Network Meta-Analysis Combining Aggregated and Individual Patient Data

• Activity Number: 372
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Anna Wiksten* and Ekkehard Glimm
• Companies: Novartis Pharma AG and Novartis
• Keywords: Network meta-analysis ; Aggregated data ; Individual patient data ; Treatment-covariate interactions ; Linear mixed models

Abstract

Network meta-analysis (NMA) is a method where multiple treatments from several trials are compared in a single analysis. Pharmaceutical companies perform NMAs to support submissions of new drugs or for market access and reimbursement purposes. Furthermore, in the development of a new drug, NMA may be used to support planning of future studies. When NMA is performed within a company, the company has access to individual patient data (IPD) of their own studies whereas usually only the aggregated data(AD) is available from the competitors' studies. In this presentation, we will introduce different models to perform NMA with IPD and AD and the methods will be applied to a case study. Both frequentist and Bayesian approaches will be considered. We will also present different NMA models for analyzing data with an individual-patient-component that includes additional covariates with a potential interaction to the treatment effect. Results from a simulation study comparing performance of the different models and different types of data will be presented.