Value of strategic statistical consultants
Many sponsors are restricting the utilization of strategic statistical consultants to sample size calculations. By doing so, they are limiting the value they receive. Statisticians are best positioned to build a bigger picture in their minds and assess larger impacts of decisions made at various stages of development on ultimate goals. They can do this either by applying logical thinking, or preferably by formal quantitative assessments. Optimal solutions are quantitative by their nature. The following are examples of what can lead to sub-optimal approaches:
- Inefficient study designs. Appropriately utilized novel designs can improve efficiency of development by reducing cost and time, and/or improving probability of success and product value.
- Infrequent assessments of efficacy/futility. The more refined our knowledge about the product is, the better decisions can be made.
- Conducting unnecessary trials. Collecting unnecessary information can delay approval time, and negatively impact sponsors and patients awaiting treatment.
- Overpowering individual trials. The shape of the power curve is such that at some point investing more in increasing sample size no longer improves probability of success. In the included figure most efficient investments would be in the region of arrow #1, and this is the region where sponsors should focus in early development.
Efficiency can further be improved by:
- Making decisions in a broader context. Simultaneous considerations of multiple programs improve overall efficiency, because investing more in one program will obstruct the development of another. In the included figure any investment in the region of arrow #3 should be moved to other programs.
In conclusion:
- Involve your statistical consultants early
- Keep an open mind regarding statistical innovation
- Develop options, and compare their characteristics
Share with friends: