Using Analytics to Support an Ambulatory ICU Model
One year ago, Stanford (Calif.) Hospital and Clinics embarked on a new approach to treating high-risk patients, including those with multiple chronic conditions. The Stanford Coordinated Care project puts sophisticated I.T. tools to the task of monitoring and treating this otherwise difficult population of patients.
At the same time, it upends traditional clinic models. Dubbed the “ambulatory ICU,” the clinic is run by two physicians, but much of the day-to-day work is delegated to a team of medical assistants, says Jorge Wilson, director of clinical and business analytics at Stanford. Wilson helped craft the analytics programs the clinic uses to stratify patient risk and monitor their progress. He’ll discuss the effort at Health Data Management’s upcoming Healthcare Analytics Symposium, July 15-17 in Chicago. (Registration information and more details about the event are at www.healthdatamanagement.com/conferences/hcs/)
“It’s a completely different primary care model,” Wilson says, with more focus on “health screenings, health behaviors and trying to keep patients out of the ED. If a patient is at-risk enough, we will even make house calls if that is what it takes. We want to change health behaviors, not just improve outcomes.” The clinic is reimbursed on a capitated, per-member, per-month basis from various payers.
The clinic targets 10 chronic diseases, including diabetes, kidney failure, and hypertension, Wilson says. Care is documented in an Epic EHR, which is a key source of information for an enterprise data warehouse Stanford is currently building out in a step-wise manner. “For regular primary care, Epic is great at managing diseases,” Wilson says. “It has many tools to help manage diabetics, for example. But we have diabetes patients who also have hypertension. We are moving into a more population-centric view than just treating individual disease groups.”
Working in conjunction with the EHR is an analytics system, which uses data visualization tools from QlikView and data extraction tools from Health Catalyst. The set-up aggregates data from Epic and presents an assessment of patients across some 30 clinical measures. The measures are color-coded in a red, yellow, green format to symbolize the continuum of risk for a given patient from high-risk to under-control. Medical assistants managing a panel of patients review these scores weekly in preparation for upcoming visits and to ensure that patients are receiving appropriate tests and interventions. “What happens is protocol-driven, but the assistants need the data in front of them to act on it,” Wilson says.
Providing patients access to their data is another way to drive change in health behavior, he adds. “We have found that having data in the hands of patients can be powerful for self-management,” he says. Providers share the quality measures with patients during visits and the next step, Wilson says, is figuring out how to present the data via Stanford’s patient portal.
Wilson’s talk, “Analytics + Clinical Effectiveness Methodology Drives Improvements in Population Health Management,” takes place Tuesday, July 16, at 1:45 p.m. at the Westin Michigan Avenue.