Randomization is the cornerstone of good trial design and, as hasbeen shown here, can be implemented using a range of practical and more technical methods, such that arguments against its use on the basis of practicalities are unfounded. There will rarely be a situation in which the simple parallel group design for a randomized trial is inappropriate, but as has been illustrated here, there are often opportunities to make better use of a limited pool of patients. Such opportunities should always be investigated.

0 0

Post a comment