Thrombosis Research Institute, London, UK.
Thrombosis Research Institute, London, UK.
Am Heart J. 2022 Jan;243:110-121. doi: 10.1016/j.ahj.2021.09.003. Epub 2021 Sep 13.
Randomized controlled trials (RCTs) are considered the gold standard for estimating the effectiveness of a treatment. However, in many instances they are impractical to conduct because of time limitations, cost restrictions, or ethical reasons. As a consequence, non-randomized observational studies have an important role in comparative effectiveness and safety research since they can address issues that would not be possible using conventional RCT methodology. Observational studies can be strategically designed to reduce the risk of potential sources of bias by emulating the design principles of an equivalent but ideal randomized trial - the target trial - that would answer the research question of interest. In this article, we review some of the necessary components of observational studies required for valid causal inference within the framework of target trial emulation, so as to avoid common methodological pitfalls of study design. We discuss the assumptions of consistency, time-zero specification, exchangeability and positivity. To illustrate these concepts in a context where existing knowledge is well-established through clinical trials, we evaluate and compare the treatment effects of vitamin K antagonists (VKA) against no VKA (No VKA) on the treatment of atrial fibrillation from two real-world observational studies, namely the GARFIELD-AF and ORBIT-AF registries. Results are compared with those of published RCTs.
随机对照试验（RCT）被认为是评估治疗效果的金标准。然而，由于时间限制、成本限制或伦理原因，在许多情况下进行 RCT 是不切实际的。因此，非随机观察性研究在比较有效性和安全性研究中具有重要作用，因为它们可以解决使用传统 RCT 方法无法解决的问题。观察性研究可以通过模拟等效但理想的随机试验（目标试验）的设计原则来战略性地设计，以降低潜在偏倚源的风险，从而回答感兴趣的研究问题。在本文中，我们回顾了在目标试验模拟框架内进行有效因果推断所需的一些必要的观察性研究组成部分，以避免研究设计中常见的方法学陷阱。我们讨论了一致性、时间零指定、可交换性和阳性假设。为了在临床试验已经确立了明确知识的背景下说明这些概念，我们从 GARFIELD-AF 和 ORBIT-AF 登记处这两项真实世界的观察性研究中评估和比较了维生素 K 拮抗剂（VKA）与无 VKA（No VKA）治疗心房颤动的治疗效果，并将结果与已发表的 RCT 进行比较。
