Abstract
Background The Cancer Screening Effectiveness and Research in Community-based Healthcare (SEARCH) project aims to create infrastructure to support conducting comparative effectiveness research in cancer screening. As a proof of principle, we have created a multi-site VDW-like PAP data structure and populated it at the Group Health Research Institute (GHRI).
Methods In this poster, we provide the PAP table structure that has been defined by SEARCH and the extract, transform, load (ETL) process employed at GHRI to populate the table. We report basic descriptives on the loaded data including the rates of PAP testing over time by population groups.
Results Group Health enrollees whose primary care clinics are owned and operated by Group Health (referred commonly as the Group Practice Division (GPD)) appear to have complete data capture of PAP tests and results. The rate of testing among the population seems to have declined in the most recent years, likely due to changes in cervical screening guidelines. In primary care clinics not owned and operated by Group Health (referred commonly as the Network), the number of PAP results found is considerably lower. This low rate of PAP resulting is due to incomplete data capture.
Conclusion Our findings show that research projects like the SEARCH cervical proof of principle that use PAP results as a key component of their study must define their population of interest based on the individual person’s primary care clinic and its relationship with Group Health to ensure complete data capture. This data structure, while created under the SEARCH project, has been designed for future use in other studies.
