Biostatistics Special Seminar

Non-inferiority (NI) studies are gaining popularity, especially in cancer and cardiovascular research. NI trials enable investigators to compare a new treatment to an active control (reference treatment) by establishing the former is not worse than the latter by more than a pre-specified NI margin. A loss of efficacy in a new treatment can be compensated by other benefits, such as fewer side effects, lower costs, and/or simpler treatment regimens. A popular approach is to express the NI margin in terms of the efficacy difference between new treatment and the active control. Many statistical methods have been developed in this framework. However, previous research mainly focused on NI studies with a single reference treatment, limiting the scope of application. In this talk, I will discuss a more complex structure in an NI trial, where multiple new treatments are simultaneously compared to multiple reference treatments with or without placebo to establish assay sensitivity. Three testing procedures will be discussed. The first one relies on the assumption that the variances of the treatments are homogeneous, and two extension procedures handle outcomes with heterogeneous variances among all treatment groups. Clinical examples will be provided for illustrative purposes.

Zoom (Link to Follow)
21 May 2021
1:30pm