Implicit Bias: Impact on Maternal Health in Minority Populations

Disclosures: 

Presenter Disclosure:

• Financial: Calista Kelly is employed by Continued.com.
• Non-Financial: Calista Kelly has no non-financial disclosures.

Content Disclosure: This learning event does not focus exclusively on any specific product or service.

Sponsor Disclosure: There is no external sponsor for this course.

Limitations/Risks

• Awareness is not impact. Reading about bias can raise knowledge and self-awareness, but it does not reliably change clinical behavior or patient outcomes on its own; completion should not be mistaken for improvement in practice.
• DEI and cultural competency. Cultural-competency framing can drift toward checklist "mastery" of groups and reinforce the very stereotyping it aims to reduce, so it must be paired with cultural humility; learners should also remember that implicit bias, structural racism, cultural competency, and DEI are often separate licensure requirements, and that this is a contested space where defensiveness can undermine engagement.
• Format and assessment. As a self-paced written activity assessed by a multiple-choice test, the course measures knowledge rather than skill or team behavior, and it can only model interprofessional collaboration rather than have learners practice it.
• Data currency and fit. The statistics are time-sensitive and will need updating. Reliable data for some high-risk populations are limited or suppressed, and national figures may not reflect a learner's local patients or setting.

Learning Outcomes

After this course, participants will be able to:

• Describe implicit and explicit bias, explain how each develops through lifespan socialization and can influence interprofessional team members’ interactions with childbearing patients, and identify resources team members can use to examine their own biases.
• Recognize how bias and power differentials between clinicians and patients can affect the patient and the patient and team relationship.
• Describe structural racism and how it forms, explain its impact on maternal health, and identify risk factors for and patterns of maternal morbidity and mortality across White, Black, Hispanic, Asian, and American Indian and Alaska Native populations in the United States.
• Describe infant mortality in the United States and its ties to maternal health, and explain the interprofessional team’s role in reducing the impact of implicit bias and structural racism for pregnant patients.

Introduction: Why This Topic Belongs to the Whole Team

Pregnancy and the year that follows it are among the most medically and socially intensive periods in a person’s life, and they are also a period in which the United States performs poorly relative to its wealth. Maternal deaths in this country have remained stubbornly high, and they fall unequally. In 2023, the national maternal mortality rate was 18.6 deaths per 100,000 live births, but the rate for Black women was 50.3 per 100,000, more than three times the rate for White women (Hoyert, 2025). These differences are not explained away by income or education, nor are they a recent accident. They are the visible surface of two deeper and interacting forces: the everyday operation of bias inside clinical encounters, and the long history of structural racism that shapes who gets sick, who is believed, and who survives.

This course is written for an interprofessional audience. The perinatal team is larger than many clinicians assume. It includes obstetric and nursing staff, but it also includes behavioral health clinicians who treat perinatal mood and anxiety disorders, physical therapists who manage pelvic floor dysfunction and musculoskeletal pain in pregnancy and postpartum, occupational therapists who support feeding, infant development, and a parent’s return to daily activities, speech-language pathologists and audiologists who screen and treat newborns and follow infants at developmental risk and respiratory therapists who manage the breathing support that premature and medically fragile newborns depend on and who help care for mothers facing respiratory complications of pregnancy. Each of these professionals interacts with childbearing patients and their families, often at moments of vulnerability, and each carries the same human cognitive machinery that produces bias. A course built for only one discipline would miss the central point of this material: that bias and its consequences are a team problem and a team responsibility.

The interprofessional framing also matters because the evidence base points toward team-level change rather than individual heroics. Education planned by and for the healthcare team, in the language of Joint Accreditation, is designed to change how team members work together, not simply to update one clinician’s knowledge. The four competency domains commonly used to organize interprofessional practice, namely values and ethics, roles and responsibilities, interprofessional communication, and teams and teamwork, will recur throughout this material because reducing the harm of bias depends on all four. A physical therapist who notices that a patient’s pain reports are being dismissed, an occupational therapist who hears a colleague describe a family in stereotyped terms, or a behavioral health clinician who recognizes that a patient has stopped trusting the care team, each holds a piece of the solution, but only if the team has the shared values, role clarity, and communication habits to act on it.

This is sensitive material. Discussions of racism, bias, and preventable death can be uncomfortable, and discomfort is an expected part of learning in this area (Campbell, 2025). The goal here is not to assign guilt but to build accurate understanding and practical skill. Bias is a feature of normal human cognition, present in well-intentioned people, and the question is not whether a clinician has biases but what the team does about them. With that framing in mind, the course moves from the psychology of bias to how bias forms and operates in clinical relationships, to the structural forces surrounding those relationships, to the maternal and infant outcomes at stake, and finally to what the interprofessional team can actually do.

A word about how to engage with this material will help. The most useful posture for a learner here is neither defensiveness nor self-flagellation, but curiosity. The aim is to leave with a more accurate model of why a group of well-meaning professionals can nonetheless participate in a pattern of unequal outcomes, and with a concrete sense of which of one’s own habits and which of the team’s routines are worth changing. Where the course presents numbers, it presents the most recent national figures available and names their source, so that a reader can see that the disparities described are not rhetorical but measured. Where it makes a claim about how bias operates, it ties that claim to the research literature rather than to anecdote. And throughout, it keeps returning to the same practical question that interprofessional education is designed to answer: not what should I, alone, believe, but how should we, together, work, so that the next patient is heard, believed, and kept safe regardless of who she is.

Section 1: Implicit and Explicit Bias Defined and Differentiated

A bias, in the most general sense, is a tendency to favor or disfavor a person, group, or idea in a way that interferes with fair judgment. In healthcare, the term usually refers to attitudes about social groups that shape how a clinician perceives, communicates with, and treats a patient. Two broad categories are useful to distinguish at the outset, because they form and operate differently and call for different responses.

Explicit bias refers to attitudes and beliefs that a person holds consciously and can report. A clinician who openly believes that members of a particular group are less reliable historians of their own symptoms holds an explicit bias. Explicit biases are, by definition, available to introspection, which means they can be stated, defended, argued with, and, in many settings, socially discouraged. Most clinicians sincerely endorse the value of treating every patient equally, and on explicit measures, healthcare professionals tend to report egalitarian attitudes (FitzGerald & Hurst, 2017). That sincerity is real and part of what makes the second category so important.

Implicit bias refers to attitudes and associations that operate outside of conscious awareness and intention. These are automatic mental shortcuts: learned associations between a social group and a set of traits, evaluations, or expectations that activate quickly and without deliberate thought. Because they are unconscious, implicit biases can run contrary to a person’s consciously held, explicitly egalitarian values, and a person can be entirely unaware that the bias is influencing a decision (Campbell, 2025). This is the crucial difference. A clinician can honestly believe that they treat all patients the same and still, in the rushed and cognitively demanding environment of clinical care, act on associations they would reject if asked about them directly.

The mechanism is rooted in ordinary brain function. Implicit associations are formed through repeated exposure to messages and patterns in the surrounding culture, and they can be rapidly activated when the brain encounters socially relevant cues, producing a fast, automatic evaluation before slower, deliberate reasoning has a chance to engage (Campbell, 2025; Josiah et al., 2023). This speed is adaptive in many contexts because the brain must constantly categorize and predict, but it becomes a liability when learned associations are inaccurate and attach negative expectations to a patient based on race, ethnicity, language, body size, age, disability, or other group membership.

A widely used research tool for detecting implicit associations is the Implicit Association Test, which measures how quickly people pair concepts, on the premise that more strongly associated concepts are paired more quickly. Research using such measures has repeatedly found that healthcare professionals, as a group, show levels of implicit bias similar to those of the general population, and a systematic review of this literature concluded that such bias among healthcare professionals is associated with measurable differences in the care patients receive (FitzGerald & Hurst, 2017; Hall et al., 2015). In other words, entering a caring profession does not exempt a person from the associations absorbed from the broader culture. It is worth noting that individual scores on these tests are not a diagnosis of a clinician’s character and do not perfectly predict any single decision; the value of the research lies in establishing, at the population level, that the phenomenon is real and pervasive.

Why does the distinction between explicit and implicit bias matter for the interprofessional team? Because the two require different countermeasures. Explicit bias can sometimes be addressed through policy, accountability, and direct feedback, since the attitude is available to the person holding it. Implicit bias is not reachable in the same way, because the person is not aware of it. Telling a team to simply try harder to be fair is therefore an incomplete strategy, since the biases most likely to slip through are the ones no one intends. Effective approaches, discussed in Section 7, work by changing the conditions under which automatic associations get a chance to drive behavior, for example, by slowing decisions at high-stakes moments, by standardizing how symptoms are assessed, and by building team habits that catch what an individual misses.

