Adaptive study design

Annex XI also emphasizes that study designs may be adapted provided that they are adequate for the purpose of classification and labelling and/or risk assessment. This may be taken to include previous studies conducted to other guidelines, including guidelines that have been superseded. 

Three comments are appropriate here. First, consideration of the traditional clinical trial design that has been the focus of attention up until this chapter is extremely worthwhile and instructive It has facilitated the introduction of fundamental design, methodology, and statistical concepts, and it will be an influential player in pharmaceutical drug development for many years to come. Second, the simple observation that the adaptive design may seem different does not in itself make it less valid, less valuable, or less important. Third, statistical approaches that are suitable for adaptive designs are, as yet, less well developed than they are for other study designs. 

Flexibility and adaptability The ability to adapt to changes in research study parameters and timelines is important to understanding the dynamic nature of the drug discovery and development process. A CSO needs to be able to interact effectively with a client to modify a study design or an ongoing research study. 

D. A. Berry, P. Muller, and A. P. Grieve, Adaptive Bayesian Designs for Dose-Ranging Trials. Case Studies in Bayesian Statistics, Vol. V. Springer Verlag, New York, 2002, pp. 99-181. 

Bandit design. An adaptive allocation design based on ideas of sequential decisionmaking with the following features. (1) Results to date are used to determine which treatment a patient will receive. (2) The patient does not necessarily receive whatever treatment is currently believed to be the best because it is recognized that there may be value in studying less-studied treatments. (3) The allocation rule attempts to maximize the (discounted) totality of future patients on the better treatment (including the one about to be treated). 

Figure 11.2 Illustration of the Armitage sequential analysis study design. Patients are paired, and one of each pair receives each alternative treatment. If the patient receiving treatment A does better than the one receiving treatment B (A > B), then the line moves upwards vice versa, if the patient receiving B does better than the one receiving A (B > A), then the line moves downward. If the treatments cannot be distinguished within a pair of patients, then the line moves horizontally. The critical boundaries (broken lines) are computed from prospective measures of a and fj (e.g. p = 0.05 and 80% power, respectively). The technique derives from an engineering control chart and, once again, can be adapted to more sophisticated forms, including limits on the study size for indeterminate results...

Preliminary pilot studies of the proposed study design are often useful to identify and anticipate outcomes and issues which would inform definitive studies and enable protocols to be adapted and would enable the views of the infants carers to be taken into account. This approach would be expected to enhance the co-operation of carers and the quality of the methodology of the subsequent definitive assessment. 

---