Jackson Dan, Rhodes Kirsty, Ouwens Mario
Statistical Innovation Group, AstraZeneca, Cambridge, UK.
Res Synth Methods. 2021 May;12(3):333-346. doi: 10.1002/jrsm.1466. Epub 2020 Dec 22.
Methods for indirect comparisons and network meta-analysis use aggregate level data from multiple studies. A very common, and closely related, scenario is where a company has individual patient data (IPD) from its own trial, but only has published aggregate data from a competitor's trial, and an indirect comparison of the treatments evaluated in these two trials is required. Matching-Adjusted Indirect Comparison (MAIC) has been developed for this situation, where we use the available IPD to adjust for between-trial imbalances in the distributions of observed baseline covariates between the two trials. We extend the current MAIC methodology, where we compute the weights that satisfy the conventional method of moments and result in the largest possible effective sample size (ESS). We show that the approach proposed by Zubizarreta in a previous study can be used for this purpose. We derive a new analytical result that shows why this alternative approach provides a larger ESS than a conventional MAIC. We also derive a new formula for the maximum ESS that can be achieved, even when permitting negative weights, when adjusting for one covariate. This can be used as an easily computed new metric that quantifies the difficulty in adjusting for covariates. What is already known: MAIC is an established way to perform population adjustment in the situation where IPD is available from one trial but only aggregate level data is available from another trial, and an indirect comparison is required. However the effective sample size (ESS) can be small after making the adjustment. What is new: We show that an alternative method can result in a larger ESS. We provide new analytical results showing why this is the case. We derive a new descriptive statistic that is based on maximising the ESS that quantifies the difficulties in adjusting for particular covariates. Potential impact for RSM readers outside the authors' field: Reweighting methods for population adjustment are becoming more commonly used and their implications for research synthesis methodology is now considerable. This paper provides important new links between the theoretical literature, and the more applied research synthesis methodology literature, relating to this topic.
间接比较和网络荟萃分析方法使用来自多项研究的汇总水平数据。一种非常常见且密切相关的情况是，公司拥有自身试验的个体患者数据（IPD），但仅拥有竞争对手试验已发表的汇总数据，并且需要对这两项试验中评估的治疗方法进行间接比较。匹配调整间接比较（MAIC）就是针对这种情况开发的，我们使用可用的IPD来调整两项试验之间观察到的基线协变量分布的试验间不平衡。我们扩展了当前的MAIC方法，计算满足传统矩法并导致最大可能有效样本量（ESS）的权重。我们表明，祖比扎雷塔在先前研究中提出的方法可用于此目的。我们得出了一个新的分析结果，表明为什么这种替代方法比传统MAIC提供更大的ESS。我们还推导了一个新公式，用于计算在调整一个协变量时即使允许负权重也能实现的最大ESS。这可以用作一个易于计算的新指标，量化调整协变量的难度。已知信息：MAIC是在一项试验有IPD而另一项试验只有汇总水平数据且需要进行间接比较的情况下进行总体调整的既定方法。然而，调整后的有效样本量（ESS）可能会很小。新内容：我们表明一种替代方法可以导致更大的ESS。我们提供了新的分析结果，说明为何如此。我们推导了一个基于最大化ESS的新描述性统计量，该统计量量化了调整特定协变量的难度。对作者领域之外的RSM读者的潜在影响：总体调整的重新加权方法越来越常用，其对研究综合方法的影响现在相当大。本文提供了与该主题相关的理论文献和更应用的研究综合方法文献之间重要的新联系。
