CoChat
Sign Up
Cited by 1,731Race, Genetics, and SocietyEthics and Social Impacts of AIArtificial Intelligence in Healthcare and Education

Hidden in Plain Sight — Reconsidering the Use of Race Correction in Clinical Algorithms

Open access
Vyas, Darshali A., Eisenstein, Leo G., Jones, David S.·2020·New England Journal of Medicine
DOI 10.1056/nejmms2004740
Open this paper in CoChat — free accountOpen PDF

Abstract

Individual subscribers can store articles and searches using a feature on the Journal's website (NEJM.org)called "My Account."Each article and search result links to this feature.Users can create personal folders and move articles into them for convenient retrieval later.

Integrity signals

Registry

Crossref · 10.1056/nejmms2004740

Cites

36 of 36 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 48 cited sources checked · 0 supported · 0 flagged

  • Interpretationreasoning
    § Obs te tric s

    Use of a calculator that lowers estimated VBAC success for people of color could worsen cesarean disparities.

    “Use of a calculator that lowers the estimate of VBAC success for people of color could exacerbate these disparities.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Nonwhite women in the United States have higher cesarean-section rates than white women.
    • Rests on an assumption: Successful vaginal deliveries have health benefits, including fewer surgical complications, faster recovery, and fewer complications in later pregnancies.
  • Interpretationreasoning
    § C ardiology

    Preoperative use of STS calculations could steer minority patients deemed higher risk away from surgery.

    “When used preoperatively to assess risk, these calculations could steer minority patients, deemed to be at higher risk, away from surgery.”
    Quote in the paperSubstantial claim
  • Premisecitation
    § C ardiology

    The AHA Get with the Guidelines-Heart Failure Risk Score predicts in-hospital death risk for admitted heart-failure patients.

    “The American Heart Association (AHA) Get with the Guidelines-Heart Failure Risk Score predicts the risk of death in patients admitted to the hospital.9”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [9] source

      Not checked in the free run.

  • Interpretationreasoning
    § A ssessment

    Most race corrections implicitly or explicitly assume that genetic difference tracks reliably with race.

    “Most race corrections implicitly, if not explicitly, operate on the assumption that genetic difference tracks reliably with race.”
    Quote in the paperNot a substantial claim
  • Premisecitation
    § A ssessment

    Studies of human population genetic structure find more variation within racial groups than between them.

    “Studies of the genetic structure of human populations continue to find more variation within racial groups than between them.34,35”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [34] source

      Not checked in the free run.

    • cites [35]

      Not checked in the free run.

  • Classified as probably not a substantial claim. 23
  • May combine two claims. 17

  • Classified as probably not a substantial claim.

    Premisecitation
    § Nephrology

    One explanation given for higher creatinine is that black people release more creatinine at baseline, partly because they are reportedly more muscular.

    “Explanations that have been given for this finding include the notion that black people release more creatinine into their blood at baseline, in part because they are reportedly more muscular.11,25”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [11] source

      Not checked in the free run.

    • cites [25] source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation
    § C ardiology

    The AHA score adds three points to patients identified as nonblack, so black patients are classified as lower risk.

    “It assigns three additional points to any patient identified as “nonblack,” thereby categorizing all black patients as being at lower risk.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [9] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § A ssessment

    Epidemiologists continue to debate how to make causal inferences based on race responsibly.

    “Epidemiologists continue to debate how to responsibly make causal inferences based on race.33”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [33] source

      Not checked in the free run.

  • May combine two claims.

    Interpretationnone given
    § Hidden in Plain Sight -Reconsidering the Use of Race Correction in Clinical Algorithms

    Even though evidence indicates race is not a dependable proxy for genetic difference, the opposite belief has become embedded in medical practice.

    “For example, despite mounting evidence that race is not a reliable proxy for genetic difference, the belief that it is has become embedded, sometimes insidiously, within medical practice.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation
    § C ardiology

    Clinicians are advised to use the AHA risk score to guide cardiology referral and resource allocation.

    “Clinicians are advised to use this risk score to guide decisions about referral to cardiology and allocation of health care resources.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [9] source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisenone given
    § Obs te tric s

    The VBAC algorithm predicts trial-of-labor risk after previous cesarean.

    “The Vaginal Birth after Cesarean (VBAC) algorithm predicts the risk posed by a trial of labor for someone who has previously undergone cesarean section.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationcitation
    § A ssessment

    Such situations are exceedingly unlikely.

