Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
3463173 | Contemporary Clinical Trials | 2009 | 12 Pages |
This paper suggests the use of the program package CADEMO when designing clinical trials. Using this package a number of frequently incorrect or inefficient applications of statistical methods can be avoided. For instance, by determining the sample size for statistical tests in advance, the type-II-risk is guaranteed to be taken into account and, foremost, only relevant effects are likely to become significant. Although this is also true for some other program packages, CADEMO in addition exclusively analyzes data by sequential testing, specifically by sequential triangular tests according to Schneider [Schneider, B. 1992. An interactive computer program for design and monitoring of sequential clinical trials. In Proceedings of the XVIth international biometric conference (pp. 237–250). Hamilton, New Zealand.]; which on average lead to a lower sample size. Moreover CADEMO serves to determine the sample size for a confidence estimation of an unknown parameter and for the so-called selection procedures, which should be applied instead of using multiple comparisons of means, if the true objective of a study is to select only the best of all possibilities. Finally, CADEMO serves to factually randomize subjects into any more-way designed experiment. For each of the indicated applications an example is given and the respective handling of CADEMO is illustrated.