Step down procedure for comparing several treatments with a control based on multivariate normal response

In this study we consider a multiple comparison of several treatments with a control based on multivariate normal response in clinical trials. Specifically we construct a step down procedure by referring to Dunnett and Tamhane (1991). Furthermore we formulate the all-pairs power by using the recursive formula derived by Hayter and Tamhane (1991) and Dunnett, Horn and Vollandt (2001). Finally we compare the step down procedure with the single step procedure proposed by Nakamura and Imada (2005) in terms of some numerical examples regarding the power of the test.     

A note on penalty-free interim analyses of clinical studies with binary responses

In oncology studies with binary responses, it is often desirable to demonstrate the anti-tumor activity of a test drug before the planned study completes. We propose penalty-free interim analysis strategies that do not change the rejection criterion for the test of null hypothesis at the end. The strategies are applied to standard one-arm and two-arm studies. Although the motivating studies are from oncology, the findings in this research note are equally applicable to similar settings. A broad list of references is provide to stimulate further research on the topic.     

On the Performance of the SPRT for Correlation Coefficient: Normal and Mixtures of Normal Populations

The sequential probability ratio test (SPRT) for the correlation coefficient is examined in the normal and non-normal situations. In the latter case, we evaluate the robustness of the normal procedure when sampling from the mixtures of normal distributions. Two models are introduced for this type of nonnormality. It is shown that in the case of moderate nonnormality, the normal procedure is robust while for extreme nonnormality, the normal procedure is not recommended.     

Estimation and Confidence Intervals after Adjusting the Maximum Information

In a comparative clinical trial, if the maximum information is adjusted on the basis of unblinded data, the usual test statistic should be avoided due to possible type I error inflation. An adaptive test can be used as an alternative. The usual point estimate of the treatment effect and the usual confidence interval should also be avoided. In this article, we construct a point estimate and a confidence interval that are motivated by an adaptive test statistic. The estimator is consistent for the treatment effect and the confidence interval asymptotically has correct coverage probability.