International Circulation: There maybe some limitations in the way that we use registries right now， or if they are to become a replacement for randomized control trials. Do you think your colleagues will accept these as an RCT?
　　Dr. James: This is a randomized trial， and has key features to avoid selection biases. It is not a classical RCT because we cannot collect thousands of variables. We have to concentrate on the most important ones. It is ideal for looking at mortality， which is a hard endpoint， but non-ideal for soft-endpoints， which need to be defined. It is limited in resources and in the number of variables we can look at. If you want a more sophisticated analysis on softer endpoints， one will have to add that information to the existing registry. We are working towards this. In the registries we have adjudication of events， more monitoring， and if we are deciding a trial that needs to be look specific details that we do no capture today we add a module on those issues and do stricter controlled monitoring and adjudication of those specific details.
　　《国际循环》：目前我们使用的注册数据库可能仍存在一些局限性，是否有可能将来替代随机对照试验（RCT）还有很多争议您认为业界同仁会接受将这种研究视为RCT的观念吗？
　　Dr. Stefan James: 这是一种随机化试验，并具有避免选择性偏倚的关键特征。这不是传统的RCT，因为我们不能收集数以千计的变量，我们必须将视线集中于最重要的一点上。这种研究用来观察硬终点死亡率非常理想，但观察需要定义的软终点则不够理想。其主要受到来源及能够观察的变量的数量的限制。如果要对软终点进行更有效的分析，必须在现有注册数据的基础上加入额外的信息。这也是我们正在努力的方向。在注册数据中我们可以评判事件以及进行更多的监测；如果要进行一项观察特定细节的试验，而今天未捕捉到该事件，那么我们可以加入一个有关该问题的模块，即可进行更严格的对照监测，评判那些特定的细节。