文献速递流行病学与生物统计医学教育与科研方法

骨病手法治疗随机对照试验可信度调查:系统综述与Meta分析

Investigating the trustworthiness of randomized controlled trials in osteopathic research: a systematic review with meta-analysis.
PubMed 文献 · J Clin Epidemiol · 2025-04-17

该前瞻性注册系统综述检索7个数据库,纳入2021年1月至2024年6月发表的61项骨病手法治疗随机对照试验,评估偏倚风险与可信度。结果显示74%试验偏倚风险高,仅7%声明作者职业关系潜在利益冲突,仅12%摘要无spin,Meta分析发现两个离群值和五个效应量极大研究,19%数据不足以合并。作者认为骨病随机对照试验的最佳实践依从性需改善。

要点

  • 74%的骨病手法治疗随机对照试验偏倚风险高,仅7%声明作者职业关系潜在利益冲突。
  • 仅12%的摘要无spin,38%存在选择性分析,36%未声明多重检验,12%存在结局替换。
  • Meta分析发现两个离群值和五个效应量极大的研究,19%的研究数据不足以合并。

核心结论

  • 骨病随机对照试验对最佳实践的依从性需要改善,以增强循证决策、减少研究浪费并提高可重复性。
  • 作者认为进一步研究应探索这些发现是否适用于其他小型、资源不足的领域。

研究设计与样本

  • 前瞻性注册的系统综述与Meta分析(PROSPERO-ID: CRD42023457697),检索MEDLINE、EMBASE、CINAHL、AMED、PEDro、ostmed.dr和Chiroindex,检索时间为2021年1月至2024年6月。
  • 纳入61项评估骨病治疗的随机对照试验,样本量中位数45(四分位距30-76),主要研究健康志愿者(29%)。
  • 偏倚风险采用Cochrane工具2评估,可信度采用Cochrane妊娠与分娩筛查工具和REAPPRAISED清单评估,同时评估期刊可信度、摘要中的误导性表述(spin)和结果合理性(通过Meta分析),发现采用描述性合成。

主要发现

  • 74%的试验偏倚风险高,仅7%声明作者职业关系潜在利益冲突。
  • 没有期刊出现在警示名单上,但23%的文章在投稿后2个月内发表。
  • 仅27%可联系到的作者回应了审稿人,7篇摘要(12%)无spin。
  • 方法学问题包括缺失数据处理不佳(31%)、选择性分析(38%)、未声明多重检验(36%)和结局替换(12%)。
  • Meta分析发现两个离群值和五个进一步效应量极大的研究,19%的研究提供的数据不足以合并。

临床怎么用

  • 原文未据此给出推荐。

局限性

  • 局限性包括报告不完整和缺乏经过验证的可信度评估工具。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

OBJECTIVES: To systematically investigate trustworthiness (methodological rigor, transparency, good governance, research integrity, and absence of misconduct) in randomized controlled trials (RCTs) of osteopathic manual therapy. METHODS: This prospectively registered review (PROSPERO-ID: CRD42023457697) searched MEDLINE, EMBASE, CINAHL, AMED, PEDro, ostmed.dr, and Chiroindex for RCTs evaluating osteopathic treatments (January 2021-June 2024). Risk of bias (RoB) was assessed using Cochrane tool 2, while trustworthiness was assessed with the Cochrane Pregnancy and Childbirth Screening Tool and the REAPPRAISED checklist. Journal trustworthiness, misleading representations in abstracts ("spin"), and results plausibility (via meta-analysis) were also assessed. Findings were synthesized descriptively. RESULTS: Sixty-one RCTs were included (median sample size 45, interquartile range (IQR) 30-76), largely studying healthy volunteers (29%). Most had high RoB (74%), and only 7% acknowledged potential conflicts from authors' professional ties. No journals appeared on cautionary lists, although 23% of articles were published within 2 months of submission. Only 27% of contactable authors engaged with reviewers. Seven abstracts (12%) were free of spin. Methodological concerns included poor missing data handling (31%), selective analyses (38%), unacknowledged multiple testing (36%), and outcome switching (12%). Meta-analysis found two outliers and five further with very large effects, while 19% provided inadequate data for pooling. Limitations include incomplete reports and lack of validated trustworthiness assessment tools. CONCLUSION: Adherence to best practices in osteopathic RCTs needs improvement to enhance evidence-based decision-making, reduce research waste, and enhance reproducibility. Further research should explore whether these findings apply to other small, under-resourced fields. PLAIN LANGUAGE SUMMARY: Clinical trials are studies that test if medical treatments work. Doctors and others use these studies to decide how to care for patients or which treatments should be paid for. For clinical trials to be helpful, they need to follow rules to show they can be trusted. For example, researchers can build trust by sharing their plans before they start, reporting all their results honestly and answering questions about their work. In this project, we looked at 61 clinical trials from a 3.5-year period that tested a hands-on treatment called "osteopathic manual therapy." We checked how well these trials followed the rules for trustworthy research. We found that many trials had problems. For example, important research steps were skipped (such as properly registering the study on appropriate online platforms before starting, following the steps described in the registration documents, and correctly examining the collected data). Often, results were also made to look better than they really are, or it was not clearly explained what happened during the study. Only a few researchers answered questions when we asked them. This shows that some osteopathic trials need to do better so people can trust the results. We also suggest ways researchers can improve trustworthiness in the future.

来源:J Clin Epidemiol,查看原文。版权归原作者所有。

引用

  1. Sénéquier A, Draper-Rodi J, Alvarez Bustins G, et al. Investigating the trustworthiness of randomized controlled trials in osteopathic research: a systematic review with meta-analysis. J Clin Epidemiol. 2025. PMID: 40252854. doi:10.1016/j.jclinepi.2025.111788

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。