It also helps to see how an implicit association becomes a clinical decision, because the pathway is rarely dramatic. Clinical reasoning depends heavily on rapid pattern recognition, the experienced clinician’s ability to size up a situation quickly, and that same speed is what allows an unexamined association to enter the judgment unnoticed. The bias does not announce itself as prejudice. It shows up as a slightly lower index of suspicion for one patient than another, a quicker readiness to attribute a complaint to anxiety or drug-seeking, a marginally shorter visit, less eye contact, fewer open-ended questions, or a faster move to reassurance rather than investigation. Any one of these is small, and any one could have an innocent explanation, which is exactly why bias is so difficult to catch in the moment and so consequential in the aggregate. Across thousands of encounters, small differences in suspicion, attention, and effort accumulate into measurable differences in the care that groups of patients receive (Campbell, 2025; FitzGerald & Hurst, 2017).

A further nuance is that implicit and explicit attitudes can point in different directions within the same person, and the relationship between what a clinician sincerely believes and how a clinician automatically reacts is loose rather than fixed. A clinician may hold strong, genuine egalitarian commitments and still carry implicit associations absorbed from the surrounding culture, and under ordinary conditions of fatigue, time pressure, and cognitive load, it is the automatic system that often wins (Campbell, 2025). This is why the sincerity of a team’s stated values, though real and important, cannot be the whole of its safeguard. Values shape the explicit layer; they do not automatically reach the implicit one. Recognizing this gap is not cynicism about clinicians’ character. It is an accurate account of how the mind works, and it points toward solutions that do not depend on individuals simply wanting to be fair.

It is also worth being precise about what implicit-bias measures can and cannot tell us, because misunderstanding this point has fueled both overclaiming and backlash. Tools such as the Implicit Association Test are research instruments that reliably detect automatic associations at the group level, and the body of evidence using them demonstrates that clinicians, like everyone else, hold such associations (FitzGerald & Hurst, 2017). What these tools do not do well is predict, with precision, what a particular clinician will do with a particular patient on a particular day. A clinician should therefore neither treat a personal test result as proof of guilt nor dismiss the entire phenomenon because the prediction at the individual level is imperfect. The defensible reading of the science is the moderate one: implicit bias is real, common among well-intentioned clinicians, and capable of influencing care, while remaining only one of several forces acting in any given encounter (Campbell, 2025; FitzGerald & Hurst, 2017).

For every profession on the perinatal team, the practical translation is the same. The audiologist interpreting a parent’s response to a newborn hearing screen, the speech-language pathologist judging whether a feeding concern warrants follow-up, the physical therapist deciding how seriously to take a postpartum patient’s report of pain, the occupational therapist assessing a parent’s engagement with an infant, and the behavioral health clinician forming an impression of a patient’s risk, are all making rapid judgments under time pressure and uncertainty, which are exactly the conditions in which implicit associations exert the most influence. None of these clinicians is choosing to be unfair, and most would be distressed to learn that an outcome had turned on an association they never endorsed. That is precisely the point. Recognizing that these everyday, well-meant judgments are vulnerable is the first step toward protecting them, and protecting them is a task that the following sections show is best carried out by the team rather than left to the individual.

It is worth naming explicitly the conditions under which an automatic association is most likely to slip past a clinician’s better judgment, because those conditions describe ordinary clinical life rather than rare exceptions. Bias tends to exert the greatest influence when a decision is made quickly, when the clinician is fatigued or stressed, when cognitive load is high, when the information available is ambiguous, and when the clinician must rely on first impressions rather than accumulated knowledge of the individual patient (Campbell, 2025). Each of these is a routine feature of perinatal care: the triage decision made in a busy unit, the late-shift assessment by a tired clinician, the brief first encounter with a patient whose history is incomplete. This matters for the team because it means that the moments of greatest vulnerability to bias are predictable, and predictable vulnerabilities can be designed around. If the team knows that the rushed, ambiguous, first-impression decision is where bias is most likely to enter, it can build in the deliberate pause, the second look, and the standard protocol precisely at those points, rather than relying on individuals to summon extra fairness at the very moments when their capacity to do so is most depleted (Campbell, 2025; Hagiwara et al., 2024; Kirwan Institute, 2013).

Bias also comes in more than one form, and a shared vocabulary helps a team recognize the particular pattern at work in a given encounter. Beyond the racial and gender bias most often discussed, the cognitive-science and healthcare literatures describe a family of related tendencies: affinity bias, the pull toward people who resemble us; confirmation bias, the tendency to notice and weight information that fits an expectation already formed; anchoring, the overreliance on the first piece of information encountered; attribution bias, the tendency to explain another person’s behavior in terms of character rather than circumstance; the halo and horns effects, in which one salient trait colors the whole impression; and name bias and bias based on perceived weight, body type, age, education, or socioeconomic status (Kirwan Institute, 2013). These are not separate phenomena from implicit bias but specific shapes it takes, and in the perinatal setting, each has a concrete cost. Anchoring on an initial impression of a patient as anxious can crowd out later physical signs; confirmation bias can lead a clinician to register findings that fit a benign explanation and discount those that do not; and weight or age bias can quietly lower the seriousness with which a symptom is treated. Naming the specific pattern is useful precisely because the countermeasure differs: confirmation bias calls for deliberately seeking disconfirming evidence, while anchoring calls for revisiting the initial assessment as new information arrives. 

Section 2: How Bias Forms: Socialization, Stereotypes, Prejudice, and "Isms" Across the Lifespan

If implicit biases are learned associations, the natural question is where they are learned. The short answer is that they are absorbed over a lifetime from the social world a person grows up in and lives in. No one is born with a fully formed set of biases, and no one chooses them deliberately. They accumulate through the ordinary process of socialization, the lifelong process by which a person takes in the norms, categories, images, and expectations of family, community, schooling, media, language, and institutions (Campbell, 2025). Understanding this process helps the team see bias as a product of the environment rather than a personal moral failing, which, paradoxically, makes it easier to confront it honestly.

Several related terms are useful to define here, because they are often used loosely. A stereotype is a generalized belief about the characteristics of a group, applied to individuals who belong to it. Stereotypes are cognitive: they are the brain’s categories and the traits it attaches to them. Prejudice is an attitude or feeling, usually negative, directed toward a group or its members. Where a stereotype is a belief, prejudice is the affective charge, the liking or disliking, that often accompanies it. Discrimination is behavior: the differential treatment of people based on their group membership. The suffix “ism,” as in racism, sexism, ageism, ableism, or classism, names a system in which stereotypes and prejudice become organized and reinforced at the level of institutions and culture, so that the disadvantage no longer depends on any single biased individual. These distinctions matter because an intervention aimed at a belief, a feeling, a behavior, or a system will look different in each case.

The lifespan dimension is important. Children absorb associations from the earliest years, often before they have the language to question them, simply by noticing patterns in who is present, who is praised, who is feared, and who is absent in their environment. These associations are then reinforced across decades by repeated exposure to the same patterns in entertainment, news, professional training, and the structure of the institutions a person passes through (Campbell, 2025). For healthcare professionals, there is a specific occupational layer to this story. Clinical training and longstanding medical literature have at times transmitted inaccurate beliefs about biological differences between racial groups, and such beliefs, once learned, can persist and distort clinical judgment, for example, in the historically documented tendency to underestimate the pain of Black patients (Josiah et al., 2023). A clinician, therefore, arrives at the bedside carrying both the general biases of the surrounding culture and any specific misconceptions transmitted through professional socialization.

It is worth dwelling on each of the key terms, because precise language helps a team diagnose what it is actually facing. A stereotype can be positive in valence and still cause harm; the belief that a particular group is unusually resilient or tolerant of pain is a stereotype, and in maternity care, a stereotype of toughness can translate directly into undertreatment. Prejudice need not be intense hostility; a faint, automatic discomfort or a reflexive lowering of warmth toward members of a group is prejudice in the sense that matters clinically, and it shapes the emotional tone of an encounter that the patient feels even when nothing is said. Discrimination, the behavioral layer, includes not only overt acts but also the quiet differentials described in Section 1, the shorter visit and the lower suspicion. And the “ism” layer is what makes the problem self-sustaining: once stereotypes and prejudice are built into curricula, guidelines, institutional routines, and the surrounding culture, they reproduce themselves in each new cohort of professionals without anyone deciding to pass them on (Campbell, 2025; Njoku et al., 2023). A team that can identify which layer it is dealing with is better positioned to choose an appropriate response.

