Use this URL to cite or link to this record in EThOS:
Title: Incorporating external evidence syntheses in the design and analysis of trials
Author: Clayton, Gemma
ISNI:       0000 0004 7972 7674
Awarding Body: University of Bristol
Current Institution: University of Bristol
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
When a trialist is designing a trial, related work will often be used to inform several aspects. This information is often used informally, such as using a systematic review to indicate whether a gap in the current evidence base justifies a new trial. External evidence can be used more formally, by explicitly incorporating it in a Bayesian framework through a prior distribution. Funders often highlight the importance of taking into account existing evidence when planning a new trial. Researchers and trialists acknowledge that existing evidence should be used to inform new research to reduce research waste. However, the prevalence of explicitly using external evidence through informative prior distributions is low and there is still much controversy around its use in all stages of a trial. In this thesis, we explore whether and how trialists could use a synthesis of external evidence in the design and analysis of a clinical trial, through a Bayesian analysis. We begin with a survey and qualitative study to capture the current use of evidence synthesis by trialists and reasons why it might not be used in practice. In the remainder of the thesis, we assess and extend methods in areas where external evidence could have the most benefit to a trialist. We focus on the following three case studies:- External evidence on likely bias in a trial, based on information from meta-analyses within meta-epidemiological studies. - External evidence on the likely effect size in a trial, based on information from similar trials within meta-analyses. - External evidence on likely outcomes in a trial, based on information from similar patients within trials.
Supervisor: Jones, Hayley ; Higgins, Julian Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available