    “Such situations, however, are exceedingly unlikely.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § Nephrology

    Black people have higher rates of end-stage kidney disease and death from kidney failure than the overall population.

    “while black people already have higher rates of end-stage kidney disease and death due to kidney failure than the overall population.25”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [25] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Premisenone given
    § Hidden in Plain Sight -Reconsidering the Use of Race Correction in Clinical Algorithms

    Since 2003, knowledge about race and human genetics has grown, but clear guidance on using race in medicine has not appeared.

    “Our understanding of race and human genetics has advanced considerably since 2003, yet these insights have not led to clear guidelines on the use of race in medicine.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given
    § Hidden in Plain Sight -Reconsidering the Use of Race Correction in Clinical Algorithms

    Physicians use these algorithms to tailor risk assessment and guide clinical choices.

    “Physicians use these algorithms to individualize risk assessment and guide clinical decisions.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § Nephrology

    Analyses have cast doubt on the claim that black people release more creatinine at baseline, but race-corrected eGFR remains standard.

    “Analyses have cast doubt on this claim, 26 but the “race-corrected” eGFR remains the standard.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [26] source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation
    § A ssessment

    Some algorithm developers offer no explanation for racial or ethnic differences; others offer rationales that trace to outdated racial science or biased data.

    “Some algorithm developers offer no explanation of why racial or ethnic differences might exist.Others offer rationales, but when these are traced to their origins, they lead to outdated, suspect racial science or to biased data.22,30,31”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [22] source

      Not checked in the free run.

    • cites [30] source

      Not checked in the free run.

    • cites [31] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Cardiac surgery

    Black people have higher rates of maternal mortality.

    “troubling because black people already have higher rates of maternal mortality.30”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [30] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning
    § A ssessment

    It is insufficient to translate a data signal into a race adjustment without determining what race represents in that context.

    “Given this complexity, it is insufficient to translate a data signal into a race adjustment without determining what race might represent in the particular context.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisereasoning
    § A ssessment

    In the discussed cases, researchers used regression on clinical outcome and patient-characteristic datasets to identify factors correlated with outcomes.

    “They examined data sets of clinical outcomes and patient characteristics and then performed regression analyses to identify which patient factors correlated significantly with the relevant outcomes.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given
    § A ssessment

    Researchers decided it was appropriate or essential to adjust for race in their models.

    “Researchers then decided that it was appropriate -even essential -to adjust for race in their model.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation
    § C ardiology

    STS calculators estimate death and complication risk during cardiac surgery.

    “The Society of Thoracic Surgeons produces elaborate calculators to estimate the risk of death and other complications during surgery.10”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [10] source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning
    § Nephrology

    Use of the KDRI may reduce donation from black people and thus contribute to the wait-time disparity.

    “Use of the KDRI may do just that.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation
    § C ardiology

    Changing to any other race or ethnicity does not raise estimated death risk versus white, but changes risk of renal failure, stroke, or prolonged ventilation.

    “Changing to any other race or ethnicity does not increase the estimated risk of death as compared with a white patient, but it does change the risk of renal failure, stroke, or prolonged ventilation.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [10] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationreasoning
    § A ssessment

    If empirical differences between racial groups were due to genetic differences, race adjustment might be justified.

    “If the empirical differences seen between racial groups were actually due to genetic differences, then race adjustment might be justified: different coefficients for different bodies.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisereasoning
    § Nephrology

    eGFR equations estimate kidney function from serum creatinine because direct measurement is cumbersome.

    “Since it is cumbersome to measure kidney function directly, researchers have developed equations that determine the estimated glomerular filtration rate (eGFR) from an accessible measure, the serum creatinine level.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § C ardiology

    STS calculators include race and ethnicity because of observed outcome differences among racial and ethnic groups, though the mechanism is unknown.

    “The calculators include race and ethnicity because of observed differences in surgical outcomes among racial and ethnic groups; the authors acknowledge that the mechanism underlying these differences is not known.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [10] source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning
    § Nephrology

    Race adjustments that yield higher kidney-function estimates for black patients might delay specialist referral or transplantation and worsen outcomes.

    “Conversely, race adjustments that yield higher estimates of kidney function in black patients might delay their referral for specialist care or transplantation and lead to worse outcomes”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation
    § A ssessment

    Because the health care system spent more on average on white patients than black patients, the tool returned higher risk scores for white patients than black patients.

