“Many predictive factors are included in structured data within today's EMR systems, but a lot of it is hidden in doctors' notes, discharge papers and other sources of unstructured data,” explained Sean Hogan, VP of global healthcare for IBM, in a release. “By tapping into the unstructured data, our clients have more complete and accurate information that allows them to make targeted interventions when appropriate that can help prevent more severe and costly medical complications.”
Through content analytics and predictive modeling techniques, Carilion was able to identify at-risk patients with an 85% accuracy rate, including 3,500 patients at risk for developing heart failure who would have been overlooked using traditional methods. The effort was part of a pilot program designed for early intervention.
In considering physiological data, prescription drug use, obesity, previous diagnoses, lifestyle and environmental factors, the program detected patients at risk of developing heart disease within a year so that early and less costly interventions could be applied.
These predictors might have otherwise been missed if not extracted from doctors' notes, which IBM's content analysis software is designed to do. In conjunction with Epic's EMR system, IBM's software extracts information from physicians' notes and translates that information into a format that can be uploaded to the patient record and even correlated with diagnosis and treatment codes.
Heart failure affects more than 5 million adults in the U.S. and costs the country $32 billion annually.
Follow Rachel Landen on Twitter: @MHrlanden