires the reader to make flawed assumptions about where to start if they wished to increase or decrease the chances of capturing the next mouse. Testing hypothesis featuring birds or bowls would be useless if we are actually needing interventions featuring cats. Furthermore, one need not spend extensive time or resources studying the behaviors and habits of mice to try to deduce from scratch who or what is catching them. We already know the agent from this example: its cats. As in Virchow’s teachings, all other endeavors would produce knowledge that would be “unsure” and “not genuine” because none of it would center the key relationship between cats and mice.

The AMA Manual of Style similarly explicitly instructs authors to “use the active voice, except in instances in which the actor is unknown or the interest focuses on what is acted on [11].” In Virchow’s framework expanded to the context of American medical racism, racism is far from “unknown” or uninteresting as the driver of harm to minoritized populations, which notably includes iatrogenic patient harms born of medicine’s own racial bias [14]. Furthermore, just last year, the AMA pledged to “Embed equity and racial justice throughout the AMA by expanding capacity for understanding and implementing anti-racist equity strategies [1].” Similarly, given that the faculty authors of this article are both psychiatrists, the American Psychiatric Association (APA) apologized for its racist legacy and committed to undoing racism in psychiatry in January 2021 [15].

Putting it all together, writing styles that fail to explicitly name racism as the cause of racialized health inequity are limited in their ability to contribute to a knowledge base that is capable of informing solutions to these societal inequities. To illustrate problematic writing styles common to disparities research, we adapt examples from the APA’s Mental Health Disparities for African Americans [16]. The website unfortunately fails to explicitly name “racism” even once, as it was first published before the public apology and commitment to undo embedded racism. The following categories are based on our observations.

Passive Comparison Without Agent: Mice are more likely to be caught than cats.

The patient (mice) receives action from an unknown, omitted agent. However, the agent (cats) is vaguely referenced as a normative comparative group instead of the driver. Here, the comparative group is less impacted by the omitted agent’s action than the patient: cats typically catch mice, not each other. Readers of this sentence would not know that cats are the ones catching the mice unless they already knew this information from elsewhere.

APA Disparity Example: Compared with the general population, African Americans are less likely to be offered evidence-based medication therapy.

This sentence does not convey knowledge that providers’ racial prejudice drives this inequity because iatrogenic racism is omitted entirely. The sentence instead focuses on racially minoritized patients (being the semantic patients) being comparatively “inferior” to a privileged, “normative,” predominately white group that is unspoken to be less impacted by iatrogenic racism (the agent’s actions). Traditional disparity research frequently compares racially minoritized groups to privileged, white groups without intentionally naming the agents of racism upholding white privilege and harming the racially minoritized. Furthermore, omitting the agent from this sentence fails to critically ask or answer: who is the one failing to offer African Americans standard of care? The answer may be more uncomfortable for clinicians to recognize, as we are the ones committing this wrongdoing. Rather than naming and spotlighting the providers who are failing to offer evidence-based medication therapy due to medical racism, the stylistic writing of much current disparity research requires scholars to make flawed assumptions about the mechanisms of inequity — thereby limiting solutions. Worse, spotlighting African Americans instead of racism prompts readers to postulate biased ideas about whether or not African Americans have inherent traits that somehow deserve or precipitate subpar treatment, perhaps like their culture or personal choices or “biological differences.” Efforts would be better spent postulating, implementing, and improving how to dismantle medical racism.

Statistic/Outcome Without Agent: Mice make up 14% of the country’s caught animals.

Whether using an active or passive voice, this style presents quantified outcomes/statistics about the patient while omitting the agents driving the outcome. The quantified data are factual (or rather fictional for this illustrative example), but the facts again do nothing to point the reader where to start if they wished to increase or decrease the chances of capturing the next mouse.

APA Disparity Example: CDC estimates that African Americans represented more than one-third (40%) of all people living with HIV in 2015.

