Felker G Michael, Anstrom Kevin J, Rogers Joseph G
Duke Clinical Research Institute, Durham, NC 27705, USA.
J Card Fail. 2008 Jun;14(5):368-72. doi: 10.1016/j.cardfail.2008.01.009. Epub 2008 May 27.
Although previous studies of mechanical circulatory support devices (MCSDs) have used all-cause mortality as a primary end point, trends in device technology and implementation will increasingly necessitate the use of composite end points for MCSD studies. The ideal composite end point for studies of MCSDs is uncertain.
We reviewed the advantages and disadvantages of traditional composite end points in MCSD studies from both a clinical and statistical perspective. Although composite end points may increase statistical power, they also introduce potential problems in interpretation of data, and in some situations may interpret a given clinical course in ways very different than it would be viewed by patients or providers. A global ranking end point, which ranks various aspects of the clinical course based on a prespecified hierarchical ranking system, may provide many of the advantages of composite end points while avoiding many of the pitfalls. One version of such an end point for use in MCSD studies is proposed.
Use of a global ranking end point for incorporating various components of the clinical course into a single end point retains much of the benefit of composite end points while avoiding many of the limitations associated with the use of composites. Adoption of such end points may facilitate the development of MCSD therapy.
尽管先前关于机械循环支持设备(MCSD)的研究主要将全因死亡率作为主要终点,但设备技术和应用的发展趋势将越来越需要在MCSD研究中使用复合终点。MCSD研究的理想复合终点尚不确定。
我们从临床和统计学角度回顾了MCSD研究中传统复合终点的优缺点。尽管复合终点可能会提高统计效能,但它们也会在数据解读方面带来潜在问题,并且在某些情况下,对特定临床过程的解读方式可能与患者或医疗服务提供者的看法大相径庭。一种基于预先设定的分层排序系统对临床过程的各个方面进行排序的综合排序终点,可能会具备复合终点的许多优点,同时避免许多陷阱。本文提出了一种用于MCSD研究的此类终点的版本。
使用综合排序终点将临床过程的各个组成部分纳入单一终点,保留了复合终点的许多益处,同时避免了与使用复合终点相关的许多局限性。采用此类终点可能会促进MCSD治疗的发展。