The developmental account also explains why exhortation alone fails. By the time a person reaches professional training, decades of socialization have already laid down the associations, and a sincere adult decision to be fair operates on top of that foundation rather than erasing it (Campbell, 2025). This is not a counsel of despair. Associations that were learned can be weakened and counteracted, and the surrounding environment that produced them can be changed. But it does mean that the goal of self-examination is not to reach a state of being bias-free, which the evidence does not support as a realistic endpoint, but to become reliably aware of one’s vulnerabilities and to build the personal and team habits that keep those vulnerabilities from reaching patients (Campbell, 2025; Hagiwara et al., 2024).

The occupational layer deserves particular emphasis for a clinical audience, because it is the part of socialization that professional training is directly responsible for. Historically, medical and clinical education transmitted false beliefs about biological differences between racial groups, including the durable and damaging myth that Black patients experience or tolerate pain differently, and such beliefs, absorbed during training, can persist into practice and distort assessment and treatment long after they have been formally discredited (Josiah et al., 2023). The maternal health literature treats this history as directly relevant to present-day disparities, because pain and symptom assessment are exactly the judgments through which warning signs of serious complications are recognized or missed (Josiah et al., 2023; Montalmant & Ettinger, 2024). A team that understands this history is less likely to be surprised when a patient’s reported pain is discounted and more likely to build checks to catch such discounting.

Turning to resources, several categories support honest self-examination, and they work best in combination. Validated self-assessment and program-level instruments allow individuals and teams to measure awareness of and mitigation of bias in maternity care and to track change over time, replacing impressions with data and enabling assessment of whether anything is actually improving (Bower et al., 2023). Structured reflective practice, in which clinicians examine specific encounters and their own emotional reactions to particular patients, helps surface associations that stay hidden under the smooth surface of routine. Continuing education, including activities like this one, provides the shared vocabulary and evidence that allow a team to discuss bias without each conversation collapsing into defensiveness or blame. Confidential implicit-association measures can be a useful prompt for personal reflection, provided they are framed as conversation-starters rather than verdicts. And cultural humility, the disposition to approach each patient as the authority on their own life and to remain a learner rather than assume expertise about a patient’s experience, gives self-examination a durable orientation rather than a finish line.

Putting these resources into practice is itself a skill, and a few principles make the effort more honest and more durable. Reflection is most useful when it is specific rather than general: examining an actual encounter that left a clinician uneasy, or a particular patient toward whom the clinician felt an unexplained reluctance, surfaces more than an abstract intention to be fair. It is also most useful when it is routine rather than occasional, because the associations being examined are themselves the product of constant, lifelong reinforcement and will not be undone by a single sitting (Campbell, 2025). Carrying the work out in the company of colleagues, in a setting where it is safe to name discomfort without fear of judgment, tends to be more sustainable than solitary effort, both because others can see patterns a clinician cannot see in themselves and because shared reflection builds the team culture in which raising the subject of bias is normal rather than threatening. The aim of all of this is not the impossible one of becoming a person with no associations, but the achievable one of becoming a clinician who knows where their judgment is most likely to slip and who has built, with their team, the habits that catch it before it reaches a patient (Campbell, 2025; Hagiwara et al., 2024).

A word of realism is appropriate here, because it will recur in Section 7. The mere act of measuring or naming one’s biases does not, by itself, reliably change behavior, and some popular training approaches rest on weak evidence and have not demonstrated lasting effects on practice (Hagiwara et al., 2024). Self-examination is necessary but not sufficient. Its value lies in opening the door to the structural and team-level changes that do the heavy lifting, and in sustaining the personal vigilance those changes require. For the interprofessional team, the shared goal of this section is a common understanding: that every member arrived with biases they did not choose, that some of those biases were transmitted by the professions’ own training, that these biases are knowable through honest and repeated reflection supported by validated tools, and that naming them is the beginning of the work rather than the end of it.

Section 3: Bias, Power Differentials, and the Patient

The amplification works in several ways. First, because the clinician controls the interpretation of the encounter, a biased expectation can quietly become the official account. If an implicit association leads a clinician to expect that a particular patient is exaggerating symptoms, the clinician may document the visit in those terms, and that documentation then follows the patient through the system, shaping how the next clinician approaches them. Second, the power differential makes it costly for patients to push back. A patient who senses that they are not being believed may hesitate to insist, precisely because they depend on the clinician and fear being labeled difficult, which can further reduce the information the clinician receives and confirm the original misjudgment. Third, the differential is steeper for patients who already occupy a marginalized social position, so that a Black patient, a patient with limited English proficiency, a patient with a disability, or a patient with a low income may face a compounded gap in power that makes the effects of bias harder to resist (Njoku et al., 2023).

It is useful to trace how a small interpersonal moment can cascade into a clinical outcome, because the chain is what makes the power differential a safety issue rather than merely a matter of courtesy. Consider a patient whose report of severe pain or unusual symptoms is met with a subtly skeptical response shaped by an implicit association. The immediate effect is that the symptom is investigated less thoroughly. The patient, sensing skepticism and aware of the cost of being labeled difficult, may soften or withdraw the complaint, thereby removing the very information that might have prompted action. The encounter is then documented in language that reflects the original skepticism, and that documentation primes the next clinician to view the patient the same way. The patient, having learned that raising concerns is unproductive, discloses less at the next visit, attends less reliably, and waits longer before seeking help when something is genuinely wrong. In the perinatal period, where the difference between a good outcome and a catastrophe can hinge on whether a warning sign is acted on early, this cascade is not abstract. It is a plausible route from an unexamined association to a preventable harm, and it runs entirely through the power differential (Campbell, 2025; Josiah et al., 2023).

The maternal health literature describes precisely these effects. Implicit bias among clinicians has been linked to differences in communication, in the patient’s experience of care, in treatment decisions, and ultimately in outcomes, and perceived discrimination is repeatedly associated with erosion of the trust that childbearing patients place in their care teams (Josiah et al., 2023; Montalmant & Ettinger, 2024). Studies that ask patients directly find the same pattern from the other side of the encounter: Black and Latino patients have reported lower ratings of provider communication, trust in the provider, patient-centeredness, collaboration in decision-making, and the clinician’s contextual knowledge of them as a person (Blair et al., 2013). Trust is not a sentimental concern in this setting; it is functionally load-bearing. A patient who trusts the team tells the team things, returns to the team, and follows through with the team. A patient whose trust has been damaged, often by an accumulation of small dismissals rather than a single dramatic insult, may do none of these, and the resulting gaps in information and engagement are themselves risks. Because trust is built and broken across many encounters with many professionals, it is inherently a team property, which is part of why the response has to be a team response.

A particularly common channel through which bias reaches patients is the microaggression, a term for the everyday slights, dismissals, and subtle verbal or nonverbal behaviors that communicate a hostile, demeaning, or diminishing message to members of marginalized groups, often without the speaker’s awareness or intent (Sue et al., 2007). If implicit biases are the unconscious associations, microaggressions are among the behaviors that flow from them. In a clinical setting, they take recognizable forms: a comment that presumes a patient’s family structure, a term of endearment used in place of a name in a way that infantilizes, an expression of surprise at a patient’s competence, or a documentation choice that subtly casts doubt on a patient’s account. Language is central here because subtle word choices can carry biased assumptions and because the effects of biased language extend beyond a single bruised feeling, shaping health outcomes and reinforcing the belief systems behind discriminatory practices (American Medical Association, 2021). Nonverbal behavior matters as well, since bias can register in reduced eye contact, greater physical distance, or a colder tone that the patient perceives even when nothing overt is said. For the interprofessional team, the practical point is that respectful, person-centered language and attentive nonverbal communication are not merely courtesies but are themselves part of safe care, because they protect the trust on which a patient’s willingness to disclose depends.

The differential is also not uniform across patients, and the concept of intersectionality helps explain why. A patient may occupy several marginalized positions at once, for example, as a Black woman with a low income and limited English proficiency, and the disadvantages do not simply add; they interact, producing a steeper and more complex power gap than any single characteristic would predict (Njoku et al., 2023). A clinician who is attentive to race but not to language, or to disability but not to class, may still leave large blind spots. The practical implication is not that clinicians must catalog every dimension of a patient’s identity, which is impossible in the moment, but that they should hold their confident first impressions loosely, especially for patients whose social position differs most from their own, and lean on the structures and the team that the rest of this course describes.