The sentence highlights a quantified inequity that racially minoritized patients are enduring. The above example excludes the well-known agents of systemic racism creating and sustaining that disparity at the onset. However, medical racism is far from being unknown or uninteresting like constructions that appropriately use the passive voice without an agent. Yes, the statistic described about HIV is technically true. However, omitting the agent of racism from this sentence fails to offer clarity on how or why or what factors are driving this harm. Racism versus privilege is unspoken, which leaves readers to generate their own rationalizations as they read these driverless statistics. These self-generated explanations may or may not be historically accurate or evidence-based because they will vary according to each individual’s social identity, personal prejudices, lived experience, and preexisting knowledge of social medicine. For example, as it relates to HIV in the cited example, evidence suggests that provider-level stigma against persons living with HIV varies according to each person’s gender, race, and clinical specialty, and these individual biases negatively impact patient care [17]. Thus, the void left by the driverless inequity risks being filled with biased and false assumptions that would further detriment the minoritized as opposed to treating them. In short, the knowledge conveyed by this disparity example sentence portrays inequity without identifying a historically-informed, unbiased, evidence-based target towards equitable solutions.

Implications for Disparities vs Health Equity Academic Writing

Once one’s perceptual acuity is attuned to the problematic use of writing styles that omit the drivers of racism in academic medicine, the prevalence of such constructions may be recognized to be widespread. Virchow’s principles clearly demand that medical and scientific knowledge be pursued and published in a manner that communicates the social changes necessary to eliminate health inequity. One may naturally begin pondering how academic medicine may have devolved into the common use of agentless writing styles of racialized health inequity.

To begin this exploration, the concept of aversive racism describes instances when healthcare professionals perform behaviors or uphold policies and practices that manifest racism while being starkly resistant to acknowledging how their thoughts, behaviors, and policies embody racism [9]. For centuries, medical racism wrongfully ascribed health inequities between white and racially minoritized populations to the genetic and biological superiority of white Americans [6]. More recent decades of academic medicine attempted to move beyond race-based medicine into more socially conscious explanations. Yet, the act of healthcare providers assuming responsibility for mistreating subgroups our patients is aversive to our personal identities and society’s expectations for us to be healers. Otherwise stated, openly acknowledging how our own iatrogenic racial prejudice harms racially minoritized patients is unpleasant, distasteful, and unwanted. And so, as academic medical researchers quantified and reported the socially mediated harms and deaths of racially minoritized people in health disparities literature over the decades, the stylistic presentation of these reports might have understandably — albeit unknowingly — evolved to favor writing constructions that minimized our personal and collective wrongdoing and instead highlighted the experiences of the minoritized. To be clear, writing in the active voice is already the standard of academic writing as endorsed by multiple aforementioned authorities. We are highlighting how aversive racism influences academic writing to deviate from these norms to avoid accountability to iatrogenic racial harms.

While aversive racism may be influencing the avoidance of iatrogenic accountability, colorblind racism provides the mechanism. Writing constructions that stylistically omit or fail to clearly identify personal and collective racism as the agent of health inequity commit faults of colorblind racism, where the observed consequences of systemic racism are described using nonracial language, justifications, and explanations [8]. For example, many have celebrated America’s “healthcare heroes” who have been on the frontlines of battling the COVID-19 pandemic. Yet, few are intentionally acknowledging the many ways that racialized educational and employment discrimination disproportionately concentrate Black and Brown people into the most hazardous and lesser paid healthcare professions like phlebotomists or nursing assistants, thereby amplifying the disproportionately high deaths of Black and Brown communities to COVID-19 [8].

