Race-Adjusted Medicine May Disadvantage Black Patients

Imagine taking a medical test where the result changes depending on which box you check on a form.
Not your blood pressure. Not your symptoms. Not your lab value. Just… the box.
If that makes your eyebrows climb into your hairline, you’re not alone.

“Race-adjusted” (or “race-corrected”) medicine shows up when a calculator, equation, or guideline
uses race as an input or applies different “normal” ranges by race. Sometimes it’s obvious.
Sometimes it’s buried in a dropdown menu or baked into a lab report you never see.
And while these adjustments were often created with the intention of improving accuracy,
they can end up delaying diagnoses, limiting treatment options, or reducing access to specialty careespecially for Black patients.

What “Race-Adjusted Medicine” Means (and Why It Exists)

Race-adjusted medicine typically means one of two things:

• Race is an input in an algorithm (for example, a risk calculator includes “Black” as a factor).
• Race changes the reference standard (for example, “normal” lung function differs by race, so the same test score is interpreted differently).

Historically, these adjustments were justified as shortcuts for biologyideas like average differences in muscle mass,
physiology, or disease risk. The problem is that race is a social category, not a neat genetic one.
People who identify as Black can have wildly different ancestry, environment, nutrition, exposures, stress levels,
and access to care. So when race gets treated as biology, medicine can drift into “precision guessing”
rather than precision care.

Even when a race adjustment improves prediction in a dataset, it can still cause harm in real lifeespecially when:
(1) the adjustment pushes someone across a threshold that affects access, eligibility, or urgency, and
(2) the reasons behind group differences are actually structural (environment, inequity, exposure), not innate.

How a Checkbox Can Change Care

Let’s get specific. Race adjustment isn’t an abstract philosophy debate. It’s a practical issue with practical consequences.
Here are several major examples that have shaped today’s conversation.

Example 1: Kidney Function and the eGFR Race Coefficient

Estimated glomerular filtration rate (eGFR) is a common way clinicians estimate kidney function from a blood test (creatinine),
sometimes with other markers. For years, a widely used equation included a race coefficient that raised eGFR values for patients
identified as Black. In plain English: the same creatinine level could look “less concerning” on paper for a Black patient than for a non-Black patient.

Why does that matter? Because eGFR thresholds can influence:
whether chronic kidney disease is diagnosed,
when a patient is referred to nephrology,
when certain medications are started or adjusted,
andcriticallywhen someone is evaluated for transplant or placed on a waiting list.

In response to concerns that race-based eGFR could delay care for Black patients,
major kidney organizations convened a task force and recommended adopting newer race-free equations.
Newer approaches can use creatinine-based formulas without a race variable, and some equations incorporate cystatin C
(a different marker) to improve accuracy without relying on race.

The takeaway is not that eGFR is “bad.” It’s that a lab number should not swing upward or downward based on race
in a way that systematically delays recognition of disease in a group already facing higher burdens of kidney illness.

Example 2: Kidney Transplant Access and “Lost” Waiting Time

When race-based eGFR inflates kidney function estimates for Black patients, it can postpone the moment a patient crosses a clinical threshold
that triggers referral, evaluation, or waitlist eligibility. If the system starts the clock later, the patient’s place in line can be harmed.
And in transplant medicine, “later” isn’t just inconvenientlater can be life-changing.

Recognizing this, national transplant policy in the U.S. moved toward requiring race-neutral eGFR calculations.
But that still left a painful question: what about people who were already evaluated under the older approach?

In response, policy pathways were created to modify waiting time for certain Black kidney candidates who may have been disadvantaged by prior
race-inclusive eGFR calculationsessentially a way to restore time that might have been earned under a race-neutral method.
It’s an unusual step in health policy: an attempt not only to fix the calculator going forward, but to address the backward-facing damage.

If you want a crisp definition of why race adjustment can be harmful, it’s this:
a formula can quietly become a gatekeeper.
And gatekeepers don’t need bad intentions to produce bad outcomes.

Example 3: Lung Testing and Race-Based Spirometry

Spirometry and other pulmonary function tests compare a person’s results to “predicted normal” values.
For decades, many reference equations used race or ethnicity, with lower predicted values for some groups, including Black patients.
That means a Black patient and a White patient could blow the same numbers into the tubeand one might be told their lungs are “normal”
while the other is told something is wrong.

Critics argue that race-based “normal” values can under-diagnose disease or delay recognition of impairment,
because the bar for “abnormal” is effectively lowered. In response to growing evidence and debate, pulmonary leaders and organizations
have called for more race-neutral approaches and more transparent interpretation frameworks.

