PRECIS-2 用于系统评价中 RCT 的回顾性评估
本文评论 PRECIS-2 在系统评价中回顾性评估 RCT 的应用,反驳了安慰剂对照、盲法或单中心试验应直接判为完全解释性而跳过详细评估的观点,并建议增加对照组域。作者认为已发表 RCT 报告因遵循 CONSORT 标准化而更适合回顾性 PRECIS-2 评估,缺失信息域应计为空白。
要点
- 安慰剂对照、盲法或单中心不应直接覆盖 PRECIS-2 详细评估。
- 建议在 PRECIS-2 中增加描述对照组的域。
- 已发表 RCT 报告因 CONSORT 标准化更适合回顾性 PRECIS-2 评估。
核心结论
- PRECIS-2 可用于系统评价中 RCT 的回顾性评估。
- PRECIS-2 工具应扩展增加描述对照组的域。
- 已发表 RCT 报告适合用于回顾性 PRECIS-2 评估。
研究设计与样本
- 本文为针对 PRECIS-2 用于系统评价的评论文章。
主要发现
- 安慰剂在真实诊疗中少被公开处方,但知晓干预通过安慰剂效应起效可能改变部分临床与政策决策,故该特征不总排除决策用途。
- 单中心 RCT 对该中心及类似机构的决策支持价值可能很高,故单中心不应覆盖 PRECIS-2 评估。
- 许多期刊要求 RCT 报告符合 CONSORT,使系统评价中所有 RCT 的可用信息标准化。
- 注册与方案文件的信息未标准化,会削弱 RCT 间及跨系统评价的比较。
- PRECIS-2 中缺失信息的域应计为空白。
关键数值与推荐
- 建议在 PRECIS-2 中增加描述对照组的域。
- PRECIS-2 缺失信息域应计为空白。
- 更广泛使用针对实用性试验的 CONSORT 扩展可能减少缺失数据域。
临床怎么用
- 系统评价中回顾性评估 RCT 时,不应仅因安慰剂对照、盲法或单中心而直接判为完全解释性。
- 应基于已发表 RCT 报告按 PRECIS-2 各域详细评估,缺失信息域计为空白。
- 评估时应考虑增加对照组特征描述以补充 PRECIS-2 评估。
局限性
- 本文为评论文章,未报告原始研究数据或系统评价结果。
- 注册与方案文件信息未标准化,影响 RCT 间比较。
- 已发表 RCT 报告中 PRECIS-2 部分域可能存在缺失信息。
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
BACKGROUND AND OBJECTIVES: A randomized controlled trial (RCT) may be intended either to support real-world decisions on choice between alternative interventions or to help researchers understand mechanisms of action of an intervention. PRECIS-2 is widely used to help investigators match detailed design elements to their main intention for that RCT. PRECIS-2 is increasingly being used retrospectively for assessing RCTs within reviews. In this commentary, we counter arguments that RCTs with a placebo control group, masking/blinding of participants or providers, or conducted in a single center should be retrospectively assessed as completely explanatory, overriding a detailed PRECIS-2 assessment. We also counter arguments that a trial cannot be assessed using only the main peer-reviewed trial report. STUDY DESIGN AND SETTING: This is a commentary on the use of PRECIS-2 for systematic reviews. RESULTS: Although placebos are seldom openly prescribed in real-world care, knowing that an intervention achieves its impact via the placebo effect might change some clinical and policy decisions, which means that this feature does not always preclude decision-making use and so should not override a full PRECIS-2 assessment. A domain describing the comparator should be added to PRECIS-2. Conduct of an RCT in only a single centre should also not override PRECIS-2 as the decision support value of a single-centre RCT could be high for decision makers in that centre and others like it. Many journals require that submitted RCT reports meet CONSORT reporting guidelines, which standardizes the available information for all RCTs in systematic reviews; whereas information from registration and protocol documents is unstandardized and undermines comparison between RCTs and across reviews. Published RCT reports are thus more suitable for retrospective PRECIS-2 assessments, but PRECIS-2 domains with missing information should be scored as blank. Wider use of the CONSORT extension specific to pragmatic trials may reduce domains with missing data. CONCLUSIONS: PRECIS-2 can be used for retrospective assessments of trials in systematic reviews. The PRECIS-2 instrument should be expanded by including a domain describing the control group(s). Published RCT reports are suitable for retrospective PRECIS-2 assessments.
引用
- Zwarenstein M, Thorpe K, Treweek S, et al. PRECIS-2 for retrospective assessment of RCTs in systematic reviews. J Clin Epidemiol. 2020. PMID: 32565215. doi:10.1016/j.jclinepi.2020.06.023
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。