QPID Launches NLP Intelligence Platform for EHRs

QPID, a new health care IT software company, has launched with a Web-based natural language processing engine to run on top of electronic health record applications.

QPID, a health care IT software company spun off from Partners HealthCare in Boston, has launched with an intelligence engine of the same name to allow clinicians to extract data from electronic health records (EHRs) and convert it into an actionable format.

Announced Feb. 14, the Web-based search engine uses natural language processing (NLP) to allow doctors to search health records using no-click navigation capabilities. Short for Queriable Patient Inference Dossier, QPID is platform-agnostic, so doctors will be able to use it on top of any EHR software.

A complement to EHRs, QPID will allow doctors, nurses and billing staff to aggregate health record data and search unstructured text, such as dictated physician notes, scanned documents and email messages. The dashboard summary of search results includes treatments received and current and past medications.

The software also allows for more efficient billing, as administrative staff can perform faster searches, QPID reported.

"What QPID does is it indexes both structured and unstructured data," Mike Doyle, CEO of QPID, told eWEEK. "It presents it to the physician or nurse, whomever they may be, exactly the way they want to see it."

When EHRs only capture patient data in a structured format, doctors' productivity is limited since the unstructured narratives are a more natural part of their workflow. Being able to search both structured and unstructured data can lead to increased clinical productivity and improved outcomes, according to the company.

The smart intelligence engine in QPID learns as it interfaces with patient information and queries built around it, said Doyle.

"It gets smarter and smarter the more it's used," he said.

The QPID software allows health care providers to avoid ordering duplicate tests by gaining access to more complete information, said Doyle.

QPID could pull up patient data showing if they had metal in their body in the past.

"Fragments from a gunshot could be in the record but might be in personal history of the patient and not in a structured data field," Doyle noted.

Metal in the body during a magnetic resonance imaging exam could severely harm the patient, so combining QPID with an EHR application could help avoid this occurrence, he said.

With the data that QPID offers, the company can help solve the problem of 100,000 patients dying in hospitals per year from physicians not having complete access to information, Doyle suggested.

Dr. Michael Zalis, an associate professor at Harvard Medical School and a radiologist at Massachusetts General Hospital, and Dr. Mitch Harris, a research scientist and technology architect at Massachusetts General Hospital, founded the company. Doctors at Massachusetts General Hospital and Partners HealthCare in Boston developed the QPID software over the last four years.

Prior to its official launch, QPID's platform has been able to improve 7,000 daily patient encounters per day, according to Doyle. Currently, 5,000 doctors in the Partners HealthCare system use the software, and they're doing more than 15 million searches or queries per month, he said.

Within the next few months, QPID will start to deploy the software in other academic medical centers as well, he said.

Other companies that offer NLP software for EHRs include Nuance and MModal, acquired by JPMorgan Chase & Co.'s private investment arm One Equity Partners in July 2012.