Minorities and medical research: Who is still excluded?

Medical research likes to introduce itself as a universal love language: randomized, controlled, double-blind, and “for everyone.”
Then you look at who actually shows up in the data…and it’s more like a group chat where half the people never got the invite.

The result is an evidence gap: treatments, devices, and guidelines built on a slice of the populationand then applied to the whole
country like a “one-size-fits-all” hoodie that fits exactly one mannequin.

Why representation isn’t a “nice-to-have”

When groups are underrepresented in medical research, the consequences aren’t abstract. They show up as missed diagnoses, wrong doses,
delayed treatment, and “we’re not sure” labels where people need clarity. If a clinical trial doesn’t reflect the real-world patients
who will use a drug, the findings can be less reliable, less generalizable, and sometimes downright misleading.

And it’s not just drugs. Research shapes medical devices, screening guidelines, risk calculators,
and even the reference images used to train clinicians. In other words: it’s not only who gets included in a trialit’s who gets included
in the definition of “normal.”

The progress we’ve made (and why it hasn’t been enough)

The U.S. has tried to address exclusion for decades. In the early 1990s, federal policy pushed NIH-funded clinical research to include women
and members of racial and ethnic minority groups. That was a major shiftand it did move the needle.

But policies don’t automatically rewrite the day-to-day realities of recruitment, eligibility criteria, site selection, language access,
transportation, time off work, childcare, disability accommodations, and the trust needed to participate. Put simply:
the rule exists, but the friction is still real.

Who is still excluded (or under-included) in U.S. medical research?

“Minorities” isn’t one group, and exclusion isn’t one problem. It’s a patchwork of who gets filtered outand why.
Here are the populations that still too often end up underrepresented, overlooked, or treated as an “edge case.”

1) Racial and ethnic minoritiesespecially Black, Hispanic/Latino, and Indigenous communities

Underrepresentation of racial and ethnic minority groups remains one of the most discussed gaps in clinical trials. In some areaslike cancer,
cardiovascular disease, and rare diseasesthe mismatch between who is sick and who is studied can be striking.

Barriers often include historical and ongoing mistrust, fewer trial sites in minority-serving communities, limited outreach from clinicians,
eligibility rules that exclude common coexisting conditions, and logistical hurdles like transportation and unpaid time away from work.
(Research participation is hard when the study schedule assumes you have a flexible job, a car, and a backup you.)

2) People with darker skin tones in device testing and clinical “reference standards”

Some of the most visible examples of exclusion show up in medical technology and training materials. A widely discussed case is
pulse oximetrythe clip-on device that estimates blood oxygen levels. Evidence has raised concerns that performance can vary
across skin pigmentation, which matters because oxygen readings can drive critical clinical decisions.

Another quieter but impactful issue: medical images. Dermatology resources and lecture materials have historically included
fewer examples of conditions on darker skin, which can delay recognition and treatment.
If the “textbook rash” is mostly shown on light skin, clinicians may miss it when it shows up differently.

3) Pregnant and lactating people

Pregnant and lactating people are frequently excluded from clinical trialseven when the medication or intervention is likely to be used
during pregnancy. That creates a familiar problem: decisions get made in the clinic with limited direct evidence.

The irony is painful: exclusion is often justified as “protective,” yet it can leave clinicians and patients without clear guidance on dosing,
safety, and effectivenessespecially during public health emergencies when real-world use happens anyway.

4) Older adults (and sometimes children): the “real patients” who get screened out

Older adults are a core population for many diseasesyet trials have historically used age cutoffs or excluded participants with multiple
chronic conditions, even though that’s basically the job description of being over 70 in America.

NIH policy now emphasizes “inclusion across the lifespan,” pushing researchers to justify age-based exclusions and to design studies that
answer questions for the people actually affected by the condition.

5) People with disabilities

Disability can be an invisible exclusion criterion: trial protocols may require frequent in-person visits, complicated instructions,
or technology access. Some studies embed eligibility criteria that unintentionally block participation without a scientific reason.

The result is a double loss: people with disabilitieswho may have higher rates of certain conditionsare left out of the evidence base,
and research misses crucial insights about accessibility, safety, and outcomes.

