CoChat
Sign Up
Cited by 110Neurobiology of Language and BilingualismStroke Rehabilitation and RecoveryDementia and Cognitive Impairment Research

Aphasia Recovery: When, How and Who to Treat?

Open access
Doogan, Catherine, Dignam, Jade, Copland, David, Leff, Alex·2018·Current Neurology and Neuroscience Reports
DOI 10.1007/s11910-018-0891-xPMC PMC6209017
Open this paper in CoChat — free accountOpen PDF

Abstract

PURPOSE OF REVIEW: We now know that speech and language therapy (SALT) is effective in the rehabilitation of aphasia; however, there remains much individual variability in the response to interventions. So, what works for whom, when and how? RECENT FINDINGS: This review evaluates the current evidence for the efficacy of predominantly impairment-focused aphasia interventions with respect to optimal dose, intensity, timing and distribution or spacing of treatment. We conclude that sufficient dose of treatment is required to enable clinical gains and that e-therapies are a promising and practical way to achieve this goal. In addition, aphasia can be associated with other cognitive deficits and may lead to secondary effects such as low mood and social isolation. In order to personalise individual treatments to optimise recovery, we need to develop a greater understanding of the interactions between these factors.

AI snapshot

Written from the paper’s full text.

TL;DR

This review argues that speech and language therapy (SALT) is effective for aphasia but that individual response varies, so treatment must be personalized by dose, timing, and patient factors. It highlights sufficient dose with distributed/spaced practice and e-therapies as practical ways to improve recovery, while noting that time post onset does not justify restricting therapy to early months.

Key findings

  • In Brady et al.’s meta-analysis of 57 trials and 3002 subjects, SALT versus no SALT improved functional outcomes with a small standardized mean difference (SMD) of 0.28; the largest effect was for speech production (SMD 1.28).
  • Dose and intensity are often confounded: Brady’s review reported a high-versus-low intensity SMD of 11.75, but six of eight studies also had higher dose; in a dose-controlled study, 48 h of therapy over 3 versus 8 weeks produced comparable functional/QoL gains, but the distributed group was significantly better on the Boston Naming Test after intervention and at 1-month follow-up.
  • E-therapies can increase dose: StepByStep (~25 h spaced practice) improved naming by 20% on trained items only, with an incremental cost-effectiveness ratio of ~£3000 per QALY; iReadMore (34 h per 4-week block) improved trained-word reading accuracy/speed by 9% (Cohen’s d = 1.38), with tDCS adding 2.6% (d = 0.41); EVA Park averaged 41 h engagement and produced ~15% functional communication improvement.
  • Predictors of response include cognitive/phonological skills explaining about 50% of between-subject variance in naming therapy response, brain factors being more important than demographic/cognitive factors for reading therapy response, and time post onset not being related to treatment response.

Methods

  • Narrative review of current evidence for predominantly impairment-focused aphasia interventions, covering dose, intensity, timing, distribution/spacing, e-therapies, and factors explaining between-subject differences.
  • Synthesizes prior meta-analytic and randomized controlled trial evidence, including Brady et al. (57 trials, 3002 subjects) and selected RCTs, with attention to standardized and simple effect sizes.
  • No new primary data are reported; conclusions are based on cited literature and the authors’ interpretation.

Limitations

  • The paper states that greater understanding is needed of interactions among factors to personalize treatment, that there is no clear evidence for a human equivalent of the animal “window of plasticity,” that dose and intensity are often confounded, and that Cochrane benefits of SALT were not evident at follow-up.
  • The paper states that e-therapies vary in quality, many are narrow/single-modality, some effects are item-specific and do not generalize, and EVA Park does not solve the lack of therapists’ time.
  • Evident from the text: as a narrative review, it does not systematically search or meta-analyze; many cited studies are small, pilot, or have mixed comparators, and the proposed adaptive dose-finding approach would require large-scale trials not yet conducted.

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

Integrity signals

Registry

Crossref · 10.1007/s11910-018-0891-x

Cites

45 of 45 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 39 cited sources checked · 0 supported · 0 flagged

  • Premisecitation

    The Brady et al. Cochrane review concluded that benefits of SALT were not evident at follow-up.

    “the somewhat overlooked conclusion of the Brady et al. Cochrane review that benefits of SALT were not evident at follow-up.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [8••]

      Not checked in the free run.

