One of the nation’s pioneers in research of healthcare information technology is joining forces with an analytics vendor in an effort to commercialize text analytics technology that seeks to gain medical value from unstructured data within providers’ digital health records.

The Regenstrief Institute and Health Catalyst announced the initiative during the HIMSS17 Conference and Exhibition in Orlando. Under the agreement, the partners would aim to bring Regenstrief’s artificial intelligence-powered text analytics technology to the provider marketplace, with additional support from Health Catalyst.

The initiative is important because it holds the promise of enabling provider organizations to apply AI against unstructured data within their records systems to permit localized, granular research. Information embedded in the unstructured text portion of records contains the vast majority of patient information in EHRs, experts say.

In addition to its long-standing efforts in HIT research, Regenstrief helped develop the Indiana Health Information Exchange, through which it internally developed its nDepth technology, which it’s used to search the text of more than 230 million records from 17 million patients.

Dale Sanders

Also See: Machine learning seen speeding diagnoses

As a commercialization partner, Health Catalyst will incorporate the nDepth solution in the company’s analytics platform, which it says is used by health systems serving 85 million patients in the U.S.

“Some 80 percent of clinical data is locked away in unstructured physician notes that can’t be read, so it can’t be accessed by advanced decision support and quality improvement applications,” says Peter Embi, MD, the Regenstrief Institute’s president and CEO. The joint effort could “help millions of patients benefit from the untapped potential hidden within unstructured data.”

Embi says nDepth can be used to analyze any document type, enabling clinicians or quality improvement professionals to discover meaningful insights. It uses natural language processing—a combination of linguistics, pattern recognition and machine learning—to derive meaning from text. To that technical mix, nDepth adds clinical domain expertise and extensive phenotype libraries that have been developed by clinicians.

Regenstrief has been able to use nDepth to find patients with metastatic melanoma; identify pre-diabetic patients for clinical trials; capture hypoglycemic events; identify patients with a family history of lung cancer; detect treatment failure for patients suffering from insomnia; and identifying “triple negative” breast cancers.

“Text and (natural language processing) technology in healthcare has been stuck in the realm of tinkering and has not had an impact on the front lines of care,” says Dale Sanders, executive vice president of development for Health Catalyst. “This partnership is going to make a breakthrough contribution across the board…in clinical research, quality of care, precision medicine and cost reduction.”

The joint effort to marry the products was assisted by Memorial Hospital at Gulfport (Miss.), a 445-bed hospital, which served as the co-development partner and first deployment site for the integration of nDepth into Health Catalyst.

“It’s important for community health systems like Memorial that have made significant investments in electronic health records to take the next step in care by unlocking the value in the EHR's unstructured data,” says Gene Thomas, the facility’s CIO. The incorporation of nDepth into Health Catalyst’s analytics tool “is enabling that value to impact the front lines of care for the first time.”

Register or login for access to this item and much more

All Health Data Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access