6) Sexual and gender minorities (LGBTQ+), especially where SOGI data isn’t collected

Many studies still don’t consistently collect sexual orientation and gender identity (SOGI) data. If the research form never asks,
the dataset can’t tell you what’s happening. This can mask disparities in screening, mental health outcomes, cancer risks, HIV prevention and care,
and more.

It’s hard to fix what you can’t measureand you can’t measure what you refuse to record.

7) People with limited English proficiency (LEP) and low health literacy

Clinical research runs on informed consent, but consent documents can be long, technical, and written above recommended readability levels.
Add language barriers, limited interpreter access, and time pressureand LEP participants can be effectively excluded even when they’re eligible.

This isn’t just a translation problem. It’s a design problem. If the process assumes everyone reads like a lawyer and speaks fluent medical-ese,
you’re going to end up studying the people who do.

8) Incarcerated people

Research involving incarcerated people is governed by additional protections because of the risk of coercion and exploitation.
Those safeguards are necessarybut they also mean participation is rare and complex, even when incarcerated populations face high burdens of illness.

9) Rural communities and people with fewer economic resources

Where trials are located shapes who can participate. If studies cluster in major academic medical centers, rural participants may face hours of
travel for each visit. Add fuel costs, missed work, and limited broadband access for digital tools, and you’ve built a system that quietly selects
for convenience.

The mechanisms of exclusion: how the “evidence gap” is manufactured

Exclusion isn’t always explicit. Often, it’s baked into the process:

• Eligibility criteria that exclude common real-world conditions (diabetes, kidney disease, multiple meds) to create a “clean” sample.
• Site selection that favors research hospitals over community clinics where many patients receive care.
• Operational burdens: frequent visits, rigid schedules, long consent forms, complex instructions.
• Cost and logistics: transportation, childcare, meals, time away from work.
• Language access gaps: limited translated materials and interpreter resources.
• Trust and history: communities remember when research treated them as raw material instead of partners.

Notice what’s missing from that list: “people don’t want to participate.” People participate all the timewhen research is accessible, respectful,
and designed for humans who have jobs, families, and a low tolerance for paperwork that reads like a microwave manual written by a hedge fund.

Trust isn’t PRit’s infrastructure

Mistrust in medical research isn’t irrational; it’s informed. In the U.S., the legacy of unethical studiesincluding the federal government’s
syphilis study at Tuskegeestill shapes community perceptions. Add modern experiences of bias in healthcare, unequal treatment,
and data misuse fears, and trust becomes a practical barrier, not a vague “attitude.”

Ethical lapses have also shaped how people think about tissue and data. The story of Henrietta Lacks and the HeLa cell line is often cited
as a turning point in public conversations about consent, respect, and who benefits from biomedical discovery.

Genomics and precision medicine: a new frontier, same old imbalance

Precision medicine promises tailored care based on genetics, environment, and lifestyle. But genomic databases have long overrepresented people
of European ancestry, creating gaps in discovery and in the accuracy of risk prediction for other populations.

Indigenous communities add another essential dimension: data sovereignty. Tribal Nations have legitimate concerns about how biospecimens
and data are collected, stored, shared, and reused. Responsible genomics requires governance that honors Tribal authority and community consentnot
just individual signatures on a form.

What’s changing now: policies, pressure, and better playbooks

In the past few years, the push for diversity in clinical research has become more explicit and more operational.
Federal agencies and research leaders have emphasized that representation affects scientific qualitynot just fairness.

FDA: toward clearer expectations for diversity planning

FDA initiatives increasingly emphasize structured planning for enrolling underrepresented populations, including guidance around
Diversity Action Plans in certain clinical studies. Meanwhile, FDA’s “Drug Trials Snapshots” program highlights who was included in key studies,
making demographics more visible to clinicians and the public.

The policy landscape can also shift with politics. For example, in January 2025, reporting indicated turbulence around federal diversity-related
clinical trial guidance and messaging. That uncertainty is a reminder that durable progress needs institutional commitmentnot just a good year
of headlines.

