COVID-19防治随机临床试验的方法学质量:一项meta流行病学研究
结果显示82.4%的试验存在高或可能高的偏倚风险,41.8%报告了统计学显著效应,随机化偏倚风险和单中心试验与发现显著效应的几率增加相关。
要点
- 82.4%的COVID-19随机对照试验存在高或可能高的偏倚风险。
- 随机化偏倚风险和单中心试验与发现统计学显著效应的几率增加相关。
- 研究者、资助者和知识使用者在设计、评估和解读RCT时应关注这些因素。
核心结论
- COVID-19试验间偏倚风险差异大,随机化偏倚风险和单中心状态影响显著效应的发现。
研究设计与样本
- 提取试验特征(中心数、资金来源、样本量),使用Cochrane RoB 2.0工具评估偏倚风险。
- 采用logistic回归评估随机化偏倚风险、单中心vs多中心、资金来源和样本量与发现统计学显著效应之间的关联。
主要发现
- 40项(44%)为单中心试验,23项(25.3%)入组患者少于50例。
- 28项(30.8%)接受产业资助,75项(82.4%)存在高或可能高的偏倚风险。
- 随机化偏倚风险和单中心试验与发现统计学显著效应的几率增加相关。
临床怎么用
- 设计、评估和解读COVID-19随机对照试验时应考虑随机化偏倚风险和单中心状态的影响。
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
OBJECTIVE: To describe the characteristics of Covid-19 randomized clinical trials (RCTs) and examine the association between trial characteristics and the likelihood of finding a significant effect. STUDY DESIGN: We conducted a systematic review to identify RCTs (up to October 21, 2020) evaluating drugs or blood products to treat or prevent Covid-19. We extracted trial characteristics (number of centers, funding sources, and sample size) and assessed risk of bias (RoB) using the Cochrane RoB 2.0 tool. We performed logistic regressions to evaluate the association between RoB due to randomization, single vs. multicentre, funding source, and sample size, and finding a statistically significant effect. RESULTS: We included 91 RCTs (n = 46,802); 40 (44%) were single-center, 23 (25.3%) enrolled <50 patients, 28 (30.8%) received industry funding, and 75 (82.4%) had high or probably high RoB. Thirty-eight trials (41.8%) reported a statistically significant effect. RoB due to randomization and being a single-center trial were associated with increased odds of finding a statistically significant effect. CONCLUSIONS: There is high variability in RoB among Covid-19 trials. Researchers, funders, and knowledge-users should be cognizant of the impact of RoB due to randomization and single-center trial status in designing, evaluating, and interpreting the results of RCTs. REGISTRATION: CRD42020192095.
引用
- Honarmand K, Penn J, Agarwal A, et al. Clinical trials in COVID-19 management & prevention: A meta-epidemiological study examining methodological quality. J Clin Epidemiol. 2021. PMID: 34274489. doi:10.1016/j.jclinepi.2021.07.002
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。