The interprofessional lens makes the power differential more visible because different team members sit at different points in the hierarchy and see different parts of the patient’s experience. A patient may feel unable to question a physician but may speak more freely with a physical therapist during a longer hands-on session, with an occupational therapist working in the patient’s own environment, or with a behavioral health clinician whose role is explicitly to listen. These professionals often function as the team’s early warning system, hearing concerns that the patient did not feel safe raising elsewhere. Audiologists and speech-language pathologists, who frequently have repeated contact with families during newborn screening and developmental follow-up, occupy a similar position of trust. The value of this distributed vantage point is lost, however, if the team lacks the communication channels and the shared sense of permission to escalate what these members hear. A concern that a patient confides to the one clinician she trusts is worth little if that clinician has no reliable way to convey it to the people who can act, or no confidence that doing so is welcome. Recognizing the power differential, therefore, points directly to a teamwork solution: structures that flatten the hierarchy enough for patient concerns to travel, regardless of which team member first receives them, and a culture in which raising another professional’s possible blind spot is treated as good care rather than as an affront.

It is worth dwelling on trust as a clinical resource because it is easy to file under bedside manner and thereby underestimate it. In the perinatal period, the patient is often the first, and sometimes the only, person who can detect that something is wrong in her own body, and the system’s ability to act on that early signal depends entirely on whether she reports it and is believed. Trust is what makes the reporting happen. A patient who trusts her care team will mention the headache, the swelling, the strange breathlessness, the dark mood, and will do so early, when intervention is easiest; a patient whose trust has been worn away by accumulated small dismissals may say nothing until a crisis forces the issue, by which point the window for easy intervention may have closed. The maternal health literature links perceived discrimination to this erosion of trust and treats the resulting disengagement as a contributor to worse outcomes rather than as a mere matter of patient satisfaction (Josiah et al., 2023; Montalmant & Ettinger, 2024). Because trust is built and spent across many encounters with many different professionals, no single clinician owns it, and no single clinician can repair it alone, which is one more reason that protecting it is a shared, interprofessional responsibility rather than an individual courtesy.

A final point concerns the patient’s own account of being treated unfairly. Patients are often accurate observers of how they are treated, and a patient’s report of feeling dismissed or disrespected is clinical information, not a complaint to be managed away (Campbell, 2025). Treating such reports as data and responding to them with curiosity rather than defensiveness is one of the most direct ways an individual clinician can reduce the harm that flows from the power differential. It is also a behavior that any member of the team can model, regardless of discipline or seniority, making it one of the few high-value responses that do not require a system change to begin.

Section 4: Structural Racism: What It Is, How It Forms, and Its Impact on Maternal Health 

How does a structure like this form? It accumulates historically. Practices established in one era, such as residential segregation, unequal access to education and employment, discriminatory lending, and concentrated policing and incarceration, become embedded in the geography and institutions of a society and continue to shape opportunity long after the original policies are formally ended (Sonderlund et al., 2022). One useful way to operationalize the concept for research is to trace specific structural domains and the pathways by which they affect health. For example, scholars have examined how the spatial and racialized clustering of incarceration, itself a product of structural racism, is associated with adverse birth outcomes in the surrounding communities, illustrating how a structure seemingly remote from the delivery room can reach into pregnancy and birth (Sonderlund et al., 2022). The general lesson is that structural racism operates through many intermediate channels, including where people can afford to live, the quality of the hospitals accessible to them, the chronic stress of navigating discrimination, and the resources their neighborhoods do or do not contain.

In maternal health specifically, several pathways are well described, and it helps to walk through them, because each suggests a different point of intervention. One is differential access to high-quality care. Because of residential and economic segregation, Black patients in the United States are disproportionately likely to give birth in a concentrated set of hospitals, and care quality across such facilities is, on average, lower on multiple measures, so that two patients with similar needs can receive systematically different care depending on where structural forces have routed them (Montalmant & Ettinger, 2024). This is a clear example of how a structure produces a disparity without any individual at the bedside intending it: the segregation that determines neighborhood, the economics that determine which hospital serves that neighborhood, and the resourcing that determines that hospital’s quality are all upstream of the delivery room, yet they shape what happens in it.

A second pathway is the cumulative physiological toll of living with chronic stress and discrimination over the life course, sometimes described through a weathering framework, in which the sustained wear of stress contributes to earlier onset and greater severity of the chronic conditions, such as hypertension and cardiovascular disease, that drive maternal risk (Njoku et al., 2023). The idea of weathering is important because it dissolves false comfort. It is tempting to assume that a patient who is educated, insured, and financially secure has escaped the effects of racism, but the weathering framework predicts, and the data confirm, that the bodily costs of navigating a discriminatory society accumulate regardless of achievement. This is one reason the Black and White disparity in maternal mortality persists, and in some analyses widens, across higher levels of education and income, with Black women who hold a college degree experiencing worse maternal outcomes than White women with far less education (Kaiser Family Foundation, 2025). If the disparity were simply a matter of poverty or schooling, it would close as those improve. It does not, which points to racism rather than race as the operative force (Njoku et al., 2023).

A third pathway connects structural racism back to the interpersonal bias of earlier sections. Structural and interpersonal racism are not separate problems but two expressions of the same system, and they reinforce each other. Structures determine who ends up in front of which clinician, under what time pressures, and with what prior documentation, and the clinician’s implicit biases then operate within those structurally shaped encounters (Montalmant & Ettinger, 2024). A patient routed by segregation to an under-resourced, overstretched facility meets clinicians who are more rushed, which, as Section 1 noted, is exactly the condition under which automatic associations exert the most influence. The structural and the interpersonal thus compound one another, and an account that addresses only one will leave the other in place. A comprehensive understanding of maternal health disparities, therefore, has to hold both levels at once. Addressing only individual bias while leaving the structures intact will leave most of the problem in place; addressing only structures while ignoring the daily encounters in which patients are believed or disbelieved will miss the mechanism by which much harm is delivered.

The history matters because it explains why the structures are so durable. Researchers operationalizing structural racism for empirical study stress that present-day inequities are the continuation of arrangements built over generations, and that structures established in one era persist through the institutions and geography they shaped long after the original policy is formally repealed (Sonderlund et al., 2022). The illustrative case of the spatial and racialized clustering of incarceration is instructive: incarceration patterns, themselves a product of structural racism, are associated with adverse birth outcomes in the surrounding communities, showing how a structure that has nothing obviously to do with pregnancy reaches into birth through the stress, instability, economic loss, and disrupted social ties it imposes on whole neighborhoods (Sonderlund et al., 2022). The general lesson for the team is that the causes of a disparity visible in the delivery room frequently lie far outside it, in domains, housing, criminal justice, education, and employment that no clinician controls but that together determine the health a patient brings to pregnancy.

A further structural pathway operates through the simple geography of where care is available. Access to maternity services is not evenly distributed, and communities that have been disinvested through the same historical processes of segregation and economic exclusion are also those most likely to face closures of obstetric units, shortages of perinatal providers, and long travel distances to the nearest facility capable of managing a complication (Montalmant & Ettinger, 2024). A patient who must travel far for care, or who has no nearby provider she trusts, is more likely to enter prenatal care late, to attend fewer visits, and to reach a higher level of care only after a complication has advanced. These access barriers are commonly described as social determinants of health, but it is important to see that they are not random misfortunes; they are the predictable products of the structural arrangements described in this section, which is why they fall along the same racial and economic lines as the outcomes they help to produce. For the interprofessional team, this pathway is a reminder that a patient’s pattern of engagement with care, how early she came, how reliably she attends, how quickly she seeks help, reflects the structure of access around her at least as much as it reflects any choice of her own, and that interpreting missed appointments or late presentation as a personal failing repeats the very error this course is meant to correct.

It is worth stating plainly what this section does and does not claim. It does not claim that any particular clinician is racist, nor that race is a biological cause of poor outcomes. The claim, grounded in the literature, is that a system built over generations distributes risk unequally by race, that this system is a major driver of the maternal health disparities documented in the next section, and that it continues to function through ordinary institutional practice unless it is deliberately changed (Sonderlund et al., 2022; Njoku et al., 2023). For the interprofessional team, the importance of naming structure is that it widens the field of possible action beyond the individual encounter, toward the policies, referral patterns, hiring, data practices, and institutional habits that the team and its organization can influence. It also guards against a demoralizing misreading of the earlier sections. If bias were purely a matter of individual minds, the persistence of disparities despite decades of good intentions would be baffling and discouraging. Once structure is in view, the persistence makes sense, and so does the strategy: change the systems, not only the minds.