Here’s the uncomfortable humor of it: if your lungs are “normal” because society treated your neighborhood’s air like a suggestion,
that isn’t normalit’s normalized harm.

Example 4: VBAC Calculators and Childbirth Choices

A VBAC (vaginal birth after cesarean) calculator estimates the likelihood that a person who previously had a C-section
will have a successful vaginal delivery in a subsequent birth. Older versions of a widely used U.S. VBAC calculator included race and ethnicity.
The effect wasn’t subtle: it systematically lowered predicted success rates for Black and Hispanic patients.

Prediction tools often influence counseling, risk framing, and whether a trial of labor after cesarean is encouraged or discouraged.
So when race reduces the predicted chance of success, it can tilt decisions toward repeat surgerysometimes even when the individual’s clinical profile
looks favorable.

In recent years, the VBAC calculator was updated to remove race and ethnicity,
and professional guidance has emphasized that calculators should be used as one input in shared decision-makingnot as a verdict.

Example 5: Heart Risk Scores and Preventive Medications

Cardiovascular risk calculators help estimate a patient’s risk of heart attack or stroke over the next decade and can guide discussions about statins,
blood pressure management, and other preventive strategies. Some widely used models are sex- and race-specific, including separate equations
for Black and non-Black patients.

Supporters argue that race-specific equations can improve calibration within the populations they were derived from.
Critics counter that race may act as a proxy for inequities (like access to care, chronic stress, neighborhood conditions)
and that “race-specific” can harden social patterns into biological fateespecially if the output is treated as destiny rather than a prompt
for deeper assessment.

This is the pattern across medicine: once race sits inside the math, it gains authority.
The number feels objectiveeven when the assumptions are not.

Why Race “Adjustments” Often Miss the Mark

1) Race is not a gene

Race categories are broad, inconsistent, and shaped by history and politics. They can vary by country, by era, and even by who’s doing the classifying.
Using race as a biological stand-in can conceal the true drivers of health differenceslike environment, access, discrimination, and cumulative stress.

2) Proxy variables create proxy harms

If an algorithm uses race to approximate a trait (say, muscle mass, exposure, or risk), it can misclassify individuals.
And when that misclassification consistently pushes one group toward less care, fewer referrals, or delayed eligibility,
the proxy becomes a mechanism of inequity.

3) Thresholds are where small shifts become big consequences

Many clinical decisions aren’t gradual. They’re threshold-based:
“If eGFR is below X, do Y.”
“If risk is above Z%, consider medication.”
“If predicted success is low, recommend a different plan.”
A modest race adjustment can move a patient across the linechanging what options are offered.

4) Algorithms can create feedback loops

If a tool predicts poorer outcomes for a group and clinicians act on that prediction by offering fewer opportunities,
the prediction can become self-fulfilling. Less access leads to worse outcomes, which “validates” the model,
which further limits access. It’s a loop with a lab coat on.

But Don’t We Need Accurate Prediction? The Real Debate

It’s tempting to frame this as: “Remove race and everything becomes fair.”
Reality is messier.
A major scientific and policy challenge is that racial inequities are realand show up in data.
If you remove race without replacing it with better variables, you can accidentally reduce predictive performance,
hide disparities, or shift error onto the people already at risk.

That’s why many experts argue for a more disciplined standard:
don’t ask, “Should we use race?”
Ask, “What is race standing in for, and can we measure that directly?”

For example, instead of using race as a proxy, tools might incorporate:
individual clinical markers (like cystatin C, blood pressure control, hemoglobin A1c),
environmental exposures (air quality, occupational risks),
access-to-care variables (medication affordability, insurance disruptions),
and social risk measures (housing instability, transportation barriers),
while transparently reporting performance across groups.
The goal is not to become “race-blind,” but to stop letting race do scientific work it cannot reliably do.

In other words: replacing race in the math doesn’t mean ignoring inequity.
It means measuring inequity with better instruments than a checkbox.

What Good Care Looks Like When We Stop “Correcting” for Race

For clinicians and health systems

• Audit your tools. Make a list of calculators, lab equations, and guideline shortcuts used in your workflow. Identify which include race.
• Adopt race-neutral standards when recommended. For kidney care, ensure your lab and EHR are aligned with current race-free eGFR reporting practices where appropriate.
• Use confirmatory testing when stakes are high. If a single estimate changes eligibility or major decisions, consider additional markers or repeat testing rather than relying on one race-tuned equation.
• Communicate uncertainty clearly. A calculator is a conversation starter, not a final judge.
• Track outcomes by group. If you change a tool, measure what happens. Equity improvements should be observed, not assumed.