NIH: inclusion policies, lifespan focus, and large-scale cohort efforts

NIH continues to reinforce inclusion expectations across race/ethnicity and sex, while also emphasizing age inclusivity through
“Inclusion Across the Lifespan.” Large cohort initiatives, including the All of Us Research Program, have also prioritized recruiting communities
historically underrepresented in biomedical research.

How to fix exclusion without turning trials into chaos

The goal isn’t to make every study perfectly mirror the U.S. census. The goal is to build evidence that actually applies to the people who will use it.
Here’s what worksbecause it addresses the real barriers:

Design for participation, not just publication

• Choose sites in community settings, not only major academic centers.
• Offer evening/weekend appointments and flexible visit windows.
• Use mobile or home visits when possible, and reduce unnecessary in-person requirements.

Pay for the friction

• Provide transportation support, childcare stipends, meals, and lodging when needed.
• Compensate participants fairly for timeespecially when visits require missed work.

Make language access a core budget item

• Translate materials with community review (not just literal word swaps).
• Build interpreter access into recruitment and follow-upnot as a last-minute scramble.
• Use consent formats that are readable, navigable, and respectful.

Build trust through shared power

• Partner with community organizations earlybefore the protocol is “final.”
• Share results back to participants and communities in plain language.
• Be transparent about data use, storage, and potential future reuse.

Stop excluding “complex” patients by default

Real patients have comorbidities. They take multiple medications. They miss appointments sometimes. If a study excludes everyone who resembles
real life, the results can become less helpfuleven if they look beautifully tidy in a journal figure.

Bottom line: the excluded are often the ones most affected

The painful pattern is this: the groups with the highest disease burden or the greatest barriers to care are often the groups most likely to be
underrepresented in the research that guides treatment. That isn’t just an equity issueit’s a scientific quality issue.

If the next generation of medicine is going to be more precise, it has to be more inclusive. Otherwise, “precision” will just mean
“accurate for some people, vibes for everyone else.”

Experiences from the real world: what exclusion feels like (and what helps)

A patient who wants to helpuntil the process makes it impossible

Many people say they’re open to research in theory, but the logistics can feel like a stress test disguised as a consent form.
A trial might require visits during business hours, multiple trips to a distant hospital, and paperwork that reads like it was written by someone
who gets paid per syllable. For a working parent, a caregiver, or someone juggling two jobs, “free participation” can still cost hundreds of dollars
in missed wages, childcare, and transportation. The experience becomes a quiet message: this wasn’t designed with you in mind.
When studies cover travel, offer flexible scheduling, and communicate clearly, that same person often goes from “I can’t” to “I’m in.”

A community that remembers, even when the brochure doesn’t

In many minority communities, hesitation isn’t about misinformationit’s about memory. People recall stories of unethical research, unequal treatment,
and being talked at instead of talked with. When researchers show up only to recruit, trust tends to evaporate. But when partnerships are built early
(through local clinics, faith groups, and community organizations), participation can risebecause respect is visible. Communities often respond well
when they can ask direct questions: Who owns the data? Will results come back to us? What happens after the study ends? The best experiences are the
ones where researchers answer plainly, share power, and keep showing up after enrollment closes.

A clinician who wants to refer patientsif the system actually supports it

Clinicians often say they’d like to offer trials more consistently, but time, workload, and fragmented referral systems get in the way.
In some settings, trial information isn’t integrated into routine workflows, and the easiest path becomes “standard care” by default.
This can hit minority and rural patients especially hard when trials are concentrated at large centers. When systems make referral easierclear eligibility
summaries, patient navigators, transportation support, and multilingual materialsclinicians report that it’s simpler to bring trials into the conversation.
The experience shifts from “extra work” to “a real option.”

A researcher who learns the hard truth: eligibility rules can discriminate

Many researchers describe an “aha” moment when they realize exclusion criteria can function like a sieve that catches certain groups more than others.
Rules that sound neutrallike limiting participants with kidney disease, multiple medications, or missed appointmentscan disproportionately exclude people
who already face health inequities. Some teams respond by revisiting what’s truly necessary for safety versus what’s simply convenient for clean data.
They pilot remote visits, reduce unnecessary labs, and budget for participant supports. The surprising outcome? Data quality often improves because the sample
becomes more realistic, retention gets better, and findings become more applicable to everyday clinical care.