It's pretty well understood now that a small subset of patients drives most of the costs in healthcare. But on any given day, physicians aren't necessarily using their time to focus on these individuals.
That's a workflow issue. But in the world of risk-based payment models, physicians will be key to intervening with high-risk patients and keeping them healthy.
In 2011, executives from the Advisory Board Co., including former CEO Frank Williams, spun out a company called Evolent Health from the consulting giant. They partnered with Pittsburgh-based health system UPMC, which like a growing number of health systems, operates as both a provider and a health plan.
UPMC had developed a technology platform known as Identifi that can analyze data from disparate sources, such as insurance claims or electronic health records and create care-management plans.
Evolent's platform aims to not only pinpoint high-risk patients, but find ways to change their behavior. A lot of that work comes from focusing on and motivating physicians.
That makes the company different from other software companies trying to help providers operate in a risk-based environment. “It was developed by a provider in a real-world setting,” Williams said at the company's San Francisco office. “We wanted to stay true to our provider philosophy … and we believe in a provider-driven model.”
The Identifi platform allows users to set the parameters for determining who falls into the specific high-risk populations they want to target. For instance, they could tell the software that they want to focus on the “18 club”—patients with nine different doctors on nine or more medications.
The platform then singles out action items (such as making sure patients know how to use their medication) and potential problems (such as flagging patients who are smokers). “You need to be able to stratify the patient population to know where to take action first,” said Chad Pomeroy, Evolent Health's chief technology officer.
The next step is to work with physicians on care-management plans. If a home visit is recommended, for instance, the technology platform can track whether it occurred. It also can identify when clinicians veer from certain protocols.