For patients and families

• Ask what equation or calculator is being used. It’s okay to ask, “Does this tool use race?”
• Request explanation in plain language. If a number affects your options, you deserve to know how it was produced.
• If you have kidney disease, ask about eGFR reporting. You can ask whether the lab uses race-neutral eGFR and whether additional testing (like cystatin C) is appropriate for your situation.
• Bring your goals into the room. Algorithms don’t know what matters to you; you do.

Frequently Asked Questions

Is using race in medicine always wrong?

Not automatically. The concern is using race as a biological shortcut instead of measuring the underlying causes directly,
and using race in ways that consistently reduce access or delay care for Black patients.
The best standard is transparency: if race is used, there should be a clear rationale, evidence it improves outcomes, and safeguards against inequity.

Does removing race from equations fix health disparities?

It helps in specific, high-impact decision points (like transplant eligibility thresholds),
but it does not erase disparities created by unequal access, environmental exposures, and structural barriers.
Think of it as removing a thumb from the scale, not rebuilding the whole table.

Why did medicine use race adjustments in the first place?

Many tools were built from historical datasets where race differences appeared in outcomes.
Developers sometimes used race to improve fit, often without fully addressing why differences existed.
Over time, these choices became “standard,” even as the scientific understanding of race and genetics evolved.

Experiences Related to Race-Adjusted Medicine (A 500-Word Reality Check)

The effects of race-adjusted medicine rarely arrive with dramatic music. They show up as quiet nudgesan appointment scheduled later,
a referral that doesn’t happen, a reassuring sentence that’s a little too reassuring. Here are common experiences clinicians and patients describe
(shared here as composite scenarios, not as any one person’s story).

The “Your kidneys look fine” moment. A patient sees “normal” on a lab portal and exhales. Months later, a new clinician reviews the trend,
notices symptoms and rising creatinine, and realizes the earlier estimate may have looked less urgent because of how the equation was applied.
Nobody lied. Nobody intended harm. But the timeline shifted, and in kidney disease, timelines matter.

The referral that never gets printed. In a busy clinic, a clinician uses a threshold: below a certain eGFR, refer to nephrology.
Above it, “watch and wait.” If an estimate is nudged upward, a patient hovers just above the line.
The printer stays silent, the specialist doesn’t meet the patient, and the window for early intervention narrows.
Later, the patient hears, “I wish we’d seen you sooner,” which is the medical version of, “Sorry about the pothole; we’ve been meaning to fix that road.”

The lung test that normalizes the abnormal. A Black patient with shortness of breath is told their spirometry is within the expected range.
They feel dismissed because their body is still sending the same urgent signal: stairs are harder, sleep is disrupted, activity shrinks.
When the interpretation changesor when a clinician looks beyond a single reference labelthe same numbers can take on different meaning:
maybe this is asthma, COPD, or occupational exposure, and maybe treatment should start now, not after “worsening.”

The childbirth conversation that feels pre-decided. A pregnant patient asks about trying for a vaginal birth after a previous C-section.
A calculator prints a probability. The counseling tone changes: “It’s probably not worth it.”
When the patient asks why, race is mentioned as if it were a medical vital sign. The patient leaves feeling like their choices were narrowed
before their preferences even entered the conversation. When newer tools remove race and counseling focuses on individual factors and informed consent,
many patients describe feeling like they’re finally being treated as a person, not a population average.

The trust tax. Perhaps the most repeated experience is emotional, not numerical:
the moment someone realizes the math is different for them. Even if the clinician explains it kindly, the patient may wonder what else is “different.”
Trust, once spent, is expensive to rebuild.

These experiences are why the debate is not just technical. It’s human. The point isn’t to pretend all group differences are imaginary.
The point is to stop using race as the lever that moves the door to care.

Conclusion: Retiring the Checkbox, Keeping the Science

Race-adjusted medicine often began as an attempt to make tools more accurate. But accuracy without equity can become a different kind of error:
one that lands on the same communities again and again.

The future is not “no data” and it’s not “ignore disparities.”
It’s better datameasures that reflect biology, exposure, environment, and accesspaired with transparency and accountability.
When medicine replaces race-based shortcuts with more meaningful variables, it doesn’t lose rigor.
It gains it.