瀑布
瀑布模型
绘图(图形)
统计
计算机科学
数学
医学
软件
地理
地图学
程序设计语言
作者
Mo Huang,Cong Chen,Linda Sun
出处
期刊:JCO clinical cancer informatics
[American Society of Clinical Oncology]
日期:2023-09-01
卷期号: (7)
摘要
Waterfall plots have gained popularity as a visualization tool to present antitumor activity of treatments in oncology, especially for phase I and II trials. The typical waterfall plot in oncology is a bar plot with each bar representing the best percent tumor size reduction from baseline for a patient sorted in descending order along the x-axis. As new therapies are routinely developed in combination with standard of care or other investigational treatments, waterfall plot comparison between combination therapy and monotherapy may facilitate development decisions in addition to overall response rate or duration of response. However, waterfall plots are often assessed heuristically in practice with lack of statistical rigor. In this work, we examine the correspondence between the waterfall plot and the empirical cumulative distribution function. We demonstrate how to derive key summary statistics directly from the waterfall plot. Using real examples from published waterfall plots, we show how comparisons of waterfall plots can elucidate clinically meaningful information, such as treatment effect patterns in progression-free survival and overall survival.
科研通智能强力驱动
Strongly Powered by AbleSci AI