Elizabeth Stojanovski
Bayesian MetaAnalysis to Account for Heterogeneity in Studies Relating Life Events to Disease
231 - 234
2018
12
11
International Journal of Mathematical and Computational Sciences
https://publications.waset.org/pdf/10009797
https://publications.waset.org/vol/143
World Academy of Science, Engineering and Technology
Associations between life events and various forms of cancers have been identified. The purpose of a recent randomeffects metaanalysis was to identify studies that examined the association between adverse events associated with changes to financial status including decreased income and breast cancer risk. The same association was studied in four separate studies which displayed traits that were not consistent between studies such as the study design, location, and time frame. It was of interest to pool information from various studies to help identify characteristics that differentiated study results. Two randomeffects Bayesian metaanalysis models are proposed to combine the reported estimates of the described studies. The proposed models allow major sources of variation to be taken into account, including study level characteristics, between study variance and within study variance, and illustrate the ease with which uncertainty can be incorporated using a hierarchical Bayesian modelling approach.
Open Science Index 143, 2018