Why Early Planning and Biostatistics Are Central to Trial Success

In a recent video interview with Applied Clinical Trials, Marc Buyse, ScD, founder and CEO of IDDI, discussed how growing trial complexity is creating new challenges for data confidence and interpretability, emphasizing that transparency and reproducibility must remain non-negotiable as Bayesian and adaptive designs become more common. He highlighted key operational gaps—including opaque methodologies, unchecked reliance on synthetic controls, and the risks of black-box analytical approaches—while making a strong case for preserving the randomized controlled trial as the gold standard wherever feasible. Buyse also reflected on the lessons of the COVID era, arguing that the clinical trial enterprise needs to become dramatically more pragmatic and cost-efficient, not just statistically sophisticated. He stressed the critical role of early planning and the estimand framework in anticipating and managing trial conduct problems, and closed with an enthusiastic case for generalized pairwise comparisons and the win ratio as methods he believes will fundamentally reshape how trial outcomes are analyzed and interpreted.

Editor's note: This transcript is a lightly edited rendering of the original audio/video content. It may contain errors, informal language, or omissions as spoken in the original recording.

ACT: What role does biostatistics and early data planning play in reducing risk and improving trial outcomes?

Buyse: Early planning and a lot of thinking in the design of a trial is essential. Far too often—and I'm not talking about the pharma industry, because pharma does this very well—but in academic and investigator-initiated trials, people have a good idea and they implement it without thinking carefully about the design and analysis. The more you can plan, not just for the expected but for the unexpected, the better.

So many trials have problems during the conduct—patients who drop out, patients who don't comply with the treatment, patients who cross over because they don't like the treatment they were assigned. All of these things pollute the analysis and the results. Even in some pharmaceutical trials we've been involved with at IDDI, the analysis was really fatally flawed by things that occurred during the conduct, and it wasn't until the end of the trial that the sponsor realized there was a problem. That made the analysis much less convincing than it otherwise would have been.

Statisticians have worked on this a great deal over the last several years. One of the outcomes of that work is the estimand framework—a bit of a strange word, but it really consists of asking a very specific question about what you're trying to estimate and what could come in the way of that estimation. For example, if you have missing data, well, that's a problem. You need to reduce the probability of missing data as much as possible, but that's easier said than done. And whatever you do, there will always be some. So how do you handle it? How do you account for it in the analysis?

That's the purpose of the estimand framework—to define ahead of time, before you even start randomizing patients, everything that can happen, and then define the methods you will use to attenuate the effects of those problems, if not eliminate them completely.