Morita S
Department of Anesthesia, Teikyo University School of Medicine, Ichihara Hospital.
Masui. 1993 May;42(5):733-7.
If one of the monitors indicates that perioperative myocardial ischemia is occurring, is it telling the truth or what is the probability of its telling the truth? This kind of question is imperative in the conduct of safe anesthesia practice. For this reason, we need to know, so called, the predictive value, which is essentially an application of Bayesian estimation. In this paper, fundamental knowledge and pitfalls in the issue of Bayesian estimation in relation to a predictive value are discussed and summarized. Prior probability, identical to the concept of prevalence is joined together with available information to form a posterior probability or a predictive value to make any clinically relevant decisions. The terms, sensitivity and specificity are defined and their relation to the predictive value is discussed as well. The architecture of Bayesian estimation is explained by taking examples from the literature.
如果其中一个监测仪显示围手术期心肌缺血正在发生，它说的是真的吗？或者它说出真实情况的概率是多少？在安全的麻醉实践中，这类问题至关重要。因此，我们需要了解所谓的预测价值，它本质上是贝叶斯估计的一种应用。本文讨论并总结了与预测价值相关的贝叶斯估计问题的基础知识和陷阱。先验概率与患病率的概念相同，它与可用信息结合在一起，形成后验概率或预测价值，以便做出任何临床相关决策。文中定义了敏感性和特异性这两个术语，并讨论了它们与预测价值的关系。通过引用文献中的例子来解释贝叶斯估计的架构。
