Representing and Leveraging Heterogeneity Between Data Sources in Multi-database Pharmacoepidemiologic Studies

Background: Regulatory authorities often require data from multiple data sources to be used in a single study, to enhance the generalizability of results and/or obtain sufficient sample size. However, heterogenous results across data sources are often observed. Interpretation of such heterogeneity would profit from a clear understanding of the context and purpose of each data source, and its diversity A number of efforts have been made to characterise data sources used in pharmacoepidemiologic research, including the development of metadata catalogues to describe data sources, and to leverage data diversity to improve results interpretation.

Objectives: To illustrate initiatives aimed at proposing a unified framework to characterize data diversity, including metadata lists, and addressing diversity across geographic regions
To illustrate strategies to leverage data diversity to improve intepretation of study results

Description: This symposium is led by the DIVERSE initiative of the ISPE Database SIG, which is conducting a scoping review on how to represent and leverage data diversity for pharmacoepidemiology. Updates on recent activities in this field will be shared, including DIVERSE itself, and the MINERVA project, which was funded in Europe with the aim of defining a list of metadata to chatacterise data sources. The experience gained will then be discussed in a broad international perspective. Finally, this symposium will provide an example on how knowledge of data diversity can be used to improve quality of generated evidence: the case of computation of background incidence rates of adverse events of special interest for COVID-19 vaccines. The presentations of the speakers will be followed by 30 minutes for audience engagement and panel discussion, guided by the symposium chair

Chair: Robert Platt, Gillian Hall
Romin Pajouheshnia (DIVERSE): Results from the ISPE-sponsored DIVERSE scoping review on representation of data diversity: existing recommendations and tools for the representation, reporting and leveraging of data source heterogeneity in multi-database studies
Susana Perez-Gutthann (MINERVA): Findings from the EMA-funded MINERVA project on data representation: adoption of a conceptual framework to represent data sources, recommended metadata to collect and lessons learned
Soko Setoguchi (DIVERSE): Data diversity between America, Australia and Asia
Corrinne Willame (ACCESS): Leveraging data diversity to improve quality of evidence in a study of background incidence rates of adverse events of special interest for COVID-19 vaccines