As many doctors are starting to realize, they are not passive agents in a racist society and health care system: their actions (and inactions) contribute to racial health disparities. And their biases show up on medical charts.

Doctors are more likely to describe Black patients as “angry,” “aggressive” or “noncompliant,” or question their symptoms. University of Chicago investigators analyzed 40,000 patient records and found that Black patients were 2.5 times more likely to have negative descriptors included in their records, compared to white patients.  

While much work has focused on racial biases in patient care and medical algorithms and instruments, these findings underscore the need to change medical documentation. Health care professionals spend 49 percent of their office time charting. Given growing documentation burnout, the pursuit of efficiency may contribute to racial bias: time crunch exacerbates cognitive shortcuts and the cursory reliance on (and copying forward of) previous notes.

My colleagues and I recently published a paper in The New England Journal of Medicine outlining seven charting habits of doctors that maintain and exacerbate racism. Here are the tools to counteract them:

Ask why a patient’s race is in their chart: Clinicians may document race uncritically, as a matter of routine practice. We should ask why. Mentioning race at the beginning of a clinical note puts it ahead of more critical case notes. For instance, Black patients are 2.4 times more likely to be incorrectly diagnosed with a psychotic disorder and, if they do have a psychotic disorder, are more likely to receive less effective (first-generation) antipsychotics. When clinicians do come across the mention of race in notes, this should prompt self-reflection against ingrained racist beliefs or mental shortcuts that could erroneously link a patient and a specific medical condition or treatment.

Allow patients to self-identify: Clinicians typically attribute race based on a patient’s physical appearance or “social assignment.” This relies on clinicians’ cultural paradigms of how certain physical characteristics relate to specific racial groups (e.g., associating darker skin tones with Black individuals), associations often rooted in flawed and harmful assumptions. In contrast, racial self-identification based on a patient’s own social, cultural, familial and political affiliations, provides different information. A person can identify as Asian American, despite surface appearances for example, because of their family history or cultural affiliation. Allowing patients to name their racial background can shed light on other relevant parts of their identity, such as country of origin or immigration status.

Let patients describe their lives: Doctors may believe they are being virtuous in adopting a “colorblind” or “post-racial” ideology. Yet, race is an omnipresent facet of our society. A growing literature links exposure to discrimination to poor health. Race and culture are not synonymous, but racial identity ties to cultural experiences such as religion, food and health-related beliefs and behaviors. Doctors should use initial probing questions in assessing sensitive issues like racial trauma to gain consent and avoid potentially retraumatizing individuals. A structural vulnerability assessment, cultural formulation interview or UnRESTS (the University of Connecticut Racial/Ethnic Stress and Trauma Scale) can help guide conversations about a patient’s racial or cultural experiences and their perception of their health issues, help-seeking behaviors and relationship to the health care system.

Don’t assume other parts of a patient’s identity: In light of racial health disparities, some doctors may overcorrect, victimizing or homogenizing groups based on their race. Despite idiosyncratic social determinants of health, they may incorrectly use race as a proxy for housing status, geography, transportation or employment. Structural stereotyping not only harms patient rapport but also mismanages diagnoses and treatment plans. Instead, patients’ understanding of their health (and the health care system), and of upstream factors impacting their health (e.g., barriers to food, housing, transportation, insurance and social support), should be explicitly elicited and documented. Practitioners may assume patients’ mistrust and hostility towards health care based on their race. Clinicians also wrongly attribute health disparities to patients’ decisions or perceived cultural dysfunction, and documentation often reflects this bias. We should instead identify and document underlying racist policies and institutions that exacerbate health risks. Finally, people (and their health) are not defined by their vulnerabilities: it is vital to also document strengths and protective factors (e.g., familial relationships, community support, spiritual engagement, stable housing, etc.) within a patient’s social history.

Do not use race in the one-liner: Initial clinical notes trigger pattern recognition and clinical reasoning in physicians, so introducing race in the first line identifying a patient in the note (the one-liner) can lead both the writer and readers away from critical reflection on race (and racism) and towards cognitive shortcuts, implicit biases and explicit stereotypes. Doctors are more likely to incorrectly associate Black patients with diagnoses of HIV, hypertension, obesity, sarcoidosis, sickle cell anemia and stroke. When race headlines a patient’s chart, it's presented without necessary context. Additionally, placing race beside past medical history and chief complaint can incorrectly imply that race, a social construct that has been used to enslave and persecute people, has some biological importance. Race is a scientific myth that is not a reliable indicator of genetics and biological risk factors. A patient’s self-identified race and related life experiences can offer valuable clinical insights when thoughtfully documented in the social history section of a medical record, allowing doctors to aptly address patients’ concerns and refine treatment plans. Finally, a more detailed understanding of patients’ racial and ethnic experiences can inform advocacy efforts to combat the causes of healthcare disparities.

Foster patient choices: If racial identifiers are noted during registration, patients should have choices (including the option to opt out) and be told why such information is collected. Many have faced doctors skeptical of their symptoms or accounts and, later when charting, clinicians may downplay or question their viewpoint. While using direct quotations from patients may appear to center their voice, research indicates that quotations and other charting conventions can convey stigma and doubt instead. The three recurring linguistic features used by clinicians in patient records include quotation marks (e.g., “the patient had a ‘reaction’ to the medication”), judgment words implying doubt (e.g., the patient “claims” or “insists”) and “evidentials”—sentence structures where symptoms are presented as hearsay. Crucially, doctors do not employ these habits equally across patients: they most frequently use these undermining techniques when documenting encounters with Black or female patients. 

Deploy debiasing strategies: Clinicians should acknowledge that clinical assessments are cross-sectional snapshots perhaps influenced by mental shortcuts or emotions. These biases, whether implicit or explicit, can be mitigated by actively using debiasing strategies, such as self-reflection, perspective-taking and intentional pauses for thoughtful contemplation.

We must be thoughtful. Doctors are still predominantly white and, given the history of race-based medical atrocities in the U.S., patients may feel uncomfortable discussing aspects of their identity or feel it is irrelevant (or potentially harmful) to their care. As a white, female doctor, I can understand how race may be the last thing a patient wants to discuss with me. Therefore, in addition to using clinical judgment to ascertain the relevance of documenting identity information, doctors must always put the patient’s perspective first in their efforts to decrease bias and more accurately capture health information. 

With the advent of direct patient access to the electronic medical record (EMR) and clinical notes, we hope clinicians and patients can use these tools to advance antiracism in medicine. When viewing their EMR, patients should feel empowered to understand these tools and evaluate how their doctors approach notes. Is there a level of thoughtfulness regarding race and ethnicity, and other aspects of identity, that acknowledges the power and influence held by such documentation? 

Many rallied behind the Black Lives Matter movement or spoke out against racial health disparities during the pandemic, but embodying the antiracist society for which we advocated requires changing medicine, including clinical documentation. We hope that, with this set of tools, antiracism can be a choice.

This is an opinion and analysis article, and the views expressed by the author or authors are not necessarily those of Scientific American.