CoChat
Sign Up
Cited by 8Treatment of Major DepressionAnxiety, Depression, Psychometrics, Treatment, Cognitive ProcessesMental Health Treatment and Access

Clinical factors associated with relapse in depression in a sample of UK primary care patients who have been on long-term antidepressant treatment

Open access
Duffy, Larisa, Lewis, Gemma, Marston, Louise, Kendrick, Tony, Kessler, David, Moore, Michael, Wiles, Nicola, Lewis, Glyn·2024·Psychological Medicine
DOI 10.1017/s0033291723002659
Open this paper in CoChat — free accountOpen PDF

Abstract

Background: This paper investigates whether age of onset of depression, duration of the last episode, number of episodes, and residual symptoms of depression and anxiety are associated with depression relapse in primary care patients who have been on long–term maintenance antidepressant treatment and no longer meet ICD10 criteria for depression. Methods: An observational cohort using data from ANTLER (N = 478), a double-blind placebo-controlled trial. The primary outcome was time to relapse using the retrospective CIS-R. Participants were followed for 12 months. Results: Primary outcome was available for 468 participants. Time to relapse in those with more than five previous episodes of depression was shorter, hazard ratio (HR) 1.84 (95% confidence interval [CI] 1.23–2.75) compared to people with two episodes; HR 1.57 (95% CI 1.01–2.43) after adjustment. The residual symptoms of depression at baseline were also associated with increased relapse: HR 1.05 (95% CI 1.01–1.09) and HR 1.06 (95% CI 1.01–1.12) in the adjusted model. There was evidence of reduced rate of relapse in older age of onset group: HR 0.86 (95% CI 0.78–0.95); HR attenuated after adjustment HR 0.91 (95% CI 0.81–1.02). There was no evidence of an association between duration of the current episode and residual anxiety symptoms with relapse. Conclusions: The number of previous episodes and residual symptoms of depression were associated with increased likelihood of relapse. These factors could inform joint decision making when patients are considering tapering off maintenance antidepressant treatment or considering other treatments to prevent relapse.

AI snapshot

Written from the paper’s full text.

TL;DR

In a secondary analysis of the ANTLER trial, UK primary care patients on long-term antidepressants who had more previous depressive episodes or residual depressive symptoms at baseline had a higher 12-month relapse risk. Age of onset showed weak evidence of association, while duration of current episode and residual anxiety symptoms were not associated with relapse after adjustment.

Key findings

  • More than five previous episodes vs up to two episodes: 49% vs 32% relapsed; HR 1.84 (95% CI 1.23–2.75) adjusted for group allocation, and HR 1.57 (95% CI 1.01–2.43) after full adjustment.
  • Residual depressive symptoms on PHQ-9 as a continuous measure: HR 1.05 (95% CI 1.01–1.09) adjusted for group allocation, and HR 1.06 (95% CI 1.01–1.12) fully adjusted; PHQ-9 score ≥9 had 51% relapse vs 41% for score ≤3.
  • Older age of onset was associated with lower relapse in the group-adjusted model: HR 0.86 (95% CI 0.78–0.95); this attenuated after full adjustment: HR 0.91 (95% CI 0.81–1.02).
  • No evidence of association was found for duration of current episode (over 10 years vs 2–10 years: fully adjusted HR 1.24, 95% CI 0.91–1.68) or residual anxiety symptoms (GAD-7 continuous fully adjusted HR 1.02, 95% CI 0.96–1.08).

Methods

  • Secondary observational cohort analysis of ANTLER, a double-blind placebo-controlled trial; 478 participants recruited from UK primary care, with primary outcome available for 468 participants and survival analysis excluding 10 participants with missing outcome/timing data.
  • Participants were aged 18–74, had at least two episodes of depression, had taken antidepressants for 9 months or more, and did not meet ICD-10 criteria for current depression; relapse was assessed over 12 months using the retrospective CIS-R at 12, 26, 39, and 52 weeks.
  • Cox proportional hazards models examined age of onset, number of episodes, duration of current episode, baseline PHQ-9, and baseline GAD-7 as exposures, with models adjusted for randomized group, clinical factors, and sociodemographic factors; analyses used STATA 14.

