CoChat
Sign Up
Cited by 29Substance Abuse Treatment and OutcomesMental Health Treatment and AccessOpioid Use Disorder Treatment

The substance use disorder treatment gap among US college students: Findings from the 2021 National Survey on Drug Use and Health

Open access
Pasman, Emily, Blair, Lisa, Solberg, Marvin A., McCabe, Sean Esteban, Schepis, Ty, Resko, Stella M.·2024·Drug and Alcohol Dependence Reports
DOI 10.1016/j.dadr.2024.100279PMC PMC11403413
Open this paper in CoChat — free account

Abstract

Background: Substance use and substance use disorders (SUD) are prevalent among college students. Information about the gap between substance use treatment need versus treatment receipt can guide efforts to increase service access. This study examined past-year DSM-5 SUD and receipt of treatment among US college students. Methods: Past-year DSM-5 SUD and treatment receipt were estimated among a sample of 6115 college students aged 16 and older and a comparison group of non-students from the 2021 National Survey on Drug Use and Health, weighted to be nationally representative. Among the college student sample, multiple logistic regression was used to identify factors associated with past-year SUD. Bivariate analyses were used to compare socio-demographic and substance use differences between college students who received treatment and those who had an SUD but did not receive treatment. Results: Weighted prevalence of past-year SUD among college students was 21.8 %. Only 4.6 % of students who had an SUD received treatment in any setting. Relative to non-students with SUD, proportionately fewer college students with SUD received treatment. Among college students, age, sex, past-year psychological distress, and past-year substance use were significantly associated with past-year SUD; and receipt of treatment differed significantly by age, insurance type, level of education, and enrollment status. College students who received treatment had greater prevalence of stimulant, opioid, tranquilizer, and poly-SUDs and more severe SUD symptomology than those who did not receive treatment. Conclusion: Additional efforts are needed to engage college students with SUDs in acceptable, evidence-based treatment services.

AI snapshot

Written from the paper’s full text.

TL;DR

In the 2021 National Survey on Drug Use and Health, 21.8% of US college students met criteria for past-year DSM-5 substance use disorder (SUD), but only 4.6% of those with SUD received treatment in any setting. Treatment receipt was lower among college students with SUD than among non-students with SUD (8.1%).

Key findings

  • Weighted past-year SUD prevalence among college students was 21.8%; among students with SUD, 4.6% received treatment in any setting, compared with 8.1% of non-students with SUD (p = .0169).
  • Among college students, past-year SUD was significantly associated with age 21–25 vs 16–20 (aOR = 1.60, 95% CI 1.21–2.12), male sex (aOR = 1.54, 95% CI 1.15–2.05), each 1-point increase in past-year psychological distress (aOR = 1.07, 95% CI 1.05–1.10), and past-year use of alcohol (aOR = 5.49), cannabis (aOR = 4.98), hallucinogens (aOR = 2.27), nonmedical prescription opioids (aOR = 3.80), stimulants (aOR = 2.89), and benzodiazepines (aOR = 4.63).
  • Students who received treatment had greater prevalence of stimulant use disorder (OR = 7.17, 95% CI 2.19–23.48), opioid use disorder (OR = 8.22, 95% CI 1.94–34.72), tranquilizer use disorder (OR = 6.74, 95% CI 2.21–20.60), poly-SUD (3+ SUDs: 39.8% vs 4.3%), and moderate-to-severe SUD (90.6% vs 49.5%).
  • Compared with non-students with SUD, college students with SUD had higher prevalence of alcohol use disorder (69.1% vs 62.8%) and cannabis use disorder (43.2% vs 32.9%), lower prevalence of opioid use disorder (4.6% vs 12.7%) and tranquilizer use disorder (2.7% vs 4.7%), and greater prevalence of moderate-to-severe SUD (49.5% vs 40.2%).

Methods

  • Secondary analysis of the 2021 NSDUH, a cross-sectional survey weighted to be nationally representative; analytic sample included 6,115 college students aged 16 and older and a comparison group of non-students aged 16 and older with SUD.
  • Past-year DSM-5 SUD and treatment receipt were coded dichotomously; treatment settings included specialty, physical/mental healthcare, mutual aid, criminal legal system, virtual/telehealth, and any setting. Psychological distress was measured with the K-6 scale.
  • Weighted Rao-Scott chi-square bivariate analyses compared groups; multiple logistic regression examined factors associated with past-year SUD among college students. A planned logistic regression for treatment receipt was not feasible due to the small number treated (weighted 1.5%, unweighted n = 56 of 1,211 with SUD), so bivariate analyses were used instead.

