脊椎操作療法研究における潜在的クラスター：メタ疫学研究
Casper Nim1, Michelle Frederiksen2, Sasha Aspinall3
1Medical Research Unit, Spine Centre of Southern Denmark, Lillebaelt Hospital, University Hospital of Southern Denmark, Kolding, Denmark; Department of Regional Health Research, University of Southern Denmark, Odense, Denmark; Center for Muscle and Joint Health, Department of Sports Science and Clinical Biomechanics, University of Southern Denmark, Odense, Denmark.
Background:
Reliable reporting and publication practices are essential for trustworthy evidence synthesis and clinical decision-making.
Objective:
We aimed to identify latent classes of randomized controlled trials (RCTs) evaluating spinal manipulative therapy (SMT) based on trial reporting and publication practices, and to examine whether these classes influenced treatment effects.
Design:
Meta-epidemiological study.
Method:
Trials were evaluated on whether they met criteria for trial reporting and publication practices across six domains. Latent class analysis was used to identify trial subgroups. Random-effects meta-regression models assessed whether class membership predicted pooled estimates of treatment effects for pain and disability.
Results:
We included 239 RCTs and identified four classes: Dated (23 %), older trials (mostly pre-2010) with consistently low proportions of criteria met; Non-contributing (30 %), newer trials that inconsistently met the criteria, had small samples, and short follow-ups; SMT-focused (15 %), which reported SMT details and fidelity more consistently but otherwise resembled the Non-contributing class; and Pragmatic (33 %), consisting of larger trials, meeting most criteria, but often underreported SMT-specific and fidelity details. Reporting practices had larger impact on class membership than publication practices. Despite differences class membership was not associated with treatment effect estimates and explained minimal outcome variability (R2 ∼1 %).
Conclusions:
Although trial reporting and publication practices varied substantially across SMT trials, these differences were not associated with differences in treatment effects. The widespread failure to meet key criteria raises concerns about the interpretability and credibility of the SMT evidence base. To strengthen transparency and scientific value, future trials should adhere more rigorously to reporting guidelines.
