Meta-Analysis of controlled clinical trials: Models and assumptions, inferential principles, and computational algorithms
Research Objectives:
Meta-analyses of published clinical trials have received increased attention recently with some meta-analytic publications having had a big impact on the cost-benefit assessment of important drugs. Much of the methodological work in meta-analysis has so far focused on evidence synthesis of summary statistics from published clinical trials. Less work has been done on meta-analysis on individual patient data. Both topics are relevant to the pharmaceutical industry. The research in meta-analysis methodology has in parts been isolated from other fields of mathematical statistics and is lacking an integrative framework clearly separating statistical models and assumptions, inferential principles, and computational algorithms. The very extensive past research on ANOVA and MANOVA of un- balanced designs, variance component models, generalised linear models with fixed and/or random effects, provides a wealth of potentially useful approaches and insights. This refers to both Bayesian and frequentist statistical inference.
Description of work:
The research will focus on:
(a) giving a unified view of the literature in these fields,
(b) delineating similarities and differences with approaches popular in meta- analyses applications, and clarifying the place of meta-analysis approaches in the wider context of these theoretical frameworks,
(c) comparison of methods with recommendations, potentially new methodological suggestions, including application of suggestions and existing methods to actual clinical trial data, and
(d) possible software implementation.
Host Institution: Novartis