Commentary: Estimands in cluster trials: thinking carefully about the target of inferenceand the consequences for analysis choice.
Commentary: Estimands in cluster trials: thinking carefully about the target of inferenceand the consequences for analysis choice.
复制标题
DOI：
10.1093/ije/dyac174
复制
发表时间：
2023-02-08
影响因子：
7.7
通讯作者：
中科院分区：
文献类型：
--
作者：
整群随机试验（CRT）非常复杂。1在方案制定阶段，我们必须选择一个适当的随机化单位（可能取决于干预实施单位）和一个适当的分析单位（可能取决于观察单位）。2，3如果随机化的单位与分析的单位不同，我们必须考虑来自同一聚类的多个观测之间的聚类-这是一个很好理解的要求。2分析单位可以是个人或群组，最好是根据统计理由作出选择（虽然实际上可能反映个人的偏好、便利或经验）。[3]卡汉及其同事的论文建议我们，我们还需要选择一个先验的推理单位，这种选择对于选择单位和分析方法都至关重要。[4]我们认为，在指定分析方法之前，需要考虑推断的目标，这一点在迄今为止的聚类试验文献中尚未得到足够的重视。
Cluster randomized trials (CRTs) are complex. 1 At the protocol development stage, we have to select an appropriate unit of randomization (which may depend on the unit of intervention delivery) and an appropriate unit of analysis (which may depend on the unit of observation). 2, 3 If the unit of randomization is different from the unit of analysis, we must account for clustering among multiple observations from the same cluster—a requirement that is well appreciated. 2 The unit of analysis may be either the individual or the cluster, with the choice ideally made on statistical grounds (although in practice it may reflect personal preference, convenience or experience). 3 The paper by Kahan and colleagues advises us that we also need to choose an a priori unit of inference and this choice is critical in selecting both the unit and the method of analysis. 4 We believe that the need to consider the target of inference before specifying the method of analysis has not received adequate attention in the cluster trials literature to date.
登录
查看更多内容
影响因子：
105.7
作者：
Campbell, Marion K.;Piaggio, Gilda;Altman, Douglas G.
通讯作者：
Altman, Douglas G.
影响因子：
5.5
作者：
Atienza, AA;King, AC
通讯作者：
King, AC
影响因子：
2.2
作者：
Jin, Man;Liu, Guanghan
通讯作者：
Liu, Guanghan
影响因子：
2
作者：
Seaman, Shaun;Pavlou, Menelaos;Copas, Andrew
通讯作者：
Copas, Andrew
有影响力的方法报告针对群体随机试验和相关设计。
影响因子：
2.7
作者：
Murray, David M.
通讯作者：
Murray, David M.