2015 Up-and-Comer: Health Fidelity: NLP for Health Risk Assessment
To accompany our Healthcare Informatics 100 list of the largest companies in U.S. health information technology every year, we like to give readers a heads-up on some fast-growing companies that could very well make the HCI 100 in years to come. In fact, one of the companies we chose as an Up-and-Comer in 2014, Evolent Health, recently registered for a $100 million initial public offering.
Some of the firms in this group may not have much revenue yet, but their growth trajectory suggests they could have a significant impact on the healthcare sector. Others are coming at seemingly intractable problems in healthcare from completely new angles. Keep your eye on these six.
Health Fidelity is one of those startups that had to veer slightly from its original business plan, but has now found a very promising niche. Started in 2011 by a few former health IT executives and clinicians, the company sought to help organizations organize and analyze all the data being collected through the EHR boom. “We saw parallels between other IT systems that had collected a lot of data in operations such as ERP systems,” says Anand Shroff, chief technology and product officer, and one of the co-founders. He calls Health Fidelity a classic Bay Area garage startup. “We spent time understanding what can be done with that data to see if a viable company could be built,” says Schroff, who previously was vice president of health information exchange and EHR products at Optum. He came to Optum by way of acquisition of Axolotl Corp., where he headed product strategy and development.
But they soon ran into a familiar problem: unstructured data. “We figured out that most data from EHRs was unstructured and we couldn’t run analytics on that data, even though it is high-value data,” Shroff explained. “It was a dead end. That was disappointing.”
Anand Shroff
The Health Fidelity team contacted Prof. Carol Friedman at Columbia University, who is an expert in natural language processing (NLP). She had been working on the exact same problem, and they struck a deal to license some intellectual property that allowed Health Fidelity to build a high-quality NLP engine that would let it put the unstructured data into a form that would make it computable for business intelligence.
They also spent time interviewing executives from several leading academic medical centers to understand their pressing needs. “Outcomes analysis was one,” Shroff says. “ACO [accountable care organization] cohort identification was another. A really interesting one that came to the surface was risk adjustment in Medicare Advantage populations.”
The San Mateo, Calif.-based Health Fidelity worked with UPMC Health Plan in Pittsburgh, to use EHR data to find appropriate conditions to identify risk factors correctly. “UPMC wanted to apply NLP to patient charts so that chronic conditions could be found proactively,” Shroff explained. “We worked with them over a year and built out our risk adjustment portfolio powered by our NLP technology.” It went live in April 2014. UPMC was able to find conditions that existed in charts that generated significantly increased reimbursements for 2014, he adds. At that point, UPMC Health System’s venture arm, UPMC Enterprises, made a significant investment in the company to fund further growth.
The privately held company does not divulge revenue, but it has grown from 10 people in April 2014 to 40 now and will likely end 2015 with more than 50. “This is the year we are scaling up both headcount and revenue,” Shroff says, “and our trajectory is expected to be significant over the next three years.”
To lead that growth, Steve Whitehurst recently joined the company as CEO. He previously served as senior vice president and general manager of Stericycle, a biohazard waste disposal company. “He has had a number of experiences taking developed technology and commercializing and scaling companies. We were at the stage to do that,” Shroff says. “The most exciting thing is that we are generating real returns for clients.”