Limitations

  • The paper states that the sample represented only 6% of 23,553 patients approached, so representativeness may be limited; subgroup numbers were relatively small, limiting power to detect small differences; baseline history relied on recall and may have reduced precision; and the 2-month tapering period could have caused withdrawal symptoms, though ICD-10 criteria were used to assess relapse.
  • Evident from the text: this was a secondary analysis of an RCT cohort, and the authors did not examine whether clinical factors moderated the effect of discontinuation because interaction tests would have low statistical power; the reported associations are average associations across both trial arms.

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

Integrity signals

Registry

Crossref · 10.1017/s0033291723002659

Cites

28 of 28 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 23 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    In primary care patients on long-term maintenance antidepressant treatment who no longer meet ICD10 criteria for depression, residual symptoms of depression at baseline are associated with increased relapse (unadjusted HR 1.05, 95% CI 1.01-1.09; adjusted HR 1.06, 95% CI 1.01-1.12 per one-unit PHQ-9 change).

    “The residual symptoms of depression at baseline were also associated with increased relapse: HR 1.05 (95% CI 1.01-1.09)and HR 1.06 (95% CI 1.01-1.12) in the adjusted model.”
    • Also stated in: § Discussion · boosted
      “The results of our secondary analysis of the ANTLER trial provide strong evidence that number of previous episodes and residual symptoms of depression at baseline increased the risk of relapse in a sample of primary care patients, after adjustment for baseline characteristics and other clinical factors.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The PHQ-9 and GAD-7 accurately measure residual symptoms of depression and anxiety.
  • Findingown data

    In primary care patients on long-term maintenance antidepressant treatment who no longer meet ICD10 criteria for depression, older age of onset of depression is associated with a reduced hazard of relapse (unadjusted HR 0.86, 95% CI 0.78-0.95; fully adjusted HR 0.91, 95% CI 0.81-1.02).

    “There was evidence of reduced rate of relapse in older age of onset group: HR 0.86 (95% CI 0.78-0.95);HR attenuated after adjustment HR 0.91 (95% CI 0.81-1.02).”
    • Also stated in: § Sample characteristics · boosted
      “Cox regression analysis of the age of onset as a continuous variable produced strong evidence of association with hazard of relapse: HR 0.86 (95% CI 0.78-0.95) in the model adjusted for group allocation; the HR attenuated in the fully adjusted model HR 0.91 (95% CI 0.81-1.02).”
    • Also stated in: § Discussion · hedged
      “There was some evidence of the age of onset being associated with relapse, though we did not find evidence that duration of last episode and residual symptoms of anxiety were associated with relapse after adjustment for baseline characteristics and other clinical factors.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Self-reported history of depression (age of onset, number of episodes, duration of current episode) is accurate.
  • Findingown data

    Longer duration of the current episode (over 10 years) compared with duration between 2 and 10 years was weakly associated with increased risk of relapse (HR 1.17, 95% CI 0.88-1.55 in the model adjusted for group allocation; HR 1.24, 95% CI 0.91-1.68 in the fully adjusted model).

    “There was weak evidence that longer duration of the current episode (over 10 years) compared with duration between 2 and 10 years, was a risk factor of relapse: HR 1.17 (95% CI 0.88-1.55) in the model adjusted for group allocation and HR 1.24 (95% CI 0.91-1.68) in the fully adjusted model.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Self-reported history of depression (age of onset, number of episodes, duration of current episode) is accurate.
  • Findingown data

    In primary care patients on long-term maintenance antidepressant treatment who no longer meet ICD10 criteria for depression, there is no evidence of an association between residual symptoms of anxiety at baseline and relapse.

    “There was no evidence of an association between duration of the current episode and residual anxiety symptoms with relapse.”
    • Also stated in: § Sample characteristics · plain
      “There was no evidence of an association between residual symptoms of anxiety at baseline and relapse.23)”
    • Also stated in: § Discussion · plain
      “we did not find evidence that duration of last episode and residual symptoms of anxiety were associated with relapse after adjustment for baseline characteristics and other clinical factors.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The PHQ-9 and GAD-7 accurately measure residual symptoms of depression and anxiety.
  • Findingown data

    In primary care patients on long-term maintenance antidepressant treatment who no longer meet ICD10 criteria for depression, having more than five previous episodes of depression is associated with a shorter time to relapse compared with having two episodes (unadjusted HR 1.84, 95% CI 1.23-2.75; adjusted HR 1.57, 95% CI 1.01-2.43).

