Getting Real About Real-World Data From Electronic Health Records

PURPOSE: Real world evidence (RWE) is only trusted by regulatory bodies, health technology assessment (HTA) organizations, payers, clinicians and patients if the underlying real-world data (RWD) is demonstrated to be high quality and fit-for-purpose. Electronic health records (EHRs) are an important source of real-world data: they are the primary record of the patient symptoms, diagnoses, treatments, outcomes, and in many cases, preferences and experiences. However, EHRs were not designed for research, and clinician engagement and interaction with EHRs varies widely from scope of practice to health care organization type. Clinician documentation in EHRs can often be limited by time pressure and administrative burden. Moreover, much information is captured in unstructured formats within the EHR, and coding systems and other factors often vary between health system EHRs. Patient records may also be split between multiple health systems and private practice EHRs. In this workshop, we will explore how all of these factors impact RWE generated from EHR data, and how these challenges can be overcome to improve trust and transparency around uses and the reliability of EHR data for research and regulatory purposes.

DESCRIPTION: In this workshop, Rachele Hendricks-Sturrup will describe health system practices in generating RWE to support learning health system goals, and Duke-Margolis’ current focus on operationalizing EHR data relevance, reliability, and quality to build trust and transparency among health system stakeholders. Mac Bonafede, Elise Berliner and Louis Ehwerhemuepha will take some of those examples and describe the strengths and weaknesses of EHR data for the specific questions and how they are working to improve the validity and quality of EHR data to answer these questions. The panel will take questions on specific challenges audience members are facing using EHR data and other RWD, and brainstorm solutions with panel and audience members.