DMAA Helps Buyers Assess Care Management Predictive Models

WASHINGTON, D.C.—The Disease Management Association of America (DMAA) this week released a comprehensive, practical "buyer’s guide" to assess the value of predictive models in care management settings.

Predictive modeling tools, used in care management to predict high cost, high-risk cases, offer tremendous value for directing and prioritizing resources. The tools are particularly effective when linked harmoniously with a care management program’s specific interventions and desired outcomes, DMAA says in its new guide.

"In care management, it is important to predict high cost cases, but it is also beneficial to predict high cost cases that can benefit from the specific care management program," the 200-member trade association says in the guide, which it formally released at its 8th Annual Disease Management Leadership Forum, in Denver.

With this perspective, the DMAA "Predictive Modeling Buyer’s Guide" reviews how to assess the value of predictive models in care management settings and describes how assessment of "impactibility"—being amenable or receptive to intervention—strengthens predictive models.

"This guide will give health plans, physicians, hospitals, governments and many other decision makers a solid framework and easy-to-use tool for understanding and assessing the value of predictive models in care management settings," said Michael Cousins, PhD, chair of DMAA’s Predictive Modeling Subcommittee and vice president, health informatics consulting, at Impact Makers Inc., a non-profit information technology and informatics company.

Areas of focus in the Guide include:

• reviewing different modeling tools and their applications;
• describing how models can be assessed for impactibility and statistical performance;
• defining and understanding the practical implications of impactibility in care management modeling; and
• suggesting what to consider when making build versus buy decisions.

The Guide also offers an index of selected peer-reviewed publications and white papers, a cross-reference to the predictive modeling terms in the DMAA "Dictionary of Disease Management Terminology," a list of companies that use or provide predictive modeling services, and a buyer’s guide checklist.

"A framework for evaluating predictive models and an opening of the ‘black box’ will have practical value to users and, ultimately, decision makers and patients," Cousins says, commenting on the working assumptions behind the guide’s development. "We also believe there is a common process to evaluate care management predictive models that is independent from specific care management applications."

Cousins said the guide is a good start and early reviews have been positive. He added that the guide will get better with time. "The DMAA is committed to this being a dynamic body of work that will evolve as we learn more and readers provide feedback," he said. For information on obtaining a copy of the guide, contact DMAA at (202) 737-5980 or dmaa@dmaa.org.

# # #

About DMAA
The Disease Management Association of America (DMAA) a non-profit membership association, represents all stakeholders in disease management and care coordination. DMAA promotes the role of disease management in raising the quality of care, improving health outcomes and reducing health care costs for individuals with chronic conditions. DMAA has more than 200 corporate and individual members representing all aspects of disease management and care coordination—from large health plans, disease management organizations and employers, to individual physicians, researchers and nurses. Learn more by visiting DMAA online at www.dmaa.org.