Geng Report: Doubtful 'Double-Blind' Design and Unverifiable Data in a Stroke Rehabilitation Study (Zhang Anmin et al., 2013)

Academic fraud report · Geng Detector

Summary

This report assesses a 2013 paper published in the PLA Journal of Preventive Medicine (DOI: 10.13704/j.cnki.jyyx.2013.03.028) titled 'Observation of the Efficacy of Comprehensive Rehabilitation Training on Stroke Patients with Hemiplegia' by Zhang Anmin, Wu Jiejie, and Jiang Li. Overall verdict: highly suspicious (orange). The central issue is a methodological implausibility: the authors claim to have used a 'double-blind' design for a comprehensive rehabilitation intervention that explicitly includes exercise therapy, occupational therapy, speech therapy, traditional physical therapy, acupuncture, and massage. Such hands-on interventions are, by their nature, incompatible with blinding of either the therapist or the patient, and no sham procedures are described. This wording appears to be a template phrase copied from drug-trial methodology rather than a legitimate description. A secondary concern notes that only summary statistics (means ± SD) are provided, with no raw images, trajectories, or blinding-protocol details, so the data's provenance cannot be verified. The earlier '7T MRI' allegation was retracted after review revealed it to be an OCR artifact. Confidence in the methodological finding is high; confidence in fabrication of numerical data is low due to insufficient evidence.

Verdict

Highly suspicious (orange). The paper contains a methodologically nonsensical claim of double-blinding in a hands-on rehabilitation trial. The retracted '7T MRI' allegation is excluded; the remaining concerns are substantive but not by themselves proof of data fabrication.

Key findings

Implausible 'double-blind' claim (severity: red, confirmed). The Methods section states that 'a double-blind method was used' while delivering acupuncture, massage, exercise, occupational, and speech therapy. No sham procedures, placebo devices, or blinding protocols are described. Physical rehabilitation cannot be blinded to either operator or patient without elaborate sham controls.
Suspected template/copy-paste wording. The phrasing reads as a boilerplate term borrowed from pharmacological trial design rather than a genuine methodological statement, suggesting superficial or careless (and possibly deceptive) reporting.
Unverifiable underlying data (severity: yellow, insufficient evidence). Only Table 1 summary statistics (mean ± SD) are presented; no raw scores, imaging, functional trajectories, or assessor-blinding details are supplied, so the origin of the numbers cannot be independently confirmed.
OCR-related false positive (retracted). An earlier flag regarding a '7T' MRI reference was removed after verification showed it stemmed from OCR artifacts in the PDF, not from the authors' claims.

Evidence highlights

Direct quotation from the paper (Methods, p. 1): '*采用双盲法，分别在患者训练前、训练3个个月时运用FMA运动功能评定表…*' — the double-blind assertion is made in the context of administering multiple non-blindable physical modalities.
Intervention list explicitly includes 运动疗法 (exercise therapy), 作业疗法 (occupational therapy), 语言疗法 (speech therapy), 传统物理治疗 (traditional physical therapy), 针灸 (acupuncture), and 按摩 (massage) — none of which are compatible with patient or therapist blinding as described.
DOI: 10.13704/j.cnki.jyyx.2013.03.028; Received date: 2012-11-28; Year: 2013.

Limitations and notes

No image-level (pixel) analysis was possible because the paper provides no original figures; therefore no image-manipulation finding is made.
The double-blinding claim, while implausible, could in principle reflect extremely poor reporting rather than deliberate fraud; clarification from the authors is required.
The summary statistics could be legitimate but cannot be verified without individual patient data.
All findings should be treated as grounds for editorial inquiry, not as a final determination of misconduct.