The Weight of Our Words: How Medical Communication Perpetuates Bias

Effective communication on family-centered rounds, during team-based care, and in medical documentation is fundamental to safe medical practice. Ineffective communication, therefore, can have a serious, mortal consequence for patients. In the landmark 1999 report To Err Is Human, the Institute of Medicine estimated that 98 000 lives are lost annually in the United States because of preventable medical errors, with nearly 70% of errors originating from miscommunication.¹  What followed was a concerted effort to systematize how medical professionals are trained to transmit information, leading to tools we use in daily practice: one-liners presented during rounds, I-PASS or SBAR handoffs during transitions of care, and huddles before procedures. When used well, these tools produce a shared mental model for team members, a path to safe and seamless care.

Take, for example, the following one-liner received by our inpatient pediatrics team during an admission handoff: “X is a stable 2-week-old born at 33 weeks’ gestation, presenting with umbilical erythema and swelling.” With these few words, our team was able to orient our clinical focus toward a diagnosis of omphalitis, plan out a course of intravenous antibiotics, and draft contingencies for necrotizing fasciitis and sepsis. This is an example of communication gone right, when the transmission of information leads to a shared mental model from which teams can make management decisions. Throughout this case, however, not all of our communication would be as effective.

A few days into the admission, team members noticed that the patient’s parents were at the bedside less frequently. On rounds, we discussed the importance of parental involvement in care. Our verbal sign-outs noted, “Parents tend to be absent.” We ensured that this information was passed on to the next set of providers through a carefully constructed written handoff. A plan to involve the Department of Children and Families was communicated to our multidisciplinary team. “After all,” we thought, “how could we give our patient the care they deserved if their parents were not invested?”

Days later, we learned that this family was living in a temporary shelter and that the many nights they spent at bedside after their child’s birth had put them at risk of losing their housing. Their “absences” were spent attending mandatory meetings to retain their shelter placement.

Our biased assumptions had led us to pass judgment on this family in a way that not only influenced our individual communication with them but also would affect their interactions with every person reading their health record moving forward. By incorporating our misinterpretation into formal written and spoken communication, we had disseminated and institutionalized our bias.

This type of miscommunication is abundant in the medical setting. A 15-year-old patient who presents after a sexual assault and declines a rape kit is referred to as “having risky behaviors” and “still making poor choices.” A Spanish-speaking, Latinx mother raising concerns about her daughter’s tubo-ovarian abscesses is signed out as “hysterical” and “needy.” A 14-year-old Black patient with attention-deficit/hyperactivity disorder and newly diagnosed diabetes expresses frustration that he is restricted to his hospital room and is subsequently labeled “aggressive.”

Research indicates that biased communication has major impacts on patient care. Stigmatizing language in the medical record is associated with less effective pain management and more negative attitudes toward patients compared with neutral descriptions of patients.²  This phenomenon has been extensively studied in sickle cell disease and substance use disorder, where the use of labels like “sickler” or “substance abuser” is associated with more negative attitudes toward patients.³  In associated studies, when clinicians were exposed to stigmatizing patient descriptions, they were more likely to condone punitive interventions, such as incarceration and disciplinary action for substance use, rather than treatment interventions.⁴ 

Language,⁵  therefore, is the vehicle through which we reinforce biased perceptions of patients, and when these are embedded in medical documentation, they carry institutional weight that lasts well beyond one encounter and care team. Besides having an impact on individual patients, this process has the insidious effect of reinforcing the challenging relationship between the medical establishment and marginalized communities, which are most likely to be affected by bias in the first place, making it more likely that these patients are dehumanized and their needs more easily dismissed.

So when our 15-year-old patient who presents after sexual assault reports severe nausea, her concerns are taken less seriously than those of patients we deem more responsible. When the Spanish-speaking, “hysterical” parent of our adolescent patient is present, we automatically scale back time at the bedside, particularly if an interpreter is required for updates. When our patient with diabetes and attention-deficit/hyperactivity disorder repeatedly ventures out of his room despite redirection, he risks restraint and is considered too hostile to be trusted with insulin injections. We tend to attribute symptoms of patients described as obese to their weight, sometimes overlooking other etiologies. We do more cursory examinations on patients with developmental delays, more easily accepting abnormalities as baseline and potentially missing meaningful clinical change.

How does medical practice, intended to serve our patients’ best interests, commit such disservice? The answer lies in acknowledging that our perceptions of patients are influenced by existing power structures and social norms. These biased perceptions affect care and are cyclically reinforced through communication, compromising the safety that tools like SBAR and I-PASS handoffs were crafted to optimize. To bend the arc of patient care toward greater equity, we as individuals must acknowledge the weight of our words. We must recognize that:

1. Our communication matters. The way we frame patients in summaries, handoffs, notes, and conversation with colleagues is not benign; it can reinforce existing institutional biases and health inequities. Now that families can access health records under the 21st Century Cures Act, biased assessments in patient documentation can reinforce the cycles of trauma and distrust that communities experience in relation to the medical establishment.

2. Existing power structures affect our communication. Our individual biases are learned from our social environment. Existing power structures create bias along the lines of race, gender, income, and other experiences, but because our biases are learned, they are predictable and can be unlearned.

3. We have the power to reframe our judgments. When we reframe problems as driven by systems of structural inequality and racism, rather than inherent to individual patients, we can better recognize their root causes and demand upstream systemic change to provide what our patients truly require.

Although individual providers making more deliberate language choices is essential to the safety of marginalized patients, changing behavior is ultimately in service of broader goals: challenging biased belief systems and pursuing systemic change. In this vein, medical institutions must acknowledge that they are not blameless. Historically, institutions have exploited marginalized communities, perpetuating distrust and reinforcing structural inequity. Thus, institutions have a responsibility to change the culture around biased language and restructure care practices to be more just and responsive to community need. We call on our institutions to invest time and capital into research that highlights the true costs of biased communication and in programming that mitigates the impact of biased language and dismantles the power structures that this language represents.

The mortal consequence of ineffective medical communication has been studied extensively. It is time to weigh the consequences of biased communication—and the structural inequity it reinforces—to the same degree.