NC TraCS Director, IDSci Staff Published in Pharmacoepidemiology & Drug Safety

Performance of a computable phenotype for identification of patients with diabetes within PCORnet: The Patient-Centered Clinical Research Network, a paper recently published in Pharmacoepidemiology & Drug Safety, describes the development and testing of a computable phenotype to identify patients with type 2 diabetes (T2DM) within PCORnet.

PCORnet, a distributed research network of 13 clinical data research networks and 20 patient-powered research networks, is a novel data platform for the identification of patients with chronic diseases, such as T2DM, and the efficient conduct Of clinical and epidemiological research.

John Buse, MD, PhD, Director of the North Carolina Translational and Clinical Sciences (NC TraCS) Institute and an author on the paper said, “I am so proud of this PCORnet work! The ability to identify patients with type 2 diabetes from our electronic health record with accuracy is essential for studies to evaluate health outcomes in the population as well as to recruit and randomize patients in clinical trials.” Buse emphasized that, “without this, functionally we had huge amounts of data but uncertain ability to focus on the right people to develop new insights.”

Using the computable phenotype developed at Vanderbilt University Health System, researchers identified more than 50,000 patients with T2DM across the four participating PCORnet sites. The positive predictive value of patients randomly selected for validation through structured chart reviews was 96.2%.

The University of North Carolina at Chapel Hill was one of the four PCORnet sites participating in this work. The four sites represented different electronic health record systems, patient populations, as well as geographic regions with a high prevalence of T2DM.