We thank Cherny and colleagues for their comments about our article. Our use of the term “arbitrary” to describe the cutoffs utilized in the European Society of Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) was not meant to reflect a criticism of the development of this tool. It was simply an observation that the assessments of value in health care (which in the case of ESMO-MCBS is based on both efficacy and safety) are continuous measures that exist over a spectrum. Typically, attempts to dichotomize a continuous variable are subject to some arbitrary decisions. This is supported by data that show that there is only fair correlation between the assessment of value by ESMO-MCBS and other value frameworks despite the use of the same constructs of clinical benefit (1–3). Additionally, we are familiar with the robust statistical modeling that was performed to develop the ESMO-MCBS (4). However, even with this modeling, arbitrary cutoffs were utilized. For example, the decision to base efficacy assessment on the lower limit of the 95% confidence interval for the hazard ratio may be based on long-standing statistical dogma, but it is a somewhat arbitrary cutoff.

中文翻译：

---

回复

我们感谢切尔尼和同事们对我们文章的评论。我们使用“任意”一词来描述欧洲医学肿瘤学会临床效益量表（ESMO-MCBS）中使用的临界值，并不意味着对该工具的开发提出了批评。只是观察到，医疗保健价值评估（就ESMO-MCBS而言，是基于功效和安全性的）是在整个范围内持续进行的评估。通常，将一个连续变量二等分的尝试会受到一些任意决定的影响。这表明有数据支持，尽管使用了相同的临床获益结构，但通过ESMO-MCBS进行的价值评估与其他价值框架之间只有公平的关联（1-3）。此外，我们熟悉开发ESMO-MCBS（4）所使用的强大统计模型。但是，即使使用此模型，也可以使用任意截止值。例如，将疗效评估基于危险比的95％置信区间的下限进行决策的决定可能基于长期的统计教条，但这在某种程度上是任意的。