Researchers: EHR data valuable in genetics studies

Electronic health records can be a valuable source of relevant clinical data for genomic studies, according to researchers from a consortium of U.S. healthcare organizations.

An eight-page report on the research, "Electronic medical records for genetic research: Results of the eMERGE Consortium," was published last week at ScienceTranslationalMedicine.org

The goal of the research, according to the report's 15 authors, was to determine whether data from EHRs could do a good enough job in identifying the phenotypes—outward manifestations—of certain diseases that it could then be used in genomic studies that look for links between health conditions and genetic markers.

The researchers analyzed EMR data from five care sites—Group Health, Seattle; the Marshfield (Wis.) Clinic; the Mayo Clinic, Rochester, Minn.; Northwestern University, Chicago; and Vanderbilt University, Nashville. The phenotypes to be identified were five conditions or diseases: dementia, cataracts, peripheral arterial disease, type 2 diabetes and cardiac conduction defects. Three of the sites used home-grown EHR systems; two used commercial off-the-shelf systems. The EHR systems captured patient records in both structured data and free-text forms.

Three sites also used what the authors described as "robust" natural language processing tools to extract structured data from the free-text portion of patient records. Each site has its own DNA biorepository linked to the EMR via a unique research identifier. At most sites, patients "opted in" to participate in the trial.

The positive predictive values of the EHR-extracted data ranged from 73% to 98%; negative predictive values ranged from 98% to 100%, according to the research.

The researchers concluded that EHRs in current routine clinical use can be valuable in identifying phenotypes for genetic studies and thus that widespread EHR adoption "creates the potential for a quantum shift forward in the availability of longitudinal, real-world clinical data for genetics research."