Limitations

  • Stated: weights were based on the general US population, so the college subpopulation may not be truly representative; Indigenous Peoples were excluded due to low representation; the small number who received treatment limited inferences about predictors of treatment receipt; NSDUH changes during COVID limit combining 2021 with prior waves, and the 2022 updated treatment measure cannot be compared with 2021; the study was constrained by NSDUH measures and did not examine other student- or college-level factors.
  • Evident: the cross-sectional design cannot establish directionality for associations such as part-time enrollment and treatment receipt; treatment-receipt comparisons are bivariate rather than multivariable because of small cell sizes, limiting adjusted inference.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1016/j.dadr.2024.100279

Cites

56 of 56 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 41 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    Among college students, male sex was associated with greater odds of past-year SUD than female sex (aOR=1.54, 95% CI 1.15-2.05).

    “Male students had greater odds of past-year SUD than female students (aOR = 1.54, 95 % CI 1.15–2.05).”
    Quote in the paperNot a substantial claim
  • Findingown data

    College students with SUD had greater prevalence of moderate-to-severe SUD than non-students with SUD (49.5% vs 40.2%, p<.0001).

    “compared to non-students, students had greater prevalence of moderate-to-severe SUDs (49.5 % compared to 40.2 %, p <.0001).”
    Quote in the paperNot a substantial claim
  • Findingown data

    College students with SUD had higher prevalence of cannabis use disorder than non-students with SUD (43.2% vs 32.9%, p<.0001).

    “college students had higher prevalence of alcohol use disorder (69.1 % compared to 62.8 %, p =.0038) and cannabis use disorder (43.2 % compared to 32.9 %, p <.0001), but lower prevalence of opioid use disorder (4.6 % compared to 12.7 %, p <.0001) and tranquilizer use disorder (2.7 % compared to 4.7 %, p =.0305).”
    Quote in the paperNot a substantial claim
  • Findingown data

    Among college students, no differences were found in odds of SUD for those who used heroin, cocaine, or inhalants.

    “No differences were found among those who used heroin, cocaine, or inhalants.”
    Quote in the paperNot a substantial claim
  • Findingown data

    College students with SUD reported greater past-year psychological distress than those without SUD (M=11.93, SD=7.04 vs M=7.80, SD=6.75, p<.0001).

    “Those who met criteria for an SUD also reported greater psychological distress (M = 11.93, SD = 7.04) than those without an SUD (M = 7.80, SD = 6.75, p <.0001).”
    Quote in the paperNot a substantial claim
  • May combine two claims. 11
  • Classified as probably not a substantial claim. 20

  • May combine two claims.

    Findingown data

    Among college students, past-year methamphetamine use was associated with increased odds of SUD (aOR=14.04, 95% CI 1.82-108.00), but this should be interpreted with caution due to small cell sizes and wide confidence intervals.

    “Use of methamphetamine (aOR = 14.04, 95 % CI 1.82–108.00) was also significantly associated with the odds of SUD, however this should be interpreted with caution given the small cell sizes for methamphetamine use and the wide confidence intervals.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Insurance type was associated with treatment receipt: proportionally more students with public insurance or no insurance had received treatment, whereas proportionately fewer with private insurance had received treatment.

    “Regarding insurance type, proportionally more students with public insurance or no insurance had received treatment, whereas proportionately fewer with private insurance had received treatment (p =.0100).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    No statistically significant associations were found between alcohol, cannabis, or hallucinogen use disorders and treatment receipt (ps > .10).

    “No statistically significant associations were found related to alcohol, cannabis, or hallucinogen use disorders (ps >.10)”
    • Also stated in: § Discussion · plain
      “alcohol and cannabis use disorders were not significantly associated with treatment utilization.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    Treatment utilization is greater among people with severe SUDs than among people with mild or moderate SUDs.

    “Treatment utilization is greater among people with opioid use disorder relative to people with alcohol or cannabis use disorder, and among people with severe SUDs relative to people with mild or moderate SUDs (Bohler, 2023).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Bohler, 2023

      Not checked in the free run.

