I consider Bob O'Neill's important role in promoting regulatory science and, in particular, through the very influential series of lectures he gave in Basel more than 20 years ago, the effect on statistics in European regulation and ultimately on regulatory science in the United States. I provide a simple model of disappointment in drug development. I consider three approaches to improving efficiency in drug development and conclude that there are simple things we could be doing to reduce the cost with which information is obtained in drug development.
- Analysis of covariance
- Flexible designs
- Patient-by-treatment interaction
- Responder analysis