文献速递流行病学与生物统计

检索ClinicalTrials.gov未改变一项系统评价的结论

Searching ClinicalTrials.gov did not change the conclusions of a systematic review.
PubMed 文献 · J Clin Epidemiol · 2017-07-27

该案例研究在开展一项系统评价的同时,于2016年3月9日检索ClinicalTrials.gov筛选合格试验记录,由两名独立评价者筛选,针对31项比较和2个结局比较纳入与不纳入该注册库记录时的结论及证据强度分级。共识别106项试验(仅同行评议文献53项、仅注册库23项、两者均有30项)。结果显示,检索注册库增强了对报告偏倚的怀疑,但未改变结论或证据强度分级。

要点

  • 检索ClinicalTrials.gov共识别106项试验,其中仅注册库来源23项。
  • 仅1项比较因加入注册库结果而合并效应量降低。
  • 2项比较存在选择性结局报告证据,另2项比较疑似发表偏倚。

核心结论

  • 检索ClinicalTrials.gov增强了对报告偏倚的怀疑,但未改变结论或证据强度分级。
  • 需进一步研究以确定在不同主题领域及试验结果注册新规生效后,检索ClinicalTrials.gov对系统评价结论的影响。

研究设计与样本

  • 本案例研究与一项系统评价同步开展。
  • 于2016年3月9日检索ClinicalTrials.gov以识别符合该系统评价纳入标准的试验记录。
  • 由两名独立评价者筛选ClinicalTrials.gov记录。
  • 针对31项比较和2个结局,比较纳入与不纳入ClinicalTrials.gov记录时的结论及证据强度分级。

主要发现

  • 共识别106项试验,其中仅同行评议文献53项、仅ClinicalTrials.gov 23项、两者均有30项。
  • 对于1项比较,加入通过ClinicalTrials.gov识别的结果后合并效应量降低。
  • 发现2项比较存在选择性结局报告的证据,另2项比较疑似发表偏倚。
  • 对于所有其他比较,检索ClinicalTrials.gov未改变2个结局的结论或证据强度分级。

局限性

  • 需进一步研究以确定在不同主题领域及试验结果注册新规生效后,检索ClinicalTrials.gov对系统评价结论的影响。

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: We assessed the effect of searching ClinicalTrials.gov on the conclusions of a systematic review. STUDY DESIGN AND SETTING: We conducted this case study concurrently with a systematic review. We searched ClinicalTrials.gov on March 9, 2016, to identify trial records eligible for inclusion in the review. Two independent reviewers screened ClinicalTrials.gov records. We compared conclusions and strength of evidence grade with and without ClinicalTrials.gov records for 31 comparisons and 2 outcomes. RESULTS: We identified 106 trials (53 in the peer-reviewed literature only, 23 in ClinicalTrials.gov only, and 30 in both sources). For one comparison, the addition of results identified through ClinicalTrials.gov reduced the pooled effect size. We found evidence of selective outcome reporting for two comparisons and suspected publication bias for another two comparisons. For all other comparisons, searching ClinicalTrials.gov did not change conclusions or the strength of evidence grading for the two outcomes. CONCLUSION: Our search of ClinicalTrials.gov bolstered suspicions of reporting biases but did not change either the conclusions or the strength of evidence grading. Further research is needed to determine the effect of searching ClinicalTrials.gov on the conclusions of systematic reviews in different topic areas and as the new rules for registration of trial results take effect.

来源:J Clin Epidemiol,查看原文。版权归原作者所有。

引用

  1. Wilson LM, Sharma R, Dy SM, et al. Searching ClinicalTrials.gov did not change the conclusions of a systematic review. J Clin Epidemiol. 2017. PMID: 28757261. doi:10.1016/j.jclinepi.2017.07.009

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。