Section 5: Maternal Morbidity and Mortality: Risk Factors and Patterns Across Populations

To act on maternal health disparities, the team needs a clear picture of the outcomes themselves. Two related terms anchor this section. Maternal mortality refers to the death of a person while pregnant or within a defined period after the end of pregnancy from causes related to or aggravated by the pregnancy or its management. Severe maternal morbidity refers to unexpected outcomes of labor and delivery that result in significant short-term or long-term consequences to a person’s health, a broader category that includes the many patients who survive a life-threatening event. Morbidity is far more common than mortality, and the two share many of the same risk factors and the same disparities, so that mortality, though rarer, serves as a visible index of a much larger burden of harm.

The clinical risk factors for maternal morbidity and mortality are reasonably well established and will be familiar to many on the team. They include cardiovascular conditions and hypertensive disorders of pregnancy, such as preeclampsia, hemorrhage, infection, venous thromboembolism, and complications of chronic conditions, such as diabetes and obesity, as well as mental health conditions and substance use disorders that contribute substantially to deaths in the year after birth (Njoku et al., 2023). Older maternal age raises risk, and access-related factors, including late entry into prenatal care or no prenatal care, lack of insurance, and limited access to facilities, compound clinical risk (Kaiser Family Foundation, 2025). Several features of these risk factors are worth emphasizing for an interprofessional audience. Many of the conditions are chronic and predate the pregnancy, which means the window for prevention opens long before conception and stays open well after delivery. Many of the deaths occur not during the dramatic hours of labor but in the weeks and months that follow. And the mental health and substance use contributions place behavioral health clinicians squarely among the disciplines whose work bears directly on maternal survival, not at the margins of it.

It is important to connect these risk factors back to the preceding sections rather than to read them as a list of individual patient failings. The chronic conditions that drive maternal risk are themselves unequally distributed by the structural forces of Section 4, and the weathering framework predicts their earlier onset among populations exposed to sustained discrimination (Njoku et al., 2023). The access-related factors are likewise structural rather than personal: whether a patient enters prenatal care early depends on insurance, transportation, work flexibility, the availability of nearby providers, and prior experiences of being treated with respect, far more than on individual motivation. And whether a developing complication is caught in time depends on the interpersonal dynamics of Section 3, on whether the patient’s early reports are believed and acted on. The risk factors, in other words, are where the structural and interpersonal forces described earlier become concrete clinical events.

The way these forces show up in maternity care has been documented directly. In obstetric anesthesia, for example, Black women have been found more likely than White women to receive general anesthesia for cesarean delivery, which carries greater risk than neuraxial anesthesia, and more likely to receive no analgesia for vaginal delivery (Tangel et al., 2020). This is exactly the kind of differential that the earlier sections predict: not a dramatic act of overt discrimination, but a pattern of treatment decisions that diverges by race in ways that accumulate into disparate risk. The same pattern appears in the management of pain more broadly, where Black and Hispanic patients have been shown to receive less effective pain treatment across a range of conditions, a disparity rooted partly in the historical and durable misconceptions about pain described in Section 2 (Aronowitz et al., 2020). For the perinatal team, these findings are not distant statistics but a description of the specific clinical decisions, about anesthesia, analgesia, and the seriousness with which pain is treated, through which bias becomes a measurable difference in the care a childbearing patient receives.

The timing data make this concrete and have direct implications for the team. Among pregnancy-related deaths, roughly one in five occur during pregnancy, nearly a quarter occur during labor or in the first week after birth, and more than half occur between one week and one year postpartum (Kaiser Family Foundation, 2025). The postpartum year, in other words, is not an afterthought but the period in which the majority of these deaths happen, which is precisely the period when physical therapists, occupational therapists, behavioral health clinicians, and pediatric-facing audiologists and speech-language pathologists are often the professionals a family sees most. A patient who has been discharged from obstetric care may still be in active contact with these disciplines, which positions them to notice the warning signs, severe headache, shortness of breath, chest pain, heavy bleeding, signs of severe mood disturbance, that can herald a late and preventable death. Equally important, the great majority of these deaths are considered preventable. Estimates indicate that more than eight in ten pregnancy-related deaths could be prevented (Kaiser Family Foundation, 2025). Preventability is the hopeful core of this material: these are not inevitable tragedies but outcomes that better systems, communication, and follow-up can avert, meaning the team’s vigilance is not a mere formality but a determinant of survival.

It is worth being concrete about what preventability means, because the abstract statistic can obscure the practical opportunity. When review committees judge a maternal death preventable, they identify points at which a different action could have changed the outcome, and those points are distributed across the system rather than concentrated in a single dramatic moment. They include missed or delayed recognition of warning signs, gaps in communication and coordination among clinicians, failures to escalate care in a timely manner, and breakdowns in follow-up after discharge, alongside patient and community factors and the access barriers discussed elsewhere in this course (Kaiser Family Foundation, 2025). Most of these are precisely the failures that an attentive, well-coordinated interprofessional team is positioned to catch. A warning sign recognized during a physical therapy session and escalated promptly, a postpartum mood disturbance identified by a behavioral health clinician and acted on, a feeding or developmental visit that becomes an occasion to ask a recovering parent how she is truly doing, each of these is a potential point of prevention. Seen this way, the preventability finding is not a distant policy abstraction but a description of the everyday clinical opportunities that the team already encounters, and a strong argument that improving how the team recognizes, communicates, and escalates is among the most direct contributions any discipline can make to maternal survival.

The patterns across populations are stark and the empirical heart of this course. In 2023, the overall United States maternal mortality rate was 18.6 deaths per 100,000 live births, a decline from 22.3 in 2022, with 669 maternal deaths recorded in 2023 compared with 817 in 2022 (Hoyert, 2025). That overall improvement, however, masks a persistent and severe racial disparity. In 2023, the maternal mortality rate for Black women was 50.3 per 100,000 live births, while the rates for White, Hispanic, and Asian women were 14.5, 12.4, and 10.7 per 100,000, respectively (Hoyert, 2025). The Black rate was more than three times the White rate and more than four and a half times the lowest group rate. While rates declined significantly for White and Hispanic women between 2022 and 2023, the rate for Black women did not improve and remained the highest by a wide margin (Hoyert, 2025). It is worth pausing on what these numbers represent. A rate of 50.3 per 100,000 is not an abstraction; it means that the everyday experience of pregnancy and birth carries a categorically different level of danger for Black women in the United States than for their White, Hispanic, and Asian counterparts, and that this has remained true even as the overall picture improved.

American Indian and Alaska Native populations require specific attention, both because the relevant learning outcome names them and because the standard reporting can render them invisible. National vital statistics typically report stable rates only for the largest race and Hispanic-origin groups, and for 2023, the data were insufficient to publish a reliable maternal mortality rate for American Indian and Alaska Native and for Native Hawaiian and Other Pacific Islander populations (Kaiser Family Foundation, 2025). This statistical suppression is itself a manifestation of how systems can obscure the populations at highest risk, since small population counts lead to fewer cells being detected, and the groups most affected become the hardest to see in the headline figures. Earlier data make the stakes clear: in 2021, American Indian and Alaska Native and Native Hawaiian and Other Pacific Islander people had pregnancy-related mortality rates of 118.7 and 111.7 per 100,000, respectively, the highest of any groups (Kaiser Family Foundation, 2025). Any honest account of maternal mortality across the five populations named in this course must therefore include the caveat that the groups facing some of the gravest risk are also the ones most often missing from the published rates, and that the absence of a number is not the absence of a problem.

It also bears emphasizing that mortality, however grave, is the narrow tip of a far larger burden. For every maternal death, many more patients survive a severe maternal morbidity event, a near miss in which hemorrhage, eclampsia, sepsis, cardiac failure, or another life-threatening complication is survived but often at the cost of lasting physical and psychological harm. These survivors carry the consequences into the postpartum year and beyond, where they intersect directly with the work of the rehabilitation and behavioral health professions: the patient recovering from an emergency hysterectomy, the patient managing the aftermath of a stroke or cardiac event, the patient whose traumatic birth has left both a physical injury and a mental health sequela. Severe maternal morbidity follows the same racial gradient as mortality, so that the populations dying at higher rates are also those surviving complications at higher rates, which means the inequity reaches a far wider group of patients than the mortality figures alone suggest (Njoku et al., 2023; Montalmant & Ettinger, 2024). For an interprofessional team, this is a reminder that addressing disparities is not only about preventing the rare death but about the much more common experience of harm that falls short of death and shapes a family’s health for years.

