牙髓病学临床试验中置信区间报告的不确定性边界:我们能否对读到的内容有信心?
该研究评估了牙髓病学领域临床试验中置信区间(CI)的报告率及其与发表特征的关系。纳入2016至2020年三本高影响因子牙髓病学期刊的141篇临床试验报告,仅29.1%报告了CI;包含复杂数据集分析的试验报告CI的调整比值比为11.47(95%CI 4.19-31.41,P<.001)。结论认为牙髓病学临床试验中不确定性测量指标的报告情况不理想。
要点
- 牙髓病学临床试验中仅29.1%报告了置信区间。
- 包含复杂数据集分析的试验报告CI的比值比显著更高。
- 临床阅读牙髓病学试验报告时需警惕不确定性测量缺失。
核心结论
- 牙髓病学临床试验报告中以置信区间为代表的不确定性测量指标纳入情况不理想。
研究设计与样本
- 共识别141篇临床试验报告,记录报告感兴趣结局中竞争性干预效果/安全性差异的CI的数量和比例。
- 采用单变量和多变量逻辑回归识别显著预测变量,并寻找年度线性趋势效应。
主要发现
- 141篇临床试验报告中,大多数发表于Journal of Endodontics(90/141,63.8%),其次是International Endodontic Journal(41/141,29.1%)。
- 仅29.1%的报告(41/141)确认报告了CI。
- 包含复杂数据集分析的临床试验报告CI报告的调整比值比为11.47(95%CI 4.19-31.41,P<.001)。
临床怎么用
- 口腔科医生在阅读牙髓病学临床试验报告时,应关注是否报告了置信区间以评估结果的不确定性。
- 对于未报告CI的试验结果,需谨慎解读其效果估计的精确性。
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
INTRODUCTION: The aim of this study was to assess the prevalence of reporting of confidence intervals (CIs) in clinical trials in endodontics and discover any further associations with a range of publication characteristics. METHODS: The electronic contents of the 3 leading endodontic journals with the highest impact factors (International Endodontic Journal, Journal of Endodontics, and Australian Endodontic Journal) were assessed from January 2016 to December 2020. The number and proportion of clinical trials reporting CIs for the difference in effectiveness/safety of competing interventions for the outcome of interest were recorded. Associations with journal, year of publication, study design, use of analyses for more complex sets of data, and others were assessed. Univariable and multivariable logistic regressions were used to identify significant predictor variables. Yearly linear trend effects were also sought. RESULTS: A total of 141 reports of clinical trials were identified. The majority were published in the Journal of Endodontics (90/141, 63.8%) followed by the International Endodontic Journal (41/141, 29.1%). CI reporting was confirmed only for 29.1% of the sample of reports (41/141). There was strong evidence that reports of clinical trials including analyses of more complex sets of data presented 11.47 times higher odds for CI reporting (adjusted odds ratio = 11.47; 95% CI, 4.19-31.41; P < .001). CONCLUSIONS: The inclusion of uncertainty measures as represented by the reporting of CIs is suboptimal within endodontic clinical trial reports.
引用
- Tzanetakis GN, Koletsi D. Uncertainty Bounds in Clinical Trials Published in Endodontic Journals within the Last 5 Years: Are We Confident on What We Read? J Endod. 2021. PMID: 33965453. doi:10.1016/j.joen.2021.04.025
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。