[MUSIC] Today we're going to talk about analyzing trials. Now this includes the role of the analyst, the statistical analysis plan, analysis considerations, and the types of analyses we do. Analysts play an important role in trials. Let's review the timeline of the trial. There's what we say funding, startup, enrollment and follow up, and analysis and there's what we actually mean, give me the money. No, they gave us the money and now we actually have to run the trial up and running. It's all downhill from here. Right, done not so fast, but there's a lot going on behind the scenes. And the analyst plays a big part in this. So when should an analyst be involved? They should be involved at all stages of the trial, not just at the final analysis. This includes the planning stage, during the trial, and at the final close out, and analysis. Now, why is this important? Well, analysts are trained to think about issues such as bias, design, outcome, and interpretation of results. They provide a complimentary perspective to clinicians, coordinators, patient advocates, and others who were involved in the trial. The more involved and invested in the trial the analyst is, the better they may be able to help you by anticipating problems or addressing issues as they arise. It's important to remember that prevention is always better than fancy statistical analysis techniques to deal with problems. As one of my mentors Dr. Thomas Louis said, space age analyses will not rescue stone age data. In addition, it's important to identify issues in real time, which an analyst can play a big part in. For example, it's very hard for a coordinator to respond to a query on data that was collected five years ago. They may not remember what's happened. They may have lost their notes. They may not have the details that you need. So what does an analyst actually do and when do they do it? Well, analysts where a lot of different hats, they developed the statistical analysis plan prior to the initiation of the study, which gives instructions on how to monitor data and analyze it. They can advise you on design aspects and the potential for bias both during the planning stage and during the course of the trial. This could cover issues such as the choice of outcome or changes you might want to make to the enrollment criteria. They're involved in monitoring during the trial and at the final close out stage. This could be things like clinic performance, data entry and timeliness, as well as data quality. One of the most important roles of the statistician and one that probably takes about 99% of their time is data cleaning, which has performed both during the trial and at the final stages of the trial. They often cached issues that cannot be tracked by the data system. And then finally, there's the role we're most familiar with, the actual analysis of the study. And this again occurs both during and at the end of the trial. This could include in-term analyses, looks during the course of the trial and how things are going, watching for emergent patterns, especially safety patterns that may be concerning as well as the final analysis of the data. [MUSIC]
