Bayesian Methods-Combining Evidence from Interventions

University Of Pittsburgh At Pittsburgh, Pittsburgh PA

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): Cross-design synthesis is a relatively new research strategy for combining data from studies with complementary designs. In theory, the aim is to capture the diverse strengths while minimizing weaknesses of different study designs to improve knowledge about interventions. Although the approach is anchored in meta-analytic principles, an essential component of cross-design synthesis is the critical review of studies to assess the generalizability of results and to assess the limits and modifiers of those results. Because of the potential biases due to selection effects and confounding in nonrandomized studies, statistical methods for adjustment also play a key role. Cross-design synthesis is an idea that has had limited application. Our goal is to develop the methods and statistical tools for implementation of cross-design synthesis in mental health services research. Our approach will be based on the use of Bayesian hierarchical models. We focus on Bayesian hierarchical models as one solution to the problem of making inferences when syntheses require not only modeling of within and between study heterogeneity, but also the qualitative differences of study types. The Bayesian approach will also help facilitate sensitivity analyses for assessing the robustness of inferences to various model inputs. We will also develop accessible and user friendly statistical software programs to implement the methodology that we develop. To inform our work and help ground the methodological developments with respect to meaningful assumptions, we propose to develop the methodology using a case study that investigates the effects of several interventions for long-term treatment for depression. We will focus on the important problem of identifying patient, provider, and therapeutic factors that are related to successful interventions. We propose to sponsor a workshop to disseminate our findings, to engage the research community in discussion of this potentially controversial methodology, and to facilitate the feedback process that will help us further refine the methodology.