Review of applications of Bayesian meta-analysis in systematic reviews

Abstract

Background: Systematic reviews are important sources of evidence in health care research. These reviews may or may not include meta-analysis as a statistical assimilation of the results of several studies in order to acquire a pooled estimate. Systematic review with meta-analysis is considered as a robust method of evidence synthesis. The methodology concerned with traditional meta-analysis does not incorporate external prior information. Hence, Bayesian methods are essential due to the natural process of incorporating the past information and updating the belief. Bayesian methods to meta-analysis have been developed with a motivation from the limitations of traditional meta-analysis such as dealing with missing data, problem with limited number of studies and problem with sparse event data in both the groups. The present article aims to unearth as to what extent Bayesian methods have been used in systematic reviews, evolution and its applications. This article also highlights the existing challenges and opportunities. Methods: The literature search was performed in databases such as Cochrane, PubMed, ProQuest and Scopus using the keywords “Bayesian Meta-analysis” and “Bayesian Meta-analyses”. All the methodology and application oriented papers specific to Bayesian meta-analysis were considered relevant for this review. Conclusion: Bayesian meta-analysis has gained popularity in the field of evidence synthesis of clinical trials. However, it did not pick up momentum in summarizing public health interventions, owing to the fact that public health interventions are targeted to highly heterogeneous population, multi-component interventions, and multiple outcomes and influenced by the context

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 12 citations exist upstream.