Dr Sylwia Bujkiewicz: Use of Bayesian multivariate meta-analysis to inform decision making in HTA
An Institute of Health Research Health Statistics Group seminar event
|An Exeter Medical School seminar|
|Speaker(s)||Dr Sylwia Bujkiewicz, University of Leicester|
|Date||22 February 2016|
|Time||12:30 to 13:30|
Evidence-based decision-making requires careful synthesis of available evidence.
When assessing new health technologies to make reimbursement decisions, the health technology assessment (HTA) process relies heavily on meta-analysis of effectiveness of new interventions. The evidence base is typically obtained from the systematic literature review of randomised controlled trials and sometimes observational studies. There is often a lot of heterogeneity in reporting of clinical outcomes due to, for example, variety of scales on which effectiveness can be measured, different time points at which different studies report their outcomes or different control arms.
Bayesian statistics provides flexible framework for modelling complex data structures by allowing multiple parameters to be modelled simultaneously. Network meta-analysis (NMA) facilitates simultaneous comparison of multiple treatment options whereas multivariate metaanalysis (MVMA) allows to model jointly treatment effects on multiple correlated outcomes.
While NMA is becoming a standard tool for synthesising evidence in HTA, MVMA has not been widely used despite of substantial methodological developments in this area and many advantages of this approach to evidence synthesis.
This seminar aims to introduce the concepts of Bayesian methods for efficient synthesis of evidence in HTA. MVMA methodology will be introduced and the application illustrated in scenarios for its use for purpose of (i) combining data on correlated outcomes from diverse sources of evidence and the implications for estimating the health-related quality of life values, (ii) predicting treatment effect on a target clinical endpoint from treatment effects on surrogate endpoints , and (iii) exploiting such prediction framework to inform decision modelling, based on examples in rheumatoid arthritis, multiple sclerosis and cancer.
Dr Sylwia Bujkiewicz is a Lecturer in Biostatistics in the Department of Health Sciences, University of Leicester. Her background is in Applied Mathematics and Medical Statistics. She previously held research and academic posts at Wroclaw University of Technology, University of Nottingham and University of Leicester. She joined the Biostatistics Research Group in Leicester as a lecturer in 2011 and her role focuses on research of Bayesian multi-parameter evidence synthesis methodology and the application of such methods to modelling surrogate endpoints.