Building the hypertension registry: adventures in using the VDW model [presentation] Presentation uri icon
Overview
abstract
  • Background/Aims : The Hypertension Registry was built to support research on the recognition, detection, and treatment of hypertension at three HMO Research Network (HMORN) sites. The goal was to build a data mart modeled after the Virtual Data Warehouse (VDW) containing patients identified as hypertensive. Two of the sites already had VDWs in place at the start of the project, while the third site used the project as a springboard to build a VDW.
    Methods : A team consisting of data managers and analysts from each site met weekly in order to 1) determine exactly which existing VDW elements and possibly new fields would be required to support hypertension studies, 2) help the VDW development efforts by clarifying data element creation rules and interpretations, and 3) document any new columns or tables in a Hypertension Registry data dictionary. All three sites had equal responsibility for registry development, documentation, and writing distributed code for quality checks rather than a lead site being designated for these efforts.
    Results : An algorithm for defining hypertension was used to create a “superset” of hypertensive patients. Using the protocol as a guide, the data team examined existing VDW tables to determine if all aims could be met with current data elements and developed solutions for additional variables or tables necessary to support planned papers.Challenges to harmonization of data elements were encountered in relation to data quality, data availability, and differing business models. Data team members gained a better understanding of the strengths and benefits of having a VDW in place, but also learned of areas where differing interpretations of the VDW data dictionary necessitated rework in the VDW or the need for better documentation. This presentation or poster will share how the team attempted to resolve these challenges in order to aid similar studies in the future.
    Conclusions : The Hypertension Registry is currently supporting at least 12 different hypertension papers and has been fully documented so that Hypertension Registries could potentially be built at additional sites. Continued discussion of VDW supported research efforts within the HMORN community is needed to learn how to best identify and create efficiencies.

  • Research
    keywords
  • Data Systems
  • Hypertension
  • Registries
  • Research Support