    “Since the health care system has spent more money, on average, on white patients than on black patients, the tool returned higher risk scores for white patients than for black patients.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [36] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning
    § A ssessment

    This STS decision reflects a default assumption in medicine that race adjustment is acceptable without understanding what race represents in context.

    “This decision reflects a default assumption in medicine: it is acceptable to use race adjustment even without understanding what race represents in a given context.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § A ssessment

    Some tools, including eGFR and VBAC, have been challenged, and clinicians have successfully advocated for institutions to remove race adjustment.

    “Some tools, including eGFR and the VBAC calculator, have already been challenged; clinicians have advocated successfully for their institutions to remove the adjustment for race.43,44”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [43]

      Not checked in the free run.

    • cites [44]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning
    § A ssessment

    When clinicians insert race into tools, they risk interpreting racial disparities as immutable facts rather than injustices requiring intervention.

    “However, when clinicians insert race into their tools, they risk interpreting racial disparities as immutable facts rather than as injustices that require intervention.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given
    § A ssessment

    The criminal justice system uses recidivism-prediction tools to guide bond and sentencing decisions.

    “The criminal justice system, for instance, uses recidivism-prediction tools to guide decisions about bond amounts and prison sentences.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given
    § A ssessment

    Clinicians and researchers typically use OMB categories of five races and two ethnicities.

    “Clinicians and medical researchers typically use the categories recommended by the Office of Management and Budget: five races and two ethnicities.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning
    § A ssessment

    Guidelines are silent on how to handle complex mixed-race or mixed-ethnicity cases, indicating their inadequacy.

    “Guidelines are silent on such issues -an indication of their inadequacy.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

    Interpretationreasoning
    § A ssessment

    If underlying data reflect racist social structures, using them in predictive tools cements racism into practice and policy.

    “But if the underlying data reflect racist social structures, then their use in predictive tools cements racism into practice and policy.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning
    § A ssessment

    When predictive tools influence high-stakes decisions, they propagate inequity into the future.

    “When these tools influence high-stakes decisions, whether in the clinic or the courtroom, they propagate inequity into our future.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § A ssessment

    The STS acknowledged concerns that including socioeconomic factors might adjust away disparities but still considered all preoperative factors independently associated with outcomes.

    “The Society of Thoracic Surgeons acknowledged concerns raised by clinicians and policymakers “that inclusion of SES factors in risk models may 'adjust away' disparities in quality of care.”Nonetheless, it proceeded to consider “all preoperative factors that are independently and significantly associated with outcomes””
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [10] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Premisenone given
    § A ssessment

    Understanding of race has advanced considerably in the past two decades.

    “Our understanding of race has advanced considerably in the past two decades.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § A ssessment

    COMPAS does not use race per se but uses many factors correlated with race and returns higher risk scores for black defendants.

    “One tool, COMPAS (Correctional Offender Management Profiling for Alternative Sanctions), while not using race per se, uses many factors that correlate with race and returns higher risk scores for black defendants.40”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [40]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning
    § A ssessment

    A second problem is how racial and ethnic categories are operationalized.

    “A second problem arises from the ways in which racial and ethnic categories are operationalized.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning
    § A ssessment

    These higher risk scores may have led to more referrals for white patients to specialty services, perpetuating spending discrepancies and race bias.

    “These scores may well have led to more referrals for white patients to specialty services, perpetuating both spending discrepancies and race bias in health care.”
    Quote in the paperSubstantial claim
37 of 37

  • The paper relies on this without support.

    Nonwhite women in the United States have higher cesarean-section rates than white women.

    “Nonwhite U.S. women continue to have higher rates of cesarean section than white U.S. women.”

  • The paper relies on this without support.

    Successful vaginal deliveries have health benefits, including fewer surgical complications, faster recovery, and fewer complications in later pregnancies.

    “The health benefits of successful vaginal deliveries are well known, including lower rates of surgical complications, faster recovery time, and fewer complications during subsequent pregnancies.”

  • The paper relies on this without support.

    Minority patients routinely have different health outcomes from white patients.

    “Since minority patients routinely have different health outcomes from white patients, race and ethnicity often correlated with the outcome of interest.”
3 of 3

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Race corrections in clinical algorithms often have the effect of requiring patients of color to be sicker before they qualify for the same level of care as a White patient.

