It's all about the data
For anyone who's ever conducted an old-school chart audit, some benefits of using electronic health records (EHRs) in research or quality improvement initiatives are fairly obvious. No more pulling individual paper charts to collect data slowly and tediously, one patient at a time. No need to interpret handwriting from all the different care providers who have contributed to an individual patient chart over time. No more visits to dusty, musty record rooms to pull charts that are no longer in active rotation. No more allergy attacks in the middle of a paper-based chart audit. (Maybe that last one is just me.)
Of course, the benefits of using data from the EHR for research and quality improvement initiatives extend well beyond these fairly mundane issues. In brief, some benefits include the increased availability of patient data, checking clinical trial feasibility through counts of patients meeting certain criteria, and supporting patient recruitment for research studies. However, there are also challenges that need to be overcome when using EHR data to support health research or to improve patient care, especially across health systems.
The challenge of data compatibility
One challenge faced by those who need to conduct research across health systems is that of data compatibility. Even something as seemingly straight forward as a date can be captured in numerous ways, such as YYYY-MM-DD or DD-MM-YYYY or MM-DD-YYYY. The UNC Health Care System's EHR is structured for the clinical and billing needs of our health system. Not surprisingly, other health care systems structure their EHRs to suit their specific institutional needs. If you tried to compare data from multiple systems, you'd experience issues with data compatibility. Addressing these issues can be a time-consuming and expensive undertaking, especially if you're a researcher going it alone.
Clinical Data Research Networks
Clinical Data Research Networks (CDRNs) support research using EHR data from multiple health systems through member institutions agreeing to a common data format, thus harmonizing the data and allowing for cross-institution data queries. In utilizing a CDRN, researchers don't have to go it alone and cross-institution clinical research becomes much more feasible.
There are other benefits to working through a CDRN, including assistance connecting researchers with collaborators at other institutions within the network. Additionally, each CDRN has its own oversight and governance processes that can help smooth the regulatory waters with templates for Data Use Agreements, IRB Reliance Agreements, and approval processes for data requests.
UNC currently participate in three CDRNs, with a fourth under development. Each network has different member institutions, different data available, as well as a different mission and scope. Carolina's involvement in these networks is managed by the NC TraCS Institute.
Examples of CDRN-facilitated research
CDRNs can support a wide variety of research studies – from observational studies like the PCORnet Bariatric Study to ADAPTABLE, PCORnet's first pragmatic, randomized controlled clinical trial. UNC is a study site for both of these projects.
For the bariatric surgery study, the work at UNC is being led by surgeon Dr. Meredith Duke. This study is comparing the benefits and harms of three types of bariatric surgery. The data collection phase of this study is nearing completion, so stay tuned for results!
ADAPTABLE (Aspirin Dosing: A Patient-centric Trial Assessing Benefits and Long-Term Effectiveness) is a five-year pragmatic trial comparing the effectiveness of two different daily doses of aspirin commonly used to prevent heart attacks and strokes in people with heart disease. The plan is to enroll and follow up to 15,000 participants from across the country, including right here in North Carolina. Excitingly, the ADAPTABLE team is almost halfway to their enrollment goal! The UNC site PI for ADAPTABLE is Dr. Darren DeWalt.
Next up – Join us as we explore the development of 'The Adventures in Data Sharing' comic with our Biomedical Informatics Team.