Healthcare by the Numbers

Client data is stored in MDI's headquarters, which was made to withstand a Category 5 hurricane. [Photo: MDI]

Companies also can use the algorithms in MDI's software to analyze demographic information and claims data in order to forecast trends and flag high-risk situations. Employers end up with a better gauge of their employees' potential demand for healthcare services so that they can either budget for it or negotiate appropriate contracts with their insurers.

While MDI's prison-health segment continues to thrive, the company's new business lines are growing fast. MDI is the primary data analytics provider for Wells Fargo Insurance Services and competes directly with the largest nationally recognized analytics company in the nation, Boston-based Verisk Analytics. Other customers include First Coast Advantage, the Medicaid managed care program run by UF Shands, the Community Health Network of Central Florida and Parrish Medical Center. Evangelical Community Hospital in Lewisburg, Pa., recently hired MDI to set up its customer provider network and implement a healthcare analytics system.

The growth has boosted MDI's revenue to around $100 million in 2009 and its workforce to 170.

MDI is also finding applications for its products among healthcare providers who've adopted the "medical home" approach to healthcare delivery. The medical home is a team-based model of care in which one provider or organization — the patient's primary care doctor, for example — oversees and coordinates all physicians involved in a patient's care and ensures that the patient follows treatment regimens. "We did it with UF Shands and also an accountable care organization in South Carolina — we built a medical home for all the patients into our analytics. Then you can measure outcomes of the patients that are with that particular home," says Willich. "You can't manage what you don't measure."

Willich envisions even more possibilities for healthcare analytics, as analytical tools are integrated with electronic health records and language processing technologies, which enable computers to "read" medical records and apply what they have read to completing a task — for example, applying appropriate medical billing codes to records after reading them.

The combination of tools, he says, will enable "a wider proliferation of customized treatment simulations moving from not only emergency, oncology or other specialized care to a wider array of disciplines. It will be best-treatment practices for the specific individual and not just for a specific disease."