  • Premisecitation

    Less than half of patients in that trial made an improvement of 3 points or more on the ANELT-A.

    “Less than half of patients made an improvement of 3 points or more on the ANELT-A.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [19•]

      Not checked in the free run.

  • Premisereasoning

    Dose (total amount of therapy) is almost always confounded with intensity (dose/time).

    “The issue of dose (total amount of therapy, usually measured in contact hours) is almost always confounded with intensity (dose/time).”
    Quote in the paperSubstantial claim
  • Interpretationcitation

    The negative findings of the AVERT study on very early mobilisation in stroke highlight the need to be cautious when making inferences from animal recovery data to human rehabilitation.

    “In the motor domain, the negative findings of the recent very large AVERT study on very early mobilisation in stroke [23] highlight the need to be cautious when making inferences from animal recovery data to human rehabilitation.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [23] Bernhardt J, Langhorne P, Lindley RI, Thrift AG, Ellery F, Collier J, et al. Efficacy and safety of very early mobilisation within 24 h of stroke onset (AVERT): a randomised controlled trial. Lancet. 2015;386(9988):46–55. source

      Not checked in the free run.

  • Premisecitation

    Recent findings have highlighted the important role of interactions between domain-general cognitive networks and language-specific systems in therapy-induced aphasia recovery.

    “Other recent findings have highlighted the important role of interactions between domain-general cognitive networks and language specific systems in therapy-induced aphasia recovery [41].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [41] Geranmayeh F, Leech R, Wise RJ. Network dysfunction predicts speech production after left hemisphere stroke. Neurology. 2016;86:1296–305. https://doi.org/10.1212/WNL.0000000000002537 . source

      Not checked in the free run.

  • May combine two claims. 23
  • Classified as probably not a substantial claim. 5
  • The quote only partly states this claim. 1

  • May combine two claims.

    Interpretationreasoning

    In the EVA Park trial, students of speech and language therapy who had received 2 h of training on conversation skills delivered the therapy, so EVA Park does not solve the problem of lack of therapists' time but allows patients and therapists to interact remotely.

    “In this trial, students of speech and language therapy who had received 2 h of training on conversation skills delivered the therapy. Therefore, EVA Park does not solve the problem of lack of therapists’ time, but it does mean that patients and therapists can effectively interact remotely.”
    Quote in the paperSubstantial claim

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

    Premisecitation

    Early rehabilitation is recommended in international clinical guidelines based on the notion of a 'window of plasticity' opened by recent ischemia, demonstrated in animal motor recovery experiments.

    “Regarding timing of therapy in relation to stroke onset, early rehabilitation is recommended in a number of clinical guidelines internationally. This is based on the notion of a ‘window of plasticity’ that is ‘opened’ by recent ischemia, which has been demonstrated in animal experiments in the sphere of motor recovery [22].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [22] Zeiler SR, Hubbard R, Gibson EM, Zheng T, Ng K, O'Brien R, et al. Paradoxical motor recovery from a first stroke after induction of a second stroke: reopening a postischemic sensitive period. Neurorehab Neural Re. 2016;30(8):794–800. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    In the Breitenstein et al. trial, the intervention group did significantly better than control on ANELT-A (p=0.0004) with a moderate standardised effect size (0.58).

    “The intervention group did significantly better than the control group (p = 0.0004) with a moderate standardised effect size (0.58)”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [19•]

      Not checked in the free run.

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

    Interpretationreasoning

    EVA Park is a broad therapy, more in keeping with what face-to-face SALT might deliver.

    “This is clearly a ‘broad’ therapy, more in keeping with what face-to-face SALTs might deliver.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    SALT is a complex intervention with many interacting components, many practices required by those delivering or receiving the intervention, and a number and variety of outcomes.

    “This means that it has (i) many interacting components, (ii) many practices required by those delivering or receiving the intervention and (iii) a number and variety of outcomes [12].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [12] MRC. Developing and evaluating complex interventions: new guidance. London: Medical Research Council; 2008.

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Bhogal's meta-analysis is entitled 'Intensity of aphasia therapy, impact on recovery' but its high-intensity studies are also high-dose studies (average 8.8 h/week, 98 h vs 2 h/week, 44 h).