    “Time to relapse in those with more than five previous episodes of depression was shorter, hazard ratio (HR) 1.84 (95% confidence interval [CI] 1.23-2.75)compared to people with two episodes; HR 1.57 (95% CI 1.01-2.43)after adjustment.”
    • Also stated in: § Discussion · boosted
      “The results of our secondary analysis of the ANTLER trial provide strong evidence that number of previous episodes and residual symptoms of depression at baseline increased the risk of relapse in a sample of primary care patients, after adjustment for baseline characteristics and other clinical factors.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Self-reported history of depression (age of onset, number of episodes, duration of current episode) is accurate.
  • May combine two claims. 13
  • Classified as probably not a substantial claim. 17

  • May combine two claims.

    Findingown data

    Compared with age of onset 40-75 years, early age of onset (23-39 years) is associated with increased relapse in the model adjusted for group allocation (HR 1.62, 95% CI 1.13-2.33), but the association attenuates after full adjustment (HR 1.33, 95% CI 0.87-2.02).

    “Compared to those with age of onset between 40 to 75 years, early age of onset was associated with increased relapse: those with age of onset between 23-and 39-years old group HR was 1.62 (95% CI 1.13-2.33)and in the age of onset between 18 to 22 group HR 1.37 (95% CI 0.90-1.97) in the model adjusted for group allocation (Table 4).After adjusting for baseline characteristics and other clinical”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Self-reported history of depression (age of onset, number of episodes, duration of current episode) is accurate.

  • May combine two claims.

    Premisecitation

    Previous evidence comes from studies conducted in specialist mental healthcare services and is unlikely to be generalizable to UK primary care settings.

    “Even without these limitations, the previous evidence comes from studies conducted in specialist mental healthcare services and is unlikely to be generalizable to UK primary care settings where most people with depression seek help (Kendrick et al., 2009).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Kendrick et al., 2009

      Not checked in the free run.

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

    Premisecitation

    There has been a rise in primary care patients staying on antidepressant medication long-term, with 10% of antidepressant users receiving chronic treatment (5+ years).

    “There has been a rise in primary care patients staying on antidepressant medication long-term (Mars et al., 2017;Moore et al., 2009) with 10% of antidepressant users receiving a 'chronic' treatment (5 + years) (McCrea et al., 2016).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Mars et al., 2017

      Not checked in the free run.

    • cites Moore et al., 2009

      Not checked in the free run.

    • cites McCrea et al., 2016

      Not checked in the free run.

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

    Premisecitation

    There is comprehensive evidence on the clinical effectiveness of antidepressants as maintenance treatment.

    “There is comprehensive evidence on the clinical effectiveness of antidepressants as maintenance treatment (Geddes et al., 2003;Glue, Donovan, Kolluri, & Emir, 2010;Johnson et al., 2012;Kaymaz, Van Os, Loonen, & Nolen, 2008), though to manage treatment it might also be useful to know the likelihood of a future relapse in this population.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Geddes et al., 2003

      Not checked in the free run.

    • cites Glue, Donovan, Kolluri, & Emir, 2010

      Not checked in the free run.

    • cites Johnson et al., 2012

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    The analyses were adjusted for randomized allocation.

    “In contrast to other studies, our analyses were adjusted for the randomized allocation.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    Prolonged antidepressant treatment could lead to side effects such as emotional numbness, weight gain, sleep disturbance, and sexual dysfunction.

    “However, prolonged antidepressant treatment could lead to side effects such as emotional numbness (Goodwin, Price, De Bodinat, & Laredo, 2017), weight gain, sleep disturbance, and sexual dysfunction (Fava, Gatti, Belaise, Guidi, & Offidani, 2015;Rosenbaum, Fava, Hoog, Ascroft, & Krebs, 1998) (Fava et al., 2015).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Goodwin, Price, De Bodinat, & Laredo, 2017

      Not checked in the free run.

    • cites Fava, Gatti, Belaise, Guidi, & Offidani, 2015

      Not checked in the free run.

    • cites Rosenbaum, Fava, Hoog, Ascroft, & Krebs, 1998

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The rCIS-R is a reliable and valid measure for assessing relapse in depression.

    “The results of test-retest reliability study of the rCIS-R and its construct validity in relation to a Global Rating Question about worsening mood, participants stopping their study medication and Patient Health Questionnaire (PHQ-9) scores are described in a separate paper.(Duffy et al., 2023).There is strong evidence that rCIS-R is a reliable and valid measure of assessing relapse in depression.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Duffy et al., 2023

      Not checked in the free run.

