Statistical methodology for clinical trials
Lead Research Organisation:
University of Bath
Department Name: Mathematical Sciences
Abstract
The project will concern group sequential and adaptive designs for clinical
trial designs. Group sequential monitoring allows interim analyses of
accumulating data, which can lead to early stopping for a positive or
negative outcome. In adaptive designs, interim data are used to modify
aspects of the trial while still protecting the type I error rate. There
must be sufficient information available at the time of an interim analysis
in order to make sound decisions to modify or terminate a trial. When a
study's primary endpoint is only observed a long time after treatment, there
may be little information on this endpoint at an interim analysis. In this case,
it is desirable to use information from short term endpoints instead. The
project will address the case of a trial with a time-to-event primary endpoint,
such as overall survival or time until disease progression, in which longitudinal
data on a biomarker are to be used at interim analyses. This project will
involve (i) developing joint models for the two endpoints, (ii) derivation of
efficient stopping rules, or more general adaptations, based on the total
data available at each interim analyses.
trial designs. Group sequential monitoring allows interim analyses of
accumulating data, which can lead to early stopping for a positive or
negative outcome. In adaptive designs, interim data are used to modify
aspects of the trial while still protecting the type I error rate. There
must be sufficient information available at the time of an interim analysis
in order to make sound decisions to modify or terminate a trial. When a
study's primary endpoint is only observed a long time after treatment, there
may be little information on this endpoint at an interim analysis. In this case,
it is desirable to use information from short term endpoints instead. The
project will address the case of a trial with a time-to-event primary endpoint,
such as overall survival or time until disease progression, in which longitudinal
data on a biomarker are to be used at interim analyses. This project will
involve (i) developing joint models for the two endpoints, (ii) derivation of
efficient stopping rules, or more general adaptations, based on the total
data available at each interim analyses.
Organisations
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| EP/N509589/1 | 01/10/2016 | 30/09/2021 | |||
| 1943928 | Studentship | EP/N509589/1 | 01/10/2017 | 15/08/2021 | Abigail BURDON (nee VERSCHUEREN) |
| Description | In Phase 3 clinical trials, group sequential trials can be useful for early stopping. This in turn may result in fewer patients recruited and reduced time until a decision. We study two joint models for longitudinal and survival data and present analysis methods for dealing with each joint model. Then, the joint distribution of the test statistics across interim analyses is determined and a group sequential trial can be performed. We show that by including the longitudinal data, the required sample size is dramatically reduced. |
| Exploitation Route | Group sequential trials can be performed for joint models of longitudinal and survival data. Therefore, pharmaceutical companies may wish to use a group sequential trial when assessing data of this form. This may result in early stopping which is an attractive feature to a pharmaceutical company. |
| Sectors | Pharmaceuticals and Medical Biotechnology |