Since its inception in early 2003, the Veterans Affairs Rheumatoid Arthritis registry (VARA), has seen a steady increase in the number of participants, said Ted Mikuls, M.D., its founder and principal investigator on VARA research.
More than 2,700 rheumatoid arthritis patients from 13 Veterans Affairs medical centers have enrolled in VARA, a big jump from the few hundred who enrolled during the first few years of its existence.
Dr. Mikuls, a professor in internal medicine and rheumatologist, said VARA has served as an important resource for both VA and non-VA researchers across the country, serving as a focus for a wide scope of clinical and translational research.
Work coming from VARA has ranged from studies of disease outcomes and their determinants, disease epidemiology and biomarker validation, among others. In contrast to many other disease registries, VARA links Natural Language Programming (NLP)-based capture of disease activity measures with a rich array of administrative data.
Dr. Mikuls is collaborating with investigators from the VA medical center in Salt Lake City on fully automating data abstraction using NLP technology. NLP provided an automated approach to mining relevant data directly from text fields found in electronic health records.
“Integrating NLP in the registry’s day-to-day workflow is quite powerful, as it greatly reduces or even eliminates the burden posed to our researchers and at the same time reduces data entry errors,” Dr. Mikuls said.
The investigators hope to identify and pull additional data elements from the clinical notes provided in participants’ electronic health records. These data are then linked with other datasets available in the VA that provide information such as pharmacy dispensing, laboratory and imaging studies, vital status and cause of death, as well as other health care utilization.
VA medical centers currently contributing to the registry include Omaha, Dallas, Iowa City, Philadelphia, Pittsburgh, Portland, Ore., Birmingham, Brooklyn, Washington, DC, Salt Lake City, Denver and Jackson, Miss.
“We believe that once we get these new and highly efficient approaches of data capture more widely implemented, it may have the real potential of letting us build VARA even more and make a broader impact in arthritis research,” Dr. Mikuls said.