Browsing by Author "Craig, Jennifer L."

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Using Electronic Health Record Data to Improve Community Health Assessment
    (UKY, 2016) Dixon, Brian E.; Zou, Jian "Frank"; Comer, Karen F.; Rosenman, Marc; Craig, Jennifer L.; Gibson, P.; Epidemiology, School of Public Health
    Background: Community health assessments assist health departments in identifying health needs as well as disparities, and they enable linking of needs with available interventions. Electronic health record (EHR) systems possess growing volumes of clinical and administrative data, making them a valuable source of data for ongoing community health assessment. Purpose: To produce population health indicators using data from EHR systems that could be combined and visually displayed alongside social determinants data, and to provide data sets at geographic levels smaller than a county. Methods: Data from multiple EHR systems used by major health systems covering >90% of the population in a metropolitan urban area were extracted and linked using a health information exchange (HIE) network for individuals who had at least two clinical encounters within the HIE network over a 3-year period. Population health indicators of highest interest to public health stakeholders were calculated and visualized at varying levels of geographic granularity. Results: Ten population health indicators were calculated, visualized, and shared with public health partners. Indicators ranged from the prevalence of a disease to the proportion of individuals with poor maintenance of their chronic condition. Calculating rates at the census-tract level or larger (e.g., average population size > 4000 people) is preferable to smaller geographic units of analysis. Implications: Extraction and linking of EHR system data are feasible for public health via an HIE network. While indicators can be derived, biases exist in the data that require more study. Further, HIE networks do not yet possess data for all conditions and measures desired by local public health stakeholders. The data that can be extracted, however, can be combined with public datasets on social determinants