试验结局的显著性(统计学意义)与发表率及发表时间相关
该研究评估了在ClinicalTrials.gov注册的晚期乳腺癌II期与III期临床试验的结果发表情况及相关因素。共纳入352项II期和74项III期试验,其中12.5%和31.1%已发表,46.9%和58.1%的结果可公开获取。在校正资金来源和试验分期后,结局的显著性仍是影响发表时间的显著因素,显著结局的试验发表更快。
要点
- 晚期乳腺癌II期与III期注册试验中,已发表比例分别为12.5%和31.1%。
- 结果可公开获取的比例分别为46.9%和58.1%。
- 在校正资金来源和试验分期后,结局显著性与发表时间显著相关(HR=6.02)。
核心结论
- 在晚期乳腺癌注册试验中,存在结果报告不足、阴性结果未发表或发表延迟的情况;作者认为需要采取进一步措施以紧急应对这一发表偏倚。
研究设计与样本
- 研究分析了2000年10月1日至2012年9月30日在ClinicalTrials.gov注册的晚期乳腺癌II期和III期临床试验,并通过PubMed检索及查阅注册网站所列文献来检索发表情况。
- 共纳入352项II期试验和74项III期试验;主要结局为发表率、结果可公开获取情况和发表时间。
主要发现
- 在352项II期和74项III期试验中,已发表的比例分别为12.5%和31.1%,结果可公开获取的比例分别为46.9%和58.1%。
- 与结局显著的试验相比,结局不显著的试验发表时间延迟(P < 0.001)。
- 在校正资金来源和试验分期后,研究结局的显著性仍是影响发表时间的显著因素(风险比=6.02;95%CI:3.59,10.07;P < 0.001),结局显著的试验发表所需时间短于结局不显著的试验。
关键数值与推荐
- II期试验发表率为12.5%,III期试验发表率为31.1%。
- II期试验结果可公开获取比例为46.9%,III期试验为58.1%。
- 结局显著性影响发表时间的风险比=6.02(95%CI:3.59,10.07;P < 0.001)。
局限性
- 原文未报告局限性
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: This study aims to comprehensively assess the publication of clinical trial results and factors associated with their publication. STUDY DESIGN AND SETTING: Phase II and III trials of advanced breast cancer registered on ClinicalTrials.gov between October 1, 2000, and September 30, 2012, were identified. Publications were searched by using PubMed and reviewing those listed on the registry site. The main outcomes were publication rate, public availability of results, and time to publication. RESULTS: Of 352 phase II and 74 phase III trials, 12.5% and 31.1% were published, whereas 46.9% and 58.1% had publicly available results, respectively. Compared to those with significant results, studies with nonsignificant results had delays in time to publication (P < 0.001). Even after adjusting for funding source and phase type, the significance of study outcomes was a significant factor that affected time to publication (hazard ratio = 6.02; 95% confidence interval: 3.59, 10.07; P < 0.001), with trials with significant outcomes taking less time to publish than those with nonsignificant outcomes. CONCLUSION: Underreporting of results and nonpublication or delays in the publication of negative results were identified in registered trials of advanced breast cancer. Thus, further initiatives appear necessary to urgently address such publication bias.
引用
- Song SY, Koo DH, Jung SY, et al. The significance of the trial outcome was associated with publication rate and time to publication. J Clin Epidemiol. 2017. PMID: 28238789. doi:10.1016/j.jclinepi.2017.02.009
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。