Humana Inc.'s business intelligence application recently sorted through claims to flag one particular member: a 44-year-old man hospitalized for a ruptured appendix. He had shown up as a "high utilizer" of health care services, prompting a call from one of Humana's personal nurses.
The nurse discovered that the man also had cancer and was struggling with social issues. She learned that he was having difficulty paying medical bills. She helped him better understand his medical condition and referred him to several agencies for financial help.
Like most organizations, Humana is looking for ways to keep costs down. Its homegrown BI application helps the health insurer do just that. But perhaps more important, Humana's internally developed BI tool, called Medical Metrics, helps its workers improve the lives of its sickest customers.
"Medical Metrics has allowed us to find a lot of folks who have not been identified for our disease management or personal nurse programs, folks who are high-cost, high-need members," says Gary Thompson, vice president of clinical programs.
N. Venkatraman, a professor and chairman of the information systems department at the Boston University School of Management, says BI is particularly important for health-care- related organizations because of increasing costs and competitive pressures.
Moreover, he says, BI is indeed identifying these high-cost patients who "might have fallen through the cracks."
Humana started the first phase of its BI implementations with Medical Metrics in 2003, says Bruce Sterpka, vice president for corporate information management. Louisville, Ky.-based Humana opted to develop its own BI tools because workers didn't find "an off-the-shelf solution that delivered all that our business asked for."
The company also wanted more flexibility than it could get with commercial products, says Marcus Bourke, Medical Metrics product manager, because it has to rapidly adjust to changes in the health care market.
In addition, Bourke says, Humana wanted a tool that was easy to use. "We want someone to be able to come in the door and, with an hour-and-a-half training session, be up and running with this tool," he explains.
Clinical staff members use Medical Metrics to identify high utilizers. Underwriters use it when developing rates for groups, and actuaries use it to help identify trends.
Data comes almost entirely from internal sources, such as Humana's benefits and claims systems, Sterpka says. Data is scrubbed, collected and stored in the company's data warehouse.
Thompson says the initial intent was to use Medical Metrics on the commercial side, "to be able to understand the claims that come in annually so we can stem that rising cost."
Now, Humana workers in various departments can drill down on cost drivers and get standardized, consistent information on those factors, says Mike Beavin, director of corporate information management.
Medical Metrics users have access to a half-dozen standard reports all connected to the same data source. They can also drill down to very fine details to identify trends and cost drivers.
"If I was sitting in the Louisville market office and wanted to know what was happening there regarding medical trends, I could go into this application and put up market trends and find stats on all the other markets," Beavin says.
Sterpka says users can customize the time periods being compared and drill down to selected markets, products and even subproducts. Without this application, such information was difficult to access and often late.
"What typically happened is someone noticed an abnormal trend at a high level in a particular market, and then it became a fire drill," Sterpka says. And those patients who needed help getting through the system? "We were missing people, and we did not have a solution to plug those gaps. In today's environment, it's presented in a couple of clicks," he says.
Humana hasn't calculated the ROI or total cost of ownership for its BI implementation, but Sterpka says monthly uses for Medical Metrics jumped from about 400 to nearly 900 between January 2005 and June 2006.
Pratt is a Computerworld contributing writer in Waltham, Mass. Contact her at email@example.com.