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

    Premisecitation

    NICE guidelines recommend continued maintenance treatment for up to 2 years for people at increased risk of relapse.

    “The current NICE guidelines (National Institute for Health Care and Excellence, 2022) recognize that some people are at increased risk of relapse and in those cases are recommend continued maintenance treatment for up to 2 years.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites National Institute for Health Care and Excellence, 2022

      Not checked in the free run.

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

    Premisecitation

    Conradi et al. (2008) reported that number of previous episodes were associated with relapse but not duration of the longest episode, nor residual symptoms of depression and anxiety.

    “Conradi et al. (2008) reported that number of previous episodes were associated with relapse but not duration of the longest episode, nor residual symptoms of depression and anxiety.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Conradi et al., 2008

      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

    The ANTLER trial was a double-blind, individually randomized, parallel group-controlled trial.

    “ANTLER was a double blind individually randomized parallel group-controlled trial that was registered with ISRCTN (ISRCTN15969819).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductcitation

    The study had a large sample size, recruited from a primary care setting, with an excellent follow-up rate and a reliable outcome measure.

    “The strengths of our study are the sample size, the sample recruited from a primary care setting, the excellent follow up rate and the use of reliable outcome measure (Duffy et al., 2023).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Duffy et al., 2023

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation

    The sample represents only 6% of patients approached, so representativeness of the sample may be limited because the population recruited into the RCT might differ from the potentially eligible population.

    “Our sample represents only a small proportion (6%) of patients approached (23 553) either by an invitation letter or at consultation, who agreed to take part and proceeded to screening and eligibility checks (Duffy et al., 2021;Lewis et al., 2021).Therefore, representativeness of our sample might be limited because the population recruited into RCT might differ in some ways from the population”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Duffy et al., 2021

      Not checked in the free run.

    • cites Lewis et al., 2021

      Not checked in the free run.

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

    Premisecitation

    NICE acknowledged the low quality of the evidence supporting this recommendation and the lack of research in the primary care population.

    “However, NICE acknowledged the low quality of the evidence supporting this recommendation and the lack of research in the primary care population where the vast majority of depression is treated.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites National Institute for Health Care and Excellence, 2022

      Not checked in the free run.

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

    Premisenone given

    Asking about previous history in this study is similar to how it is done in clinical practice.

    “This method of asking about previous history is similar to that used in clinical practice.”
    Quote in the paperNot a substantial claim

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

    Premisenone given

    Selection bias in a cohort study is less likely to limit the validity of comparisons within the cohort.

    “any selection bias into a cohort study are less likely to limit the validity of comparisons within a cohort.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisenone given

    Residual depressive symptoms and history of depression are the main factors that influence risk of relapse.

    “Residual depressive symptoms and history of depression are the main factors that influence risk of relapse”
    Quote in the paperSubstantial claim

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

    Premisenone given

    Antidepressants are effective at preventing relapse of depression.

    “Antidepressants are effective at preventing relapse of depression”
    Quote in the paperNot a substantial claim

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

    Premisenone given

    Existing evidence predominately comes from studies set in specialist mental healthcare services and is unlikely to be generalizable to UK primary care settings where most people with depression seek help.

    “Existing evidence predominately comes from studies set in specialist mental healthcare services and is unlikely to be generalizable to UK primary care settings where most people with depression seek help”
    Quote in the paperNot a substantial claim

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

    Premisenone given

    This study was one of few to examine clinical factors of relapse in the primary care population.

    “Our study was one of few to examine clinical factors of relapse in the primary care population.”
    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 conductnone given

    The trial had a relatively short tapering period of two months, and some relapses could have been affected by withdrawal symptoms.

    “The trial had a relatively short tapering period of two months, and some relapses could have been affected by withdrawal symptoms.”
    Quote in the paperNot a substantial claim

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

    Interpretationnone given

    If the benefits of antidepressant maintenance followed a relative risk pattern, one would expect a greater absolute risk reduction for individuals who have a higher risk of relapse.

    “For example, if the benefits of antidepressant maintenance followed a relative risk pattern, one would expect a greater absolute risk reduction for individuals who have a higher risk of relapse.”
    Quote in the paperNot a substantial claim

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

    Premisecitation

    Gopinath et al. (2007) found that residual depressive symptoms, an increasing number of prior episodes, and comorbid anxiety or panic symptoms were associated with relapse.