  • May combine two claims.

    Study design and conductown data

    The planned logistic regression to examine factors associated with receipt of treatment among college students with past-year SUD was not feasible due to the small number who received treatment (weighted 1.5%, unweighted n=56 of 1211) and small cell sizes (<5) that violate logistic regression assumptions; such an analysis would require a sample size of approximately 1667.

    “Though we initially planned a logistic regression to examine factors associated with receipt of treatment among college students with a past-year SUD, this was not feasible due in part to the small number of persons who received treatment (weighted 1.5 %, unweighted n = 56 of 1211 who met criteria for SUD) and the small cell sizes (< 5) that violate the assumptions of the logistic regression model”
    Quote in the paperSubstantial claim

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

    Study design and conductcitation

    Past-year psychological distress was measured using the K-6 scale, a validated instrument for measuring recent non-specific psychological distress.

    “Past-year psychological distress was measured using the Kessler Psychological Distress (K-6) scale (Kessler et al., 2002). The K-6 is a validated instrument for measuring recent non-specific psychological distress that is included in the PhenX toolkit (Hamilton et al., 2011) and widely used in substance use research”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Kessler et al., 2002 source

      Not checked in the free run.

    • cites Hamilton et al., 2011 source

      Not checked in the free run.

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

    Study design and conductown data

    Survey procedures with provided sample weights were used for all weighted analyses to adjust for nonresponse, noncoverage, and oversampling and to enhance generalizability of inferences.

    “Survey procedures were used for all weighted analyses with provided sample weights to adjust for nonresponse, noncoverage, and oversampling in some groups and to enhance generalizability of inferences.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Substance use and substance use disorders are prevalent among college students.

    “Substance use and substance use disorders (SUD) are prevalent among college students.”
    • Also stated in: § Introduction · boosted
      “Substance use and related harms are highly prevalent on college campuses (Arterberry et al., 2020, Patrick et al., 2023).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Arterberry et al., 2020 source

      Not checked in the free run.

    • cites Patrick et al., 2023

      Not checked in the free run.

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

    Premisecitation

    A 2009 study estimated that nearly half of undergraduate students at a large public university met criteria for an SUD at some point during their first three years of college.

    “A 2009 study of undergraduate students at a large public university estimated that nearly half of all students met criteria for an SUD at some point during their first three years of college, and only 9 % of those with an SUD sought help (Caldeira et al., 2009).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Caldeira et al., 2009 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation

    Only about 13% of individuals meeting diagnostic criteria for an SUD receive specialty treatment.

    “National estimates suggest that only about 13 % of individuals meeting diagnostic criteria for an SUD receive specialty treatment (Sahker et al., 2024).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Sahker et al., 2024 source

      Not checked in the free run.

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

    Study design and conductown data

    A regression analysis evaluated multicollinearity and all VIFs were <1.5, indicating minimal multicollinearity unlikely to affect the final model.

    “A regression analysis was performed to evaluate for multicollinearity and all VIFs were <1.5, indicating minimal multicollinearity unlikely to affect the results of the final model.”
    Quote in the paperNot a substantial claim

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

    Study design and conductreasoning

    Missingness was minimal, and the NOMCAR statement was included in the logistic regression to reduce potential bias by including cases with missing values in variance estimation.

    “Missingness was minimal in the dataset, as variables were selected when possible that contained previously imputed values. However, the NOMCAR statement was included in the logistic regression to reduce potential for bias by including cases with missing values in the variance estimation for the full model.”
    Quote in the paperNot a substantial claim

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

    Premisecitation

    Natural recovery from SUDs does occur (e.g., "maturing out").

    “Though natural recovery from SUDs does occur (e.g., “maturing out”, Lee and Sher, 2018), outcomes tend to be better with early intervention (Carney and Myers, 2012, Stockings et al., 2016).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Lee and Sher, 2018

      Not checked in the free run.

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

    Premisecitation

    Research suggests sex and gender differences in SUD are narrowing.

    “research suggests sex and gender differences in SUD are narrowing and should continue to be monitored among the college student population (McHugh et al., 2018).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites McHugh et al., 2018 source

      Not checked in the free run.

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

    Premisecitation

    Campus-based social norms campaigns have been shown to reduce alcohol and other drug use among college students, particularly those who overestimate the use of their peers.