    “It is entitled, ‘Intensity of aphasia therapy, impact on recovery’ [13], yet in table 3 the authors clearly make the point that high-intensity studies (with an average 8.8 h of therapy per week, compared to 2 h for the low intensity studies) are also the high dose studies (98 vs. 44 h).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [13] Bhogal SK, Teasell R, Speechley M. Intensity of aphasia therapy, impact on recovery. Stroke. 2003;34(4):987–93. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Of the two studies matched for dose, one compared different treatments in high and low intensity arms, and the other found no significant difference in the primary outcome for high versus low intensity treatment in the subacute phase.

    “Of the two studies that were matched for dose, one of them compared different treatments in high and low intensity arms [14] and the other found no significant difference in the primary outcome measure for high versus low intensity treatment given in the subacute phase.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [14] Pulvermüller F, Neininger B, Elbert T, Mohr B, Rockstroh B, Koebbel P, et al. Constraint-induced therapy of chronic aphasia after stroke. Stroke. 2001;32(7):1621–6. source

      Not checked in the free run.

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

    Interpretationcitation

    High intensity may be required for mass-practice, item-based approaches used to improve impairment-based outcomes, but spacing these adaptively, acknowledging evidence for sleep in language learning, may be the way forward.

    “High intensity may well be required for the mass-practice, item-based approaches used to improve performance on impairment-based outcomes, but perhaps spacing these in an adaptive way, acknowledging the burgeoning evidence base for the role of sleep in language learning [17], is the way forward.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [17] Kurdziel LBF, Mantua J, Spencer RMC. Novel word learning in older adults: a role for sleep? Brain Lang. 2017;167:106–13. https://doi.org/10.1016/j.bandl.2016.05.010 . source

      Not checked in the free run.

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

    Interpretationreasoning

    Treating associated cognitive impairments and secondary effects of brain injury (low mood and social isolation) should also improve communication and ultimately quality of life.

    “Treating associated cognitive impairments and secondary effects of brain injury (low mood and social isolation) should also improve communication and ultimately quality of life.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    In that trial, the unstandardized effect was less impressive, with an average improvement of 2.6 points (+6.5%), and the ANELT-A minimally important clinical difference is five points.

    “however, the unstandardized effect was less impressive with an average improvement of 2.6 points (+ 6.5%). The ANELT-A scale has a minimally important clinical difference value of five points.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [19•]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    StepByStep computer-based multi-modal therapy in an RCT in chronic stroke patients (n=34) involved 20 min, three days a week for five months (~25 h spaced practice) and improved naming ability by 20% on average.

    “In an RCT in chronic stroke patients (n = 34), the intervention group practiced for 20 min, three days a week for five months (~ 25 h of spaced practice) and improved their naming ability by 20% on average [29].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [29] Palmer R, Enderby P, Cooper C, Latimer N, Julious S, Paterson G, et al. Computer therapy compared with usual care for people with long-standing aphasia poststroke: a pilot randomized controlled trial. Stroke. 2012;43(7):1904–11. https://doi.org/10.1161/STROKEAHA.112.650671 . source

      Not checked in the free run.

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

    Premisecitation

    Basso's early review found no role for age, sex or handedness and concluded that the factors that really influence outcome are initial severity of aphasia and rehabilitation.

    “As Anna Basso remarked in an early review which focused on demographic variables and found no role for age, sex or handedness: ‘The factors that really do influence outcome are initial severity of aphasia (which is inextricably associated with the extent and the location of the lesion) and rehabilitation’ [35].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [35] Basso A. Prognostic factors in aphasia. Aphasiology. 1992;6(4):337–48. https://doi.org/10.1080/02687039208248605 . source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Brady's Cochrane review reports a very large effect for high versus low intensity (SMD 11.75), but six of eight studies in that comparison had higher dosage and intensity versus low dosage and intensity.

    “Brady’s Cochrane review comes out very strongly for high versus low intensity (a staggering SMD of 11.75) but again six out of eight studies used for this intensity comparison had higher dosage and intensity versus low dosage and intensity.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [8••]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    In a wait-list controlled cross-over study of EVA Park in 20 patients, over 5 weeks they logged an average of 41 h of engagement, and the intervention produced significant gains in a functional communication measure that persisted for at least 5 weeks after therapy.