In the disparity examples above and throughout academic medical writing, health disparities reporting without purposeful inclusion of the context of racism risk having readers misunderstand the etiology of these inequities in ways that perpetuate the four mechanisms of colorblind racism [8]. Those include (1) naturalizing the existence of these outcomes, (2) abstract liberalism depicting these outcomes as the result of freely chosen actions that brought about poor health outcomes amidst a landscape of equal opportunity, (3) minimizing the ongoing relevance of racism in creating and sustaining these outcomes, and/or (4) attributing poor outcomes to unhealthy cultural values amongst racially minoritized communities. When providers are left to create their own understanding of racialized inequity after the presentation of driverless disparities, their theories and rationalizations often fall along the lines of aversive and colorblind racism.

For example, let us imagine a provider with no personal or academic affinity to understanding the trauma of racism to racially minoritized communities. Let us next imagine a scenario where they read the same APA disparity example before:

Compared with the general population, African Americans are less likely to be offered evidence-based medication therapy.

Because the driver of inequity is omitted from this example, the door is left open for this provider to create their own understanding. For instance, the provider might treat the presence of this disparity as if it is an unavoidable fact or “just the way things are.” The example disparity might be viewed with a metaphorical “shoulder shrug” to symbolize the provider’s helplessness or innocence [18]. This provider may think, “That’s terrible! I’m glad I’ve got nothing to do with it. That problem is much bigger than myself.” If the provider is unfamiliar that race is a social and political concept, they may even attribute this disparity to differential medication needs based upon the myth of biological race [6]. In these instances, the subtype of colorblind racism — naturalization — is being committed because the disparity is viewed as unavoidable and possibly biological as opposed to reversible with personal accountability and sociopolitical intervention.

Writing Styles Promoting Racial Health Equity

Societal and healthcare racism are known agents of health inequity. Medical research best upholds health equity and avoids faults of colorblind and aversive racism by directing attention to racism as the explicit agent driving observed disparities to racially minoritized patients as semantic patients. Consider the revised examples below.

Active Voice with Context:

Original APA Disparity Example: Compared with the general population, African Americans are less likely to be offered evidence-based medication therapy.

Corrected Example: Compared with privileged populations less harmed by racism, providers’ racial bias decreases their adherence to evidence-based medication therapy to African Americans,.

This sentence has been reformatted to an active voice with the agent in the spotlight. Here, the agent and subject of the sentence are “providers’ racial bias.” The agent is performing an action, “decreases.” And so, readers are able to clearly identify a starting point to conceptualize, explain, and correct this inequity: generate ideas of how to correct and hold accountable our own racial bias in prescribing practices. Furthermore, factors external to the bodies and/or choices of the racially minoritized group, being African Americans, are explicitly described to be responsible for this disparity. Therefore, the chances that readers without a robust understanding of racism and social medicine would generate colorblind or aversively racist rationalizations that blame the racially minoritized group for their own minoritization are intentionally reduced.

Original APA Disparity Example: CDC estimates that African Americans represented more than one-third (40%) of all people living with HIV in 2015.

Corrected Example: Intersecting mechanisms of systemic racism, including poverty and insurance access barriers, render African Americans as representing more than one-third (40%) of all people living with HIV in 2015.

The above sentences have been reformatted to use the active voice to clearly show how systemic racism creates the context in which this unfortunate statistic is rooted. Furthermore, such a construct proactively minimizes the burden placed upon readers without a robust understanding of racism, social medicine, or HIV awareness to generate their own flawed assumptions to explain this outcome [17].

We, the authors, honor the decades of racialized health disparity research as necessary to document our baseline starting points. According to Virchow’s principles of social medicine, however, racialized health inequities that are portrayed in an agentless, colorblind, and aversively racist fashion are incapable of conveying scientific or medical knowledge that directs meaningful health equity solutions that dismantle socially mediated health disparities. It is now time to transition academic medical writing towards a style of racial health equity. Accountability for iatrogenic racism in writing is uncomfortable, but necessary. We, too, are humbly developing these skills. Medical journals and national manuals of writing style could further each field’s commitments to racial equity by requiring authors to stylistically use the active voice to explicitly name racism as the agent inflicting harm upon racially minoritized patients. Words matter.