Two further facts about the disparity deserve emphasis because they shape how the team should think about causes. First, as noted in Section 4, the Black-White gap persists across education and income levels and is not explained by socioeconomic status, which directs attention to racism rather than individual patient characteristics (Kaiser Family Foundation, 2025; Njoku et al., 2023). This finding is worth repeating because it is the single most effective counter to the common but mistaken assumption that the disparity is really about class. Second, the persistence of the disparity in a period when overall rates improved suggests that whatever drove the general improvement did not reach Black patients equally, a pattern consistent with structurally unequal access to the gains. For the interprofessional team, the central message of this section is that maternal mortality is racially patterned, largely preventable, and concentrated in a postpartum window during which many team members are actively involved, which makes the team’s vigilance and coordination directly relevant to whether a patient lives.

Section 6: Infant Mortality and Its Ties to Maternal Health 

Maternal and infant health are bound together, biologically and socially, and the disparities in one are mirrored in the other. Infant mortality is defined as the death of an infant before the first birthday, and it is widely used as an indicator of the underlying health of mothers, the quality of perinatal care, and the social conditions in which families live (Ely & Driscoll, 2025). When a population’s mothers are sicker, less well cared for, and more stressed before and during pregnancy, its infants die at higher rates, which is why infant mortality cannot be understood apart from maternal health.

The national figures provide the baseline. In 2023, the United States infant mortality rate was 5.61 deaths per 1,000 live births, essentially unchanged from 2022, with 20,162 infant deaths recorded, a 2 percent decrease from the prior year (Ely & Driscoll, 2025). The leading causes of infant death are closely tied to the course of pregnancy: congenital malformations, disorders related to preterm birth and low birth weight, sudden infant death syndrome, injuries, and maternal complications of pregnancy (Kaiser Family Foundation, 2025). Several of these causes, particularly those stemming from preterm birth and low birth weight, are direct downstream consequences of maternal health and the conditions of pregnancy, which is the mechanical link between the two outcomes. An infant born too early or too small because of an inadequately managed maternal condition, a stressful pregnancy, or barriers to timely care begins life at a disadvantage that the best neonatal care cannot always overcome.

The racial patterning of infant mortality parallels the maternal data. Infants born to Black mothers die at more than twice the rate of infants born to White mothers, with Black infant mortality reported at roughly 10.9 per 1,000 live births compared with about 4.5 for White infants and 3.4 for Asian infants (Ely & Driscoll, 2025). American Indian and Alaska Native infants also experience elevated mortality relative to White and Asian infants. The leading contributors to the Black and White gap are the higher rates of preterm birth and low birth weight among Black infants, which again routes the explanation back through maternal health and the structural and interpersonal forces that shape it (Kaiser Family Foundation, 2025). The same disparities, the same upstream causes, and the same persistence across socioeconomic lines that characterize maternal mortality reappear in infant mortality data, providing strong evidence that the two are manifestations of a single underlying problem rather than two separate ones.

For the interprofessional team, this connection is not abstract. The infants who survive complicated pregnancies and difficult births are precisely the population that occupational therapists, physical therapists, speech-language pathologists, and audiologists follow most closely. A preterm infant may need feeding support from a speech-language pathologist, developmental and motor support from occupational and physical therapists, hearing surveillance from an audiologist, and respiratory support managed by a respiratory therapist while the parent navigates the physical recovery and the mental health consequences of a high-risk pregnancy and a fragile newborn, often with behavioral health involvement. These professionals therefore meet the downstream effects of maternal health disparities every day, frequently without naming them as such. Understanding that an infant’s developmental risk and a parent’s distress can both trace back to the same structural and interpersonal forces gives the team a more accurate picture of the family in front of them and a stronger rationale for coordinated, respectful, and vigilant care. It also reframes the team’s work as part of the prevention story: supporting maternal recovery, catching warning signs early, and sustaining a family’s trust in the care system all contribute, however indirectly, to better outcomes for the next pregnancy and the next generation.

The preterm birth pathway deserves a closer look because it is the mechanical hinge connecting maternal health to infant survival and because it carries the same disparity. Preterm birth and low birth weight are among the leading contributors to infant death, and they are more common among Black infants, which accounts for much of the Black and White gap in infant mortality (Kaiser Family Foundation, 2025). The forces that drive preterm birth are not separate from the forces described in earlier sections. The chronic stress captured by the weathering framework, the differential access to high-quality prenatal care produced by structural segregation, and the interpersonal dynamics that determine whether an emerging complication is caught and treated all feed into whether a pregnancy reaches term. An infant born too early because a maternal condition went unrecognized or untreated enters a neonatal course that the most skilled team can only partly rescue. Understanding this chain helps every discipline see that the infant outcomes they manage downstream were partly written upstream, in the quality and equity of the care the mother received, and that improving maternal care is, in a direct sense, infant care as well.

There is also a reciprocal direction to the maternal and infant relationship that matters for the postpartum team. A medically fragile newborn intensifies the stress, sleep disruption, and emotional strain on a recovering parent, raising the risk of perinatal mood and anxiety disorders at exactly the time when the parent is also managing their own physical recovery and the infant’s complex needs. Behavioral health clinicians supporting these parents, and the rehabilitation professionals helping both parent and infant regain function, are therefore working on two interlocking recoveries at once. Recognizing the dyad as a single unit of care, rather than treating mother and infant as separate cases, is consistent with the interprofessional emphasis on coordinated, whole-person care and positions the team to support the family in ways that fragmented, discipline-by-discipline attention cannot.

It is also worth stepping back to see why infant mortality carries such weight as a measure. Because an infant’s first year of life is so sensitive to the health of the mother, the quality of perinatal care, and the social conditions surrounding the family, the infant mortality rate is widely treated as a barometer of the underlying well-being of a whole population, not merely as a count of individual losses (Ely & Driscoll, 2025). When that barometer shows a rate for one group more than double the rate for another, and when the gap traces back through preterm birth and low birth weight to the same maternal health and structural forces examined throughout this course, the disparity is best read as a signal that the conditions supporting healthy pregnancy and infancy are themselves unequally distributed. For the interprofessional team, this reinforces a theme that runs through the entire course: the outcomes that appear at the very end of the causal chain, an infant lost, a mother dead, a complication survived with lasting harm, are produced by forces that operate long before and far upstream, many of which the team’s everyday practice can influence. Improving the equity of maternal care is therefore not a separate project from improving infant survival; it is, in large part, the same project viewed from a different point in the lifespan.

Section 7: The Interprofessional Team's Role in Reducing Bias and Structural Racism 

The preceding sections describe a problem with two levels, the interpersonal and the structural, and the response has to operate at both. This section turns to what the interprofessional team can actually do, with attention to what the evidence supports and what it does not.

Begin with a note of humility about training. Implicit bias training has become widespread and is now legally required for healthcare professionals in several states (Shah & Bohlen, 2023). Yet a systematic review of implicit bias training for healthcare providers and trainees found that such training is frequently characterized by methodological weaknesses and translational gaps, often diverging from the underlying science and lacking strong evidence of internal, face, and external validity, which means a single training module cannot be assumed to change behavior on its own (Hagiwara et al., 2024). This finding should not be read as a reason to abandon education, which builds the shared understanding this very course aims to provide, but rather as a reason to embed education within a larger system of change and to be skeptical of any claim that a single workshop has fixed the problem. The honest message to a team is that awareness is a starting point and that durable change requires ongoing reflection and structural support (Campbell, 2025; Hagiwara et al., 2024).

At the level of the individual encounter, several evidence-informed strategies can reduce the influence of automatic associations. Honest self-assessment is the starting point, and a widely used entry point is a confidential implicit-association measure, such as those hosted by Project Implicit, which can prompt reflection by making a clinician’s automatic associations visible, provided the result is treated as a conversation starter rather than a verdict (Kirwan Institute, 2013). Counter-stereotypic imaging and the deliberate practice of individuating, that is, focusing on the specific person rather than the group category, can weaken the grip of stereotypes, and perspective-taking and structured reflection can increase a clinician’s self-awareness and empathy, with simulation-based and reflective approaches showing some of the more consistent positive effects in the literature (Campbell, 2025). Practices that reduce cognitive load and stress, including mindfulness and emotional regulation techniques, can help because bias exerts the most influence precisely when a clinician is rushed, tired, and overwhelmed, which is the ordinary condition of much clinical work (Campbell, 2025). Underpinning all of these is the disposition of cultural humility, defined as the ability to maintain an other-oriented stance that remains open to the aspects of cultural identity most important to the patient rather than assuming expertise about the patient’s experience (Hook et al., 2013). None of these is a cure, and measured effects on implicit-association scores tend to be modest and may fade, which is why they work best as habits sustained over time rather than as one-time exercises (Campbell, 2025).