    “Twenty patients took part in a wait-list controlled, cross-over study. Over 5 weeks, they logged an average of 41 h of engagement. The intervention produced significant gains in a functional communication measure that persisted for at least 5 weeks after the therapy finished.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [34] Marshall J, Booth T, Devane N, Galliers J, Greenwood H, Hilari K, et al. Evaluating the benefits of aphasia intervention delivered in virtual reality: results of a quasi-randomised study. PLoS One. 2016;11(8):e0160381. https://doi.org/10.1371/journal.pone.0160381 . source

      Not checked in the free run.

  • May combine two claims. The quote only partly states this claim.

    Premisecitation

    In a double-blind, placebo-controlled, cross-over RCT in chronic patients with a clinically significant auditory perceptual deficit (n=20), an off-the-shelf phonological e-therapy paired with donepezil produced a small but significant generalised improvement in speech perception, with more severely affected patients benefitting more.

    “There was a small but significant generalised improvement in speech perception, with more severely affected patients benefitting more.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [31] Woodhead ZVJ, Crinion J, Teki S, Penny W, Price CJ, Leff AP. Auditory training changes temporal lobe connectivity in ‘Wernicke’s aphasia’: a randomised trial. Journal of Neurology Neurosurgery and Psychiatry. 2017;88(7):586–94. source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Aguilar and colleagues found that brain factors were more important than demographic and cognitive factors when predicting individual aphasic patients' responses to a reading therapy, although the best predictive model included both sets of factors.

    “Aguilar and colleagues recently found that brain factors were actually more important than demographic and cognitive factors when predicting individual aphasic patients’ responses to a reading therapy, although the best predictive model included both sets of factors [37••].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [37••]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    In the EVA Park study, effect sizes were not reported but an estimate from the two group average scores suggests an improvement of ~15% in functional communication.

    “Effect sizes were not reported but an estimate from the two group average scores suggests an improvement of ~ 15% in functional communication.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [34] Marshall J, Booth T, Devane N, Galliers J, Greenwood H, Hilari K, et al. Evaluating the benefits of aphasia intervention delivered in virtual reality: results of a quasi-randomised study. PLoS One. 2016;11(8):e0160381. https://doi.org/10.1371/journal.pone.0160381 . source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Speech and language therapy (SALT) compared with no SALT improves functional outcomes, with a small standardised mean difference of 0.28.

    “The headline finding is that speech and language therapy (SALT) compared with no SALT improves functional outcomes, but the SMD is small at 0.28.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [8••]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    A high-profile positive group-randomised RCT in chronic patients (n=176) found significant effects following an average of 31 h of therapy (intervention) vs 4.5 h (control) over three weeks.

    “A high-profile, group-randomised, positive RCT in chronic patients (n = 176) found significant effects following an average of 31 h of therapy (intervention group) vs. 4.5 h (control group) over three weeks [19•].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [19•]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    In the StepByStep trial, the intervention cost more than usual care because of therapist time, but despite modest QALY gains (0.14) the incremental cost-effectiveness ratio was favourable (~£3000 per QALY; usual UK cut-off £20,000).

    “The intervention cost more than usual care because of the therapist’s time required to set it up, and despite rather modest gains in quality of life units (0.14 QALYs) the therapy’s incremental cost-effectiveness ratio was favourable (~£3000 per QALY, usual cut off is £20,000 in the UK).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [29] Palmer R, Enderby P, Cooper C, Latimer N, Julious S, Paterson G, et al. Computer therapy compared with usual care for people with long-standing aphasia poststroke: a pilot randomized controlled trial. Stroke. 2012;43(7):1904–11. https://doi.org/10.1161/STROKEAHA.112.650671 . source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    In the same Dignam et al. study, the distributed group was significantly better on the Boston Naming Test both after intervention and at 1-month follow-up.

    “however, on a test of impairment (the Boston Naming Test) the distributed group were significantly better, both after the intervention and at 1-month follow-up [15••].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [15••]

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Hillis et al. demonstrated that the influence of lesion load on aphasia recovery may be influenced by SSRIs; for patients with superior temporal and/or SLF/AF lesions, SSRIs improved naming outcomes independent of depression, initial aphasia severity, or lesion volume.

    “Hillis et al. [38] recently demonstrated that the influence of lesion load on aphasia recovery may be influenced by SSRIs. For patients with superior temporal and/or SLF/AF lesions, the use of SSRIs improved naming outcomes, independent of the presence of depression, initial aphasia severity, or lesion volume.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [38] Hillis AE, Beh YY, Sebastian R, Breining B, Tippett DC, Wright A, et al. Predicting recovery in acute poststroke aphasia. Ann Neurol. 2018;83(3):612–22. https://doi.org/10.1002/ana.25184 . source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    High-dose, massed practice may be necessary to affect change for item-specific, impairment-based measures; however, the spacing and distribution of this training is also important.

