All ETDs from UAB

Advisory Committee Chair

Jeff M Szychowski

Advisory Committee Members

Nengjun Yi

Inmaculada Aban

Alan Thevenet N Tita

Andres Azuero

Document Type


Date of Award


Degree Name by School

Doctor of Philosophy (PhD) School of Public Health


Multi-arm clinical trials comparing several treatments against a common control are more efficient than traditional multiple two-arm trials. Several flexible multi-arm designs which utilize the closed test procedure for strong familywise error rate control have been proposed to allow treatment selection and sample size re-estimation (SSR) at the interim analysis. Because of the interdependence among hypothesis tests for each treatment induced by the closed test procedure, SSR methods for two-arm trials are not directly applicable to multi-arm trials. Further, there is limited literature on methods for SSR in multi-arm trials. In paper 1, we derive SSR procedures based on the conditional power approach in the context of confirmatory two-stage multi-arm trials with normal outcomes for three flexible designs: the inverse normal combination test (INC), Fisher’s combination test (FiC) and the flexible group sequential design (FGS). We conduct extensive simulation studies to evaluate the performance of SSR based on the conditional power approach using both the planned effect size and the estimated interim effect size, considering two treatment selection rules. We find that FGS and INC outperform FiC in terms of power and additional sample size needed when selecting the best treatment; FGS has the highest power while INC has the smallest sample size when keeping all promising treatments. In paper 2, we extend the SSR procedure to binary outcomes and investigate the operating characteristics of SSR in the three designs. The relative performance of the three designs is similar to that in paper 1. In paper 3, we apply the SSR procedure to the data collected from a trial of higher-dose oxytocin at delivery to demonstrate a practical example of a multi-arm study that may have benefited from the SSR methods examined in this analysis.

Included in

Public Health Commons



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.