Study design normally covers many different things:
- Choice of study design literally that is a decision about the number of treatment arms, way of comparisons, number of comparisons, etc.
- Study design also includes the protocol development procedure.
- CRF development can also be considered as part of the study design.
- From statistical aspect determination of the sample size is one of the most important study design activity.
- Randomisation also can be considered as study design acttion. The randomisation method is determined during protocol developement, while the calculation is performed before study conduction where it is possible (especially in case of paper CRF) but it might happen in real-time in case of eCRF especially when adjustment for certain factors is made.
Sample size calculation
Sample size calculations can be classified into three main groups.
- Group 1: Classical sample size determination. These tasks are well-documented, easy-to-complete, majority of statistical softwares have good processes to support.
- Group 2: Classical sample size determination for a statistical endpoint which is not yet supported by commercial softwares, so simulation or further programming is required.
- Group 3: Modern sample size determination. These approaches should be used where the design is “modern” as well. Typicial examples are adaptive designs, where even e.g. the number of arms can be modified during the study. The (expected) sample size calculation requires good statistical and programmatic skills.
- Sample size calculation for a phase I first in human study for a German pharma company.
- Sample size determination for a medical review conducted in 13 countries for a UK pharma company.
- Sample size determination for a medical device study for a French company.