Big data retrieval is becoming a reality. Using a platform that can collect and standardize de-identified data from multiple healthcare systems’ various information systems, researchers were able to present a retrospective cohort study of patient characteristics associated with venous thromboembolic events (VTE), according to research published July 3 by the Journal of the American Medical Informatics Association .
“Although the theoretical informatics infrastructure exists to standardize, normalize and aggregate clinical information from different EHRs, efforts to date have not succeeded in integrating EHR data into a readily usable, scalable platform for clinical research in real time, especially across different healthcare systems,” according to the study’s lead author, David C. Kaelber, MD, PhD, MPH, of Case Western University and the MetroHealth System, both in Cleveland.
The Explorys platform, developed by a Cleveland-based health IT vendor of the same name, contains health information on several million patients from multiple, distinct healthcare systems. The platform utilizes a health data gateway server located behind the firewalls of participating healthcare organizations and collects data from their various information systems, including EHRs and billing systems. The data is then normalized according to standard ontologies, such as ICD-10 and SNOMED-CT, included in the National Library of Medicine’s unified medical language system.
Using the platform, researchers were able to demonstrate the association between patient characteristics and VTE by identifying an aggregated cohort of VTE patients and comparing them to patients without a VTE. The platform returned 21,210 incidents of VTE in 959,030 patients followed for a mean of 3.5 years each.
Researchers determined a positive correlation between VTE and increasing height and weight. They also determined a higher rate of VTE among black patients and a lower rate of VTE among Hispanic patients compared to the rate of VTE among white patients.
The results are similar to those contained in published studies, but the importance of this study is that it demonstrates the potential of clinical informatics tools to perform large-scale, retrospective studies with limited resources, according to Kaelber.
“This example demonstrates the use of clinical data from disparate healthcare systems using different EHRs in an expanded and novel way,” he wrote. “The Explorys model as a clinical data repository solution appears more efficient to set up and use than other solutions we are aware of and provides near real-time data updates.”