    “Campus-based social norms campaigns have been shown to reduce alcohol and other drug use among college students, particularly those who overestimate the use of their peers (Hembroff et al., 2021, Mattern and Neighbors, 2004, Pischke et al., 2021).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Hembroff et al., 2021 source

      Not checked in the free run.

    • cites Mattern and Neighbors, 2004

      Not checked in the free run.

    • cites Pischke et al., 2021 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation

    The SUD treatment gap among college students is larger than the treatment gap observed among non-students in this sample (8.1%), the general US population, and other subpopulations of people with SUDs, including young adults.

    “This discrepancy is larger than the treatment gap observed among non-students in this sample (8.1 %), the general US population, and other subpopulations of people with SUDs, including young adults (Krawczyk et al., 2022, Mekonen et al., 2021, Sahker et al., 2024).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Krawczyk et al., 2022 source

      Not checked in the free run.

    • cites Mekonen et al., 2021 source

      Not checked in the free run.

    • cites Sahker et al., 2024 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    In recent years, social distancing measures to reduce the spread of COVID-19 impacted college student substance use and treatment access.

    “In recent years, social distancing measures to reduce the spread of COVID-19 have impacted college student substance use and treatment access.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Firkey et al., 2022 source

      Not checked in the free run.

    • cites Saloner et al., 2022 source

      Not checked in the free run.

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

    Study design and conductown data

    Indigenous Peoples were excluded from the analytic sample due to low representation.

    “We excluded Indigenous Peoples due to low representation in the analytic sample.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Data from focus groups with college students indicate students prefer that programming acknowledge the reasons students use alcohol and other drugs and oppose content that could be interpreted as fear-mongering.

    “Indeed, data from focus groups with college students indicate students prefer that programming acknowledge the reasons students use alcohol and other drugs and oppose content that could be interpreted as fear-mongering (Andraka-Christou et al., 2019).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Andraka-Christou et al., 2019 source

      Not checked in the free run.

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

    Premisecitation

    Most adolescents with severe SUD symptoms do not age out of symptomatic use over a 32-year period.

    “Further, most adolescents with severe SUD symptoms do not age out of symptomatic use over a 32-year period (McCabe et al., 2022).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites McCabe et al., 2022 source

      Not checked in the free run.

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

    Premisecitation

    In that 2009 study, only 9% of those with an SUD sought help.

    “A 2009 study of undergraduate students at a large public university estimated that nearly half of all students met criteria for an SUD at some point during their first three years of college, and only 9 % of those with an SUD sought help (Caldeira et al., 2009).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Caldeira et al., 2009 source

      Not checked in the free run.

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

    Study design and conductnone given

    Changes to the NSDUH during COVID limit the ability to combine the 2021 data with previous waves.

    “Changes to the NSDUH during COVID limit the ability to combine the 2021 data with previous waves.”
    Quote in the paperNot a substantial claim

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

    Study design and conductnone given

    This study was constrained by NSDUH measures.

    “this study was constrained by NSDUH measures”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    Many people who feel they have resolved a past substance use concern do not consider themselves to be "in recovery".

    “Some research suggests young people with substance use concerns are often interested in non-abstinent treatment goals, and many who feel they have resolved a past substance use concern do not consider themselves to be “in recovery” (Pasman et al., 2023, Smith et al., 2023, Subbaraman and Witbrodt, 2014).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Pasman et al., 2023

      Not checked in the free run.

    • cites Smith et al., 2023 source

      Not checked in the free run.

    • cites Subbaraman and Witbrodt, 2014

      Not checked in the free run.

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

    Premisecitation

    College students often overestimate the substance use of their peers.

    “Previous research has shown college students often overestimate the substance use of their peers (Cox et al., 2019, Dempsey et al., 2016).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Cox et al., 2019 source

      Not checked in the free run.

    • cites Dempsey et al., 2016 source

      Not checked in the free run.

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

    Premisecitation

    Lack of SUD recognition is the number one reason people with an SUD do not receive treatment.

    “Lack of SUD recognition is the number one reason people with an SUD do not receive treatment (SAMHSA, 2023).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites SAMHSA, 2023

      Not checked in the free run.

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

    Study design and conductnone given

    The 2022 NSDUH includes an updated measure of treatment receipt which cannot be compared to the 2021 treatment data.