    “Such “corrections” are used in a variety of areas of medicine, from endocrinology and nephrology to oncology and obstetrics, which often have the effect of requiring patients of color to be sicker before they qualify for the same level of care as a White patient.24–26”

    From Challenging Race-Based Medicine Through Historical Education About the Social Construction of Race (2023) · cites this paper as [26]

    The introduction says many race-adjusted algorithms may direct more resources to White patients, and the Nephrology section says race-corrected kidney estimates might delay Black patients’ referral for specialist care or transplantation. Table 1 likewise says a heart-failure score may raise the threshold for clinical resources for Black patients. These passages support the concern behind the paraphrase, but describe potential effects rather than establishing that patients of color often must be sicker to receive the same care. (LLM, conf=0.84)

    Experimental — accuracy not yet measured

  • If a classifier deployed in a clinical pipeline mistakenly underdiagnosed a certain subgroup (for example, Black patients) more than others due to lower prevalence of the disease, this still leads to disadvantage for members of that group and could lead to serious ethical concerns.

    “If a classifier deployed in a clinical pipeline mistakenly underdiagnosed a certain subgroup (for example, Black patients) more than others due to the lower prevalence of the disease, this still leads to disadvantage for members of that group and could lead to serious ethical concerns 8 .”

    From Underdiagnosis bias of artificial intelligence algorithms applied to chest radiographs in under-served patient populations (2021) · cites this paper as [8]

  • The choice of subgroups spanning race, socioeconomic status, sex and age is motivated by the clear history, in both traditional medicine and AI algorithms, of bias for subgroups on these axes.

    “The choice of these subgroups is motivated by the clear history, in both traditional medicine and AI algorithms, of bias for subgroups on these axes 6,8,10,11 .”

    From Underdiagnosis bias of artificial intelligence algorithms applied to chest radiographs in under-served patient populations (2021) · cites this paper as [8]

  • Some observed differences in underdiagnosis have been established in other areas in clinical care, such as underdiagnosis of female patients, Black patients and patients with low socioeconomic status.

    “We note that some of our observed differences in underdiagnosis have been established in other areas in clinical care, such as underdiagnosis of female patients 9,10 , Black patients 5,8,9 and patients with a low socioeconomic status 36 .”

    From Underdiagnosis bias of artificial intelligence algorithms applied to chest radiographs in under-served patient populations (2021) · cites this paper as [8]

  • Algorithmic bias has harmed minoritized communities in housing, banking, and education, and health care is no different.

    “However, algorithmic bias has harmed minoritized communities in housing, banking, and education, and health care is no different.1”

    From Guiding Principles to Address the Impact of Algorithm Bias on Racial and Ethnic Disparities in Health and Health Care (2023) · cites this paper as [1]

Cites

Flagged first, then by citations

Research works printed but not linked: 2

  1. Jones DS, Dorr GM, Pollock A. Symposium: race, pharmaceuticals, and medical technology. J Law Med Ethics 2008;36:439-545.
  2. Kanis JA. Assessment of osteoporosis at the primary health care level. WHO Scientific Group technical report. Sheffield, United Kingdom: World Health Organization Collaborating Centre for Metabolic Bone Diseases, University of Sheffield, 2007.

Cited by

Sorted by citations

Related

Authors
Vyas, D. A., Eisenstein, L. G., & Jones, D. S.
Affiliations
Harvard University (US)
ORCID
Venue
New England Journal of Medicine
Publisher
Massachusetts Medical Society
Published
2020-08-27
Volume
383
Issue
9
Pages
874-882
ISSN
  • 0028-4793 (ISSN-L)
  • 1533-4406
DOI
10.1056/nejmms2004740
OpenAlex
W3035885149
Semantic Scholar
07a84399ee2befea66e7ed0560fc32f0e4e8a070
Cited by
1,731
References
35
Citations by year
2026: 191 · 2025: 262 · 2024: 295 · 2023: 337 · 2022: 295 · 2021: 267 · 2020: 78 · 2013: 1
FWCI
144.54
Type
journal article
Language
en
Open access
Locations
  • New England Journal of Medicine (publishedVersion)
  • PubMed (publishedVersion)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Race, Genetics, and Society
  • Social Sciences › Safety Research › Ethics and Social Impacts of AI
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
Keywords
clinical algorithms, racial bias, algorithmic fairness, health disparities, algorithmic discrimination
MeSH terms
Algorithms, Cardiology, Female, Humans, Male, Nephrology, Obstetrics, Urology, Risk Assessment, Racial Groups, Healthcare Disparities

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-04.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-04.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review