Closely related is the team’s and the organization’s capacity for cultural and linguistic competence. Cultural competence has been defined as a set of congruent behaviors, attitudes, and policies that come together in a system or among professionals to enable effective work in cross-cultural situations (Cross et al., 1989), while linguistic competence is the capacity of an organization and its staff to convey information in a way that is readily understood by patients of limited English proficiency, low literacy, or sensory disability (National Center for Cultural Competence, n.d.). The stakes are concrete: without it, encounters produce incomplete or inaccurate histories, misunderstandings of diagnoses and treatment, problems with medication and informed consent, and avoidable repeat visits. A practical, high-yield example is the use of trained, qualified medical interpreters rather than family members or friends, with attention to difficult-to-translate terms and to the physical arrangement of the encounter. For a perinatal team serving diverse families, including the audiologists and speech-language pathologists whose work centers on communication itself, building cultural and linguistic competence into routine practice is one of the more tangible ways to prevent bias and miscommunication from reaching patients.

The more powerful levers are at the team and system levels, where the interprofessional framing pays off. Standardizing assessment and decision processes, for example, using consistent protocols for evaluating and escalating warning signs such as postpartum pain, bleeding, or mood changes, reduces the room for unexamined judgment to determine who is taken seriously. Validated instruments allow a team or organization to measure awareness and mitigation of bias and to track whether interventions are actually changing practice rather than assuming they are (Bower et al., 2023). Building communication channels that allow concerns to travel across the hierarchy means that what a physical therapist or occupational therapist hears during a long session, or what a behavioral health clinician learns about a patient’s eroding trust, can reach the people who can act on it. Treating patient reports of disrespect or dismissal as safety data and reviewing adverse outcomes with an explicit eye to whether bias or unequal access played a role turns individual events into system learning. Each of these maps onto the interprofessional competency domains: shared values and ethics that center equity; clear roles and responsibilities so that no warning sign is anyone’s blind spot; communication that crosses disciplinary lines; and genuine teamwork that distributes the work of vigilance.

First Domain: Values and Ethics for Interprofessional Practice

It is worth taking each of those four domains in turn because they give the team a practical structure for organizing its response and convert a broad commitment to equity into specific, assignable behaviors. The first domain, values and ethics for interprofessional practice, asks the team to place the patient’s interest and dignity at the center of its shared work and to treat equity as a professional obligation rather than an optional virtue. In concrete terms, this means the team agrees that a patient’s report of pain or distress is to be believed and investigated regardless of the patient’s race, language, body size, or insurance status, and that respectful communication is held as a standard of care rather than a matter of individual temperament. When equity is named as a shared value, an individual clinician who slows down to take a dismissed concern seriously is acting on team policy rather than swimming against the current, and a clinician who treats a patient curtly can be understood to have departed from an agreed standard rather than merely having an off day.

Second Domain: Roles and Responsibilities

The second domain, roles and responsibilities, asks each profession to understand both its own contribution and those of the others well enough that important observations do not fall through the cracks between disciplines. In the perinatal context, this means that the physical therapist knows that the postpartum pain she is hearing about could signal a complication that belongs to the obstetric team, that the speech-language pathologist understands that a parent’s flat affect during a feeding session may be relevant to the behavioral health clinician, and that everyone knows who is responsible for escalating a warning sign and how. Role clarity is a safeguard against bias because much of the harm described in this course occurs in the spaces between professionals, where each assumes someone else has noticed and acted. A team that has mapped its roles explicitly leaves fewer such spaces.

Third Domain: Interprofessional Communication

The third domain, interprofessional communication, asks the team to communicate in a manner that is responsive, respectful, and sufficiently structured to reliably carry concerns across the hierarchy. The relevance to bias is direct, because the distributed vantage point described earlier, in which the patient confides in the clinician she trusts most, is useless if that clinician has no dependable way to convey what she heard to the people who can act, or no confidence that doing so is welcome. Structured handoffs, shared documentation that travels with the patient, and an explicit norm that raising a colleague’s possible blind spot is an act of good care rather than an accusation all serve this domain. Communication is also where trust is protected or damaged, so the team’s communication habits with the patient, not only among its members, fall under this domain.

Fourth Domain: Teams and Teamwork

The fourth domain, teams and teamwork, asks the team to apply the values, the role clarity, and the communication into a coordinated practice that performs better than the sum of its individuals. This is the domain in which the standardized protocols, the measurement of bias awareness, the review of adverse outcomes, and the flattening of hierarchy live, because each of these is a property of how the team works together rather than of any one member’s knowledge or goodwill. The evidence reviewed throughout this course points consistently toward this conclusion: because bias is a feature of normal cognition that no amount of individual effort fully eliminates, and because structural forces operate above the level of any single encounter, the durable protections are the ones built into the team’s shared routines (Campbell, 2025; Hagiwara et al., 2024; Bower et al., 2023).

Structural racism, by definition, is not solved at the bedside, but the team and its organization are not powerless before it. Within their sphere, teams can examine their own referral patterns and access barriers, advocate for the data practices that make high-risk populations visible rather than suppressed, support workforce diversity and culturally responsive care, and align with institutional and policy efforts aimed at the upstream causes (Montalmant & Ettinger, 2024; Sonderlund et al., 2022). The point is not that an audiologist or a physical therapist can dismantle structural racism single-handedly, but that the structures are made of countless ordinary practices, many of which a team can influence, and that naming the structural level keeps it from being treated as someone else’s problem.

Measurement

Measurement deserves a place of its own in the team’s response, because without it good intentions cannot be distinguished from good results. A team that wishes to know whether it is reducing the harm of bias has to look at data, and several kinds are available. Validated instruments can assess awareness of and mitigation of bias within a maternity care setting, providing a team with a baseline and a way to detect change over time, rather than relying on the assumption that training or a policy must have helped (Bower et al., 2023). Stratifying clinical and experience measures by race and ethnicity, for example, examining whether pain is assessed and treated equally, whether warning signs are escalated equally, and whether patients of different groups report equivalent respect and trust, can make a hidden disparity visible and turn it into an actionable target. Reviewing severe morbidity and mortality events with an explicit question about whether bias or unequal access contributed converts individual tragedies into system learning. None of these measurements is punitive in spirit; its purpose is to replace the comfortable assumption that a team is doing well with evidence about whether it actually is, and to direct effort toward the points where the data show the largest gaps. This emphasis on measurement also guards against the documented risk of mistaking a single educational intervention for a solution, since the data will make it plain whether practice has changed (Hagiwara et al., 2024).

Care Models and Programs That Have Reduced Maternal Harm

The strategies in this section are not only theoretical; they are embodied in established programs whose results have been measured. The leading national example is the Alliance for Innovation on Maternal Health (AIM), a federally funded program coordinated through the American College of Obstetricians and Gynecologists that develops and disseminates maternal patient safety bundles, which are standardized, evidence-based sets of practices aimed at the leading preventable causes of maternal death and severe morbidity, including obstetric hemorrhage, severe hypertension and preeclampsia, sepsis, and perinatal mental health and substance use conditions. As of early 2025, AIM bundles were being implemented in roughly 2,000 birthing facilities across 49 states, the District of Columbia, and Puerto Rico (Health Resources and Services Administration, 2025). Each bundle is organized around the domains of readiness, recognition, response, reporting, and respectful care, and the deliberate inclusion of respectful and equitable care places the trust and bias concerns of this course within the safety framework rather than beside it (American College of Obstetricians and Gynecologists, n.d.).

This model has a measurable track record. The California Maternal Quality Care Collaborative (CMQCC) pioneered bundle-based perinatal quality improvement, and in a peer-reviewed evaluation of 99 collaborative hospitals, patients experiencing obstetric hemorrhage showed a 20.8 percent reduction in severe maternal morbidity compared with a 1.2 percent reduction in non-participating comparison hospitals, with a larger 28.6 percent reduction in hospitals that had prior collaborative experience (Main et al., 2017). At the state level, California’s maternal mortality rate fell by roughly 65 percent between 2006 and 2016, a period when the national rate was rising, a decline that CMQCC attributes substantially to these collaborative quality-improvement efforts (California Maternal Quality Care Collaborative, n.d.).

