All Times EDT
Keywords: Perspective Plan, meta-analysis
Many European countries have national registry databases. These data sources provide opportunities for medical research. However, for some research, the sample size in each data source is too small to draw a robust conclusion. But the privacy regulations in EU requires individual data accessible only within the country, and inaccessible outside the country. Some data owners insist the data accessible only to the internal researcher. This situation hinders researchers to combine data of individual patients from several data sources together to conduct analyses. However, it is allowed to conduct analyses on each data source separately. Therefore, a new framework of 2-step data analysis strategy become feasible: 1) A perspective planned analysis is conducted on each data source that to generate aggregated / summary following the standards for all sources; 2) the summary data/results from individual source will be obtained for comparison or combined for meta-analysis. In this research we will explore this 2-step analysis strategy. In order to compare results from different databases or conduct meta-analysis based on the results from different databases, it would be helpful if we could adjust for the differences in distributions of (key) covariates in the analysis cohorts derived from databases.