    “High-dose, massed practice may indeed be necessary to affect change for item-specific, impairment-based measures; however, the spacing and distribution of this training is also important.”
    Quote in the paperNot a substantial claim
23 of 23

  • The paper relies on this without support.

    Aphasia is a chronic condition for the majority of sufferers.

    “Given that aphasia is a chronic condition for the majority of sufferers, it seems reasonable that precious scientific and clinical resources should be dedicated to reducing its impact”
    “This seems even more pertinent given that aphasia is a chronic condition for many people”

  • The paper relies on this without support.

    Aphasia can be associated with other cognitive deficits and may lead to secondary effects such as low mood and social isolation.

    “In addition, aphasia can be associated with other cognitive deficits and may lead to secondary effects such as low mood and social isolation.”
    “language impairment can often lead to low mood and decreased social functioning.”

  • The paper relies on this without support.

    Stroke has a variable effect across patients and recovery is dependent on several key intrinsic and extrinsic factors.

    “Stroke has a variable effect across patients and recovery is clearly dependent on several key factors that are either intrinsic (e.g. site of lesion, cognition, including ‘capacity to learn’, mood and other personal characteristics), or extrinsic (e.g. time since stroke, therapy dose and quality) to the individual.”

  • The paper relies on this without support.

    Most, but not all, aphasic stroke patients improve over time.

    “Most, but not all, aphasic stroke patients improve over time”
4 of 4

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?

  • Aphasia severity influences inter-individual variability in treated language improvement in post-stroke aphasia.

    “It is well established that inter-individual variability in TLI in post-stroke aphasia is influenced by an array of factors, including, but not limited to, age12–15, education16–19, aphasia severity11,14,15,20,21, baseline cognitive abilities13,22,23, and, specific to bPWA, bilingual language experience and language of treatment (e.g., whether therapy is delivered in L1 or L2)24,25.”

    From Machine Learning Predictions of Recovery in Bilingual Poststroke Aphasia: Aligning Insights With Clinical Evidence (2025) · cites this paper as [20]

    In “Factors That May Explain Between-Subject Differences in Response to Therapy” (page unavailable), the review identifies initial aphasia severity as a factor in overall outcome. It also notes that a naming-therapy measure tends to emphasize gains in more mildly affected patients, but does not clearly establish that severity predicts between-person differences in improvement attributable to treatment. (LLM, conf=0.78)

    Experimental — accuracy not yet measured

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Doogan, C., Dignam, J., Copland, D., & Leff, A.
Affiliations
  • UCL Queen Square Institute of Neurology (GB)
  • The University of Queensland (AU)
Corresponding author
Doogan, Catherine
ORCID
Venue
Current Neurology and Neuroscience Reports
Publisher
Springer Science and Business Media LLC
Published
2018-10-15
Volume
18
Issue
12
Pages
90-90
ISSN
  • 1528-4042 (ISSN-L)
  • 1534-6293
DOI
10.1007/s11910-018-0891-x
PMC
PMC6209017
OpenAlex
W2897099992
Semantic Scholar
208fceffe5c634adafa344351379fdfc1093c044
Cited by
110
References
45
Citations by year
2026: 6 · 2025: 15 · 2024: 18 · 2023: 7 · 2022: 23 · 2021: 23 · 2020: 12 · 2019: 5 · 2018: 1
FWCI
1.37
Type
journal article
Language
en
Open access
Licence
Locations
  • Current Neurology and Neuroscience Reports (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • UCL Discovery (University College London) (submittedVersion)
Funders
University College London
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neurobiology of Language and Bilingualism
  • Health Sciences › Medicine › Rehabilitation › Stroke Rehabilitation and Recovery
  • Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
Keywords
aphasia, aphasia recovery, speech therapy, treatment timing, e-therapy, treatment response variability, cognitive impairment, depressive mood, social isolation
MeSH terms
Stroke Rehabilitation, Aphasia, Cognition Disorders, Humans, Speech Therapy, Stroke, Cognitive Dysfunction

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 2026-10-08.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-08.

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