    “Further, the 2022 NSDUH includes an updated measure of treatment receipt which cannot be compared to the 2021 treatment data.”
    Quote in the paperNot a substantial claim

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

    Premisecitation

    Substance use and SUDs tend to increase in adolescence and peak in young adulthood.

    “Our findings are in line with a large body of research showing substance use and SUDs tend to increase in adolescence and peak in young adulthood (McCabe et al., 2019, Patrick et al., 2023)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites McCabe et al., 2019 source

      Not checked in the free run.

    • cites Patrick et al., 2023

      Not checked in the free run.

28 of 28

  • The paper relies on this without support.

    Current and national information about past-year treatment need and utilization among US college students is limited.

    “However, current and national information about past-year treatment need and utilization among US college students is limited.”
1 of 1

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?

  • SUD symptom severity, along with poly-SUDs and mental illness severity, is a significant factor for treatment receipt.

    “These studies suggest that SUD symptom severity, along with poly-SUDs and mental illness severity, is a significant factor for treatment receipt.”

    From Treatment Use Among U.S. Adults with a Substance Use Disorder: Associations with Symptom Severity, Problem Self-Perception, Comorbid Mental Illness, and Mental Health Treatment (2025) · cites this paper as [26]

    The Results (Treatment Receipt; page unavailable) report that poly-SUDs and moderate-to-severe SUDs were more prevalent among college students who received treatment. The copy measures psychological distress and links it to having an SUD, but the cited passages do not establish that mental illness severity was significantly associated with treatment receipt. The reported treatment comparisons are bivariate, rather than evidence that these are independent factors. (LLM, conf=0.86)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Prescription-stimulant comparison overstates its p-value threshold

    Minor

    The results include prescription stimulants among comparisons described as having p < .0001, but Table 2 reports p = .0006 for that comparison.

  • Table 4 labels the untreated group with the total SUD sample size

    Minor

    Table 4 labels the treatment-not-received group as n = 1327, but its three age-category counts total 1269. Adding the 58 treatment recipients gives 1327, indicating that the header uses the combined group size.

    Reported 1327 · recomputed 1269

  • Prescription-opioid comparison cites the wrong p-value

    Minor

    The results give p = .0006 for the prescription-opioid-use comparison, whereas the corresponding Table 2 row gives p = .0032.

3 of 3

Cites

Flagged first, then by citations

2 of the 58 references couldn’t be matched to a known paper.

Research works printed but not linked: 1

  1. Sahker, 2023, Evaluating the substance use disorder treatment gap in the United States, 2016−2019: A population health observational study, Am. J. Addict.

Cited by

Sorted by citations

Related

Authors
Pasman, E., Blair, L., Solberg, M. A., McCabe, S. E., Schepis, T., & Resko, S. M.
Affiliations
  • University of Michigan (US)
  • Wayne State University (US)
  • Texas State University (US)
Corresponding author
Pasman, Emily
ORCID
Venue
Drug and Alcohol Dependence Reports
Publisher
Elsevier BV
Published
2024-09
Volume
12
Pages
100279
ISSN
2772-7246 (ISSN-L)
DOI
10.1016/j.dadr.2024.100279
PMC
PMC11403413
OpenAlex
W4402037402
Semantic Scholar
d0d9f39114970e129ec8de2687ae0f66d0a8183f
Cited by
24
References
58
Citations by year
2026: 18 · 2025: 6
FWCI
7.28
Type
journal article
Language
en
Open access
Licence
Locations
  • Drug and Alcohol Dependence Reports (publishedVersion, cc-by-nc-nd)
  • PubMed Central (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
Funders
  • National Institute on Drug Abuse
  • National Institutes of Health (R01DA031160, R01DA043691)
Topics
  • Health Sciences › Medicine › Epidemiology › Substance Abuse Treatment and Outcomes
  • Social Sciences › Psychology › Social Psychology › Mental Health Treatment and Access
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Opioid Use Disorder Treatment
Keywords
substance use disorders, university students, treatment gap, treatment receipt, treatment access, DSM-5 diagnostic criteria, stimulant use disorder, opioid use disorder, psychological distress, insurance type, educational attainment, enrollment status, National Survey on Drug Use and Health

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 2026-10-07.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-07.

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