Two features make these programs directly relevant to this course. First, the gains are produced through precisely the team and system mechanisms described above, namely standardized recognition and escalation, rehearsed emergency response, and the routine review of data, rather than through individual exhortation; the bundles have since been revised to embed respectful and equitable care, and collaboratives increasingly stratify outcomes by race and ethnicity so that improvements reach the patients at highest risk (American College of Obstetricians and Gynecologists, n.d.; Montalmant & Ettinger, 2024). Second, they give the interprofessional team a concrete picture of what success looks like: a patient is less likely to die or to suffer severe harm when every discipline recognizes the warning signs, the team has practiced its response, and the data are watched to see who is being left behind. For the disciplines represented in this activity, these programs are the structural counterpart to the individual habits of listening and believing, and they show that the layered response this course recommends is not aspirational but already at work.

Finally, every team member's role includes the conduct of individual relationships, where evidence and ethics converge. Listening to patients, believing their reports of their own bodies, communicating respectfully, and protecting trust are not soft additions to clinical care but, in a period when most maternal deaths are preventable, and many depend on a patient’s warning being heard, are themselves safety interventions. The interprofessional team’s role in reducing the impact of implicit bias and structural racism for pregnant and postpartum patients is therefore both modest and essential: modest because no single profession or person controls the outcome, and essential because the outcome is built from exactly the encounters, hand-offs, and habits that the team controls together.

Pulling these threads together, the response the evidence supports is layered rather than singular. At the level of the individual clinician, it asks for honest, sustained self-examination and for the everyday discipline of listening to patients and believing their reports of their own bodies. At the level of the team, it asks for shared values that center on equity, clear roles so that warning signs are no one’s blind spot, communication that carries concerns reliably across the hierarchy, and coordinated routines that perform better than the sum of their members. At the level of the system, it calls for standardized processes, measurements that reveal disparities rather than hide them, and engagement with structural and access barriers that route patients to unequal care. No single layer is sufficient on its own: individual goodwill without team structure leaks; team structure without individual listening has nothing to carry; and both together still operate within structures that must be named and addressed. One way to hold these layers together is the image of a cycle that can turn in either direction: structural determinants and biased decision-making can feed a vicious cycle of diminished care and worse outcomes, while sound decision-making and supportive structures can feed a virtuous cycle of improved care and better outcomes, with the clinician and the team positioned at the point where the cycle can be redirected (Vela et al., 2022). What makes this an interprofessional rather than a single-discipline task is that each of these layers depends on contributions from every profession that touches the childbearing family, from behavioral health to audiology, and that the safety of the patient emerges from how those contributions are coordinated. The hopeful conclusion, grounded in the finding that the great majority of maternal deaths are preventable, is that this is a problem largely within the team’s power to influence, provided the team takes it up together rather than leaving it to the conscience of individuals (Campbell, 2025; Hagiwara et al., 2024; Bower et al., 2023; Kaiser Family Foundation, 2025).

Application to Team-Based Practice

The material in this course reaches each represented profession differently, and naming those connections helps every reader locate their own role. Behavioral health clinicians often hold the relationships in which a patient feels safe enough to disclose distrust, perceived discrimination, or worsening symptoms, and they treat the perinatal mood and anxiety disorders that contribute to maternal deaths in the postpartum year, which places them at the center of both the trust and the clinical-risk dimensions of this material. Physical therapists frequently spend extended, hands-on time with patients managing pelvic floor dysfunction, musculoskeletal pain, and postpartum recovery, and that time is an opportunity both to hear concerns that were not voiced elsewhere and to model the practice of believing a patient’s report of pain, a domain with a documented history of biased underestimation. Occupational therapists support feeding, infant development, and a parent’s return to daily roles, often in the patient’s own environment, where they can observe context that the clinic never sees and where the power differential may relax enough for honest exchange. Speech-language pathologists and audiologists conduct newborn feeding and hearing screenings and follow developmentally at-risk infants across repeated visits, which makes them trusted longitudinal contacts for families navigating the downstream consequences of high-risk pregnancies and positions them to notice and escalate parental concerns over time. Respiratory therapists manage the ventilation and oxygen support that medically fragile newborns rely on and assist with the maternal respiratory complications that can arise during pregnancy and the postpartum period, which places them among the clinicians most likely to encounter warning signs such as shortness of breath and to act on them quickly.

Across all of these roles, the shared behaviors are the same. Each profession makes rapid judgments under pressure that are vulnerable to implicit associations, each occupies a position in a power differential relative to the patient, each meets families who carry the effects of structural forces they did not choose, and each can either protect or undermine a patient’s trust in the care system. The interprofessional opportunity is to ensure that the distinct vantage points add up to more than their sum, so that a concern heard by one discipline reaches the whole team, and so that the values, role clarity, communication, and teamwork that reduce the harm of bias are built into how the team works rather than left to individual goodwill.

A few illustrative situations show how these principles become concrete and how the team functions as a single safeguard. Consider a postpartum patient who mentions to her physical therapist, during a session for pelvic pain several weeks after delivery, that she has also been having headaches and some shortness of breath, symptoms she has not raised with anyone else because a previous clinician seemed to brush off her concerns. The physical therapist who has absorbed this material recognizes two things at once: that these symptoms can signal a serious late postpartum complication, and that the patient’s reluctance to report them is itself a product of an eroded trust that bias can cause. The values domain tells the therapist to take the report seriously; the roles domain tells her that escalation belongs to the obstetric or emergency pathway; the communication domain gives her a reliable way to convey the concern; and teamwork ensures the handoff is received rather than lost. The same chain could begin with an occupational therapist noticing a parent’s distress during a home visit, a speech-language pathologist observing that a parent seems disengaged during a feeding session, or a behavioral health clinician learning that a patient has stopped trusting her care team. In each case, the discipline that first receives the information is not expected to manage the complication alone, but to recognize its significance and to move it into the team’s shared field of action.

These scenarios also illustrate why the individual relationship and the team structure are complementary rather than competing. The therapist’s decision to believe the patient and to respond with curiosity rather than dismissal is an individual behavior, available to any clinician without waiting for a system change, and it is exactly the behavior that keeps a patient willing to disclose. But whether that disclosure leads to timely care depends on the team structures around it, the channels, the role clarity, and the shared norms that carry the concern to the people who can act. A clinician with excellent personal habits embedded in a team that has no way to escalate her observations will still watch concerns evaporate, and a well-structured team staffed by clinicians who do not listen will have nothing worth escalating. The interprofessional response to bias, therefore, asks for both at once: the individual disposition to listen, believe, and respect, and the collective structures that turn what is heard into what is done.

Summary

This course examined implicit bias and its impact on maternal health in minority populations, building from individual psychology toward structural forces and the outcomes at stake, and it returns now to each learning outcome. It differentiated implicit bias, which operates automatically and outside awareness, from explicit bias, which is consciously held, and explained how both can shape clinical interactions even in well-intentioned clinicians (LO1). It described how stereotypes, prejudice, and “isms” form and evolve through lifespan socialization, and identified self-assessment instruments, reflective practice, and continuing education as resources for examining one’s own biases, while noting that awareness is a beginning rather than an end (LO1). It examined how bias and power differentials affect the patient and the patient-team relationship, eroding communication and trust in ways that pose safety concerns in the perinatal period (LO2). It gave examples of structural racism, described how it forms and accumulates historically, and explained its impact on maternal health through pathways such as differential access to quality care and the cumulative toll of chronic stress (LO3). It identified the major risk factors for maternal morbidity and mortality and described the patterns across populations, including a 2023 Black maternal mortality rate more than three times the White rate, the persistence of that disparity across education and income, and the elevated and often statistically suppressed rates among American Indian and Alaska Native and Native Hawaiian and Other Pacific Islander populations (LO3). Finally, it described United States infant mortality and its ties to maternal health and pregnancy-related complications, and summarized the interprofessional team’s role in reducing the impact of implicit bias and structural racism for pregnant patients, a role that is modest at the level of any one profession and essential at the level of the team’s shared encounters, hand-offs, and habits (LO4). The through-line is that most maternal deaths are preventable, that the harm is racially patterned and structurally produced, and that the perinatal team, working as a team, holds many of the levers that determine whether a patient is heard, believed, and kept safe.

References

Please see the additional handout.