Geneticists Decode e-Records

General patient data contained in electronic medical records can help genetic researchers quickly and accurately identify patients for their investigations, thereby slashing the cost of collecting research data, according to a Northwestern University study.

Researchers analyzed electronic records for information generated from routine doctors' visits -- including diagnoses, medications and laboratory tests -- for signs of five diseases: type 2 diabetes, dementia, peripheral arterial disease, cataracts and cardiac conduction. They used e-records from five national sites, each of which used different vendors, according to a Northwestern news release.

The data in the electronic records allowed the researchers to correctly identify patients' diseases 73 percent to 98 percent of the time. They also were able to rule out the five targeted diseases with at least 98 percent accuracy. The researchers' predictions were confirmed with the patients' doctors.

According to an abstract of the study, the case-identification rate improves when electronic records use natural language processing that helps computers to understand nuance and jargon of human speech.

Eventually, patients' genomes could be included in their medical records, says the lead investigator, Dr. Abel Kho, an assistant professor of medicine at Northwestern's Feinberg School of Medicine. "With permission from patients," he says, "you could search electronic health records at not just five sites but 25 or 100 different sites and identify 10,000 or 100,000 patients with diabetes, for example."

The Northwestern study was published last week in Science Translational Medicine journal. The research project received financial support from the National Human Genome Research Institute, with additional funds coming from the National Institute of General Medical Sciences.