Institute for Medical Informatics, Statistics and Epidemiology, University of Leipzig, Leipzig, Germany.
PLoS One. 2012;7(6):e38960. doi: 10.1371/journal.pone.0038960. Epub 2012 Jun 15.
Analysis of clinical studies often necessitates multiple graphical representations of the results. Many professional software packages are available for this purpose. Most packages are either only commercially available or hard to use especially if one aims to generate or customize a huge number of similar graphical outputs. We developed a new, freely available software tool called KMWin (Kaplan-Meier for Windows) facilitating Kaplan-Meier survival time analysis. KMWin is based on the statistical software environment R and provides an easy to use graphical interface. Survival time data can be supplied as SPSS (sav), SAS export (xpt) or text file (dat), which is also a common export format of other applications such as Excel. Figures can directly be exported in any graphical file format supported by R.
On the basis of a working example, we demonstrate how to use KMWin and present its main functions. We show how to control the interface, customize the graphical output, and analyse survival time data. A number of comparisons are performed between KMWin and SPSS regarding graphical output, statistical output, data management and development. Although the general functionality of SPSS is larger, KMWin comprises a number of features useful for survival time analysis in clinical trials and other applications. These are for example number of cases and number of cases under risk within the figure or provision of a queue system for repetitive analyses of updated data sets. Moreover, major adjustments of graphical settings can be performed easily on a single window.
We conclude that our tool is well suited and convenient for repetitive analyses of survival time data. It can be used by non-statisticians and provides often used functions as well as functions which are not supplied by standard software packages. The software is routinely applied in several clinical study groups.
分析临床研究结果通常需要多种图形表示。为此，有许多专业的软件包可供选择。这些软件包大多要么是商业软件，要么很难使用，尤其是如果需要生成或定制大量相似的图形输出。我们开发了一种新的、免费的软件工具，称为 KMWin（Windows 下的 Kaplan-Meier），用于进行 Kaplan-Meier 生存时间分析。KMWin 基于统计软件环境 R，并提供了一个易于使用的图形界面。生存时间数据可以作为 SPSS（sav）、SAS 导出（xpt）或文本文件（dat）提供，这也是其他应用程序（如 Excel）的常见导出格式。图形可以直接以 R 支持的任何图形文件格式导出。
基于一个工作示例，我们演示了如何使用 KMWin 并展示了其主要功能。我们展示了如何控制界面、定制图形输出以及分析生存时间数据。对 KMWin 和 SPSS 在图形输出、统计输出、数据管理和开发方面进行了比较。虽然 SPSS 的一般功能更强大，但 KMWin 包含了一些对临床试验和其他应用中的生存时间分析有用的功能。例如，图形中包含病例数和风险病例数，或者提供队列系统以重复分析更新的数据集。此外，还可以在单个窗口中轻松进行主要的图形设置调整。
我们认为，我们的工具非常适合且方便重复分析生存时间数据。它可以由非统计学家使用，并提供常用功能以及标准软件包不提供的功能。该软件在几个临床研究小组中得到了常规应用。