    “Gopinath et al. (2007) in a larger cohort of 386 primary care patients in US (Gopinath et al., 2007) found that residual depressive symptoms, an increasing number of prior episodes, and comorbid anxiety or panic symptoms were associated with relapse.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Gopinath et al., 2007

      Not checked in the free run.

22 of 22

  • The paper relies on this without support.

    Self-reported history of depression (age of onset, number of episodes, duration of current episode) is accurate.

    “At baseline we measured sociodemographic factors (age, gender, ethnicity, education, employment, housing and marital status), alcohol consumption, financial difficulties, undergoing psychological therapies as well as history of depression (age of onset of depression, number of episodes, duration of the current (or index) episode) considered as relevant clinical risk factors of relapse.”

  • The paper relies on this without support.

    The PHQ-9 and GAD-7 accurately measure residual symptoms of depression and anxiety.

    “At each time point, participants also completed the Patient Health Questionnaire (PHQ-9), a nine-item self-administered questionnaire for measuring DSM-IV depressive symptoms.”
    “At the same time points participants also completed GAD-7, a seven-item questionnaire, that measures anxiety.”

  • The paper relies on this without support.

    The proportional hazards assumption holds for the Cox models.

    “We used Cox proportional hazards modeling to investigate the degree to which each clinical factor (age of onset, number of episodes, duration of the current (or index) episode, residual symptoms of depression measured by PHQ-9 and residual symptoms of anxiety measured by GAD-7 at baseline) was associated with time to relapse, adjusting for potential confounders (baseline demographic”
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?

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.

  • The "over 10 years" duration group includes more participants than the table's 11-years-or-more group

    Major

    Table 4 lists 188 participants with a current episode lasting over 10 years. Table 2 lists only 88 participants in its 11-years-or-more category (24 + 22 + 42); 188 instead equals the combined six-to-10-years and 11-years-or-more counts. This calls the duration grouping used for the reported hazard ratio into question.

  • Previous-episode category counts differ between Tables 2 and 4

    Minor

    Table 2 gives 102 participants with two episodes and 209 with more than five; Table 4 gives 103 and 208, respectively. Both tables total 477 participants across the same three episode categories.

  • Female proportion falls rather than rises across residual-depression categories

    Minor

    The text says participants with higher residual depressive symptoms were more likely to be female, but Table 3 reports female proportions of 78%, 69%, and 59% as the PHQ-9 category increases.

3 of 3

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. 2022, Depression in adults: Treatment and management full guideline

Cited by

Sorted by citations

Related

Authors
Duffy, L., Lewis, G., Marston, L., Kendrick, T., Kessler, D., Moore, M., Wiles, N., & Lewis, G.
Affiliations
  • University College London (GB)
  • The Royal Free Hospital (GB)
  • University of Southampton (GB)
  • University of Bristol (GB)
Corresponding author
Duffy, Larisa
ORCID
Venue
Psychological Medicine
Publisher
Cambridge University Press (CUP)
Published
2023-09-27
Volume
54
Issue
5
Pages
951-961
ISSN
  • 0033-2917 (ISSN-L)
  • 1469-8978
DOI
10.1017/s0033291723002659
OpenAlex
W4387077213
Semantic Scholar
1d4d6800b87be3a8886156e600f1eca82b5006e8
Cited by
8
References
26
Citations by year
2025: 5 · 2024: 3
FWCI
1.24
Type
journal article
Language
en
Open access
Licence
Locations
  • Psychological Medicine (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • ePrints Soton (University of Southampton) (submittedVersion)
  • UCL Discovery (University College London) (submittedVersion, cc-by)
  • Explore Bristol Research (submittedVersion, cc-by)
  • Bristol Research (University of Bristol) (submittedVersion, cc-by)
Funders
  • National Institute for Health and Care Research (13/115/48)
  • Department of Health and Social Care
  • University College London
  • University of Bristol
  • University College London Hospitals NHS Foundation Trust
Topics
  • Health Sciences › Medicine › Pharmacology › Treatment of Major Depression
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Anxiety, Depression, Psychometrics, Treatment, Cognitive Processes
  • Social Sciences › Psychology › Social Psychology › Mental Health Treatment and Access
Keywords
depression relapse, primary care patients, age of onset of depression, number of depressive episodes, residual depressive symptoms, relapse prevention
MeSH terms
Antidepressive Agents, Depression, United Kingdom, Humans, Primary Health Care, Recurrence, Retrospective Studies

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 2026-10-10.

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

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