Statistical methodologies for trial design with network meta-analysis
Date
2023-08
Authors
Ye, Fangshu
Major Professor
Advisor
Wang, Chong
O'Connor, Annette M.
Wu, Huaiqing
Yu, Cindy L.
Nordman, Daniel J.
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Altmetrics
Abstract
Network meta-analysis (NMA) is a tool increasingly used in human and animal health to understand the comparative effect of interventions. One of the unique features of network meta-analysis is the ability to generate estimates of comparative efficacy when no direct clinical trial exists, which introduce the possibility to leverage evidence from the existing network of trial to inform the design of a new trial. Although there are methods proposed based on NMA to plan new trials, methodology for specific applications and the potential misuse of NMA in designing new trials are lacking. Our research focused on designing a new trial for two scenarios of comparison of interest: 1) two treatments in the existing network; 2) one treatment is in the existing network and the other one is not in the existing network. In Chapter 2, we propose an NMA-based method that can increase the power for a fixed total sample size or minimize the required total sample size to reach a pre-specified power when designing a new two-arm trial. In addition, we implement our proposed method into the R package OssaNMA and launch an R Shiny app for the convenience of the application. In Chapter 3, we investigate the potential inflation of type I error risk when a new trial is conducted only when, based on a p-value of the comparison in the existing network, a “promising” difference between two treatments is noticed. In Chapter 4, we consider how to design a new trial when the comparison of interest is between a treatment in the existing network and a treatment not in the existing network. All the methods are proposed from formula derivation and validated by simulation studies. Application of our work can provide insightful guidance for researchers to design a resource-saving new trial without misuse of NMA.
Series Number
Journal Issue
Is Version Of
Versions
Series
Academic or Administrative Unit
Type
dissertation