Sophisticated modeling techniques are used to forecast everything from consumer demand to weather. Predictive modeling is a concept that has gained increased attention in the health care arena of late, and is being applied to managing the business of healthcare. Providers and payors use predictive modeling to identify risk for individuals and populations and to target programs and interventions.

Many approaches are packaged under the term predictive modeling. In evaluating the different options, each organization has to look for a model that best fits their business needs. Unfortunately, most models are built with a one-size-fits-all approach. People are different sizes, and the assorted parts of a health care organization have different needs.

• Disease managers need tools to identify members with chronic illness and target interventions based on level of severity.

• Case managers require alerts to identify and intervene when patients are sickest.

• Underwriters demand models that incorporate group turnover, renewal timing, member terminations and new enrollees.

• Network managers seek strategies to measure patient severity based on diagnoses and co-morbidities rather than physician-driven utilization measures.

BENEFITS

• Maximize intervention efforts using the most accurate high-risk methodology available
• Improve operations and set competitive premium rates with superior medical expense forecasting
• Increase patient and physician satisfaction by providing clinically-relevant, actionable information
• Rely on models "fitted" to your business environment
• Realize the benefits quickly; Risk Navigator Clinical can be installed in less than 10 weeks
  

STEERING CLEAR OF DATA PITFALLS

The first step to better predictions is clean data. Our intelligent data cleanup engine has a library of edits to help scrub out the buried inconsistencies. Then, Risk Navigator Clinical uses your data to build predictions. Two years of data is used to construct, refine and test the models. This means that our predictions are based on the distinctive nature of your population, provider network and benefit structure. Because it's your model, data characteristics and inconsistencies are accommodated within the predictive engine. Now you can rely on the results even if your data is less than perfect.

LAYING THE COURSE TO USEABLE INFORMATION

Risk Navigator Clnical presents your results in an easy-to-use web interface. A few mouse clicks and you have generated the list of members with diabetes and have a profile of Impact Opportunities areas that provide the greatest potential for intervention and management. Standard reports present clinically-relevant data about expected inpatient, pharmacy, and ER use, along with total expense projections. Detail data is presented in a COGNOS PowerPlay ® cube for more extensive ad hoc reporting.

READING THE STATISTICAL MAP

Risk Navigator Clinical delivers the best and most useable results. Predictive models are typically evaluated on the accuracy of their results at an individual level (R² statistic) and at a population level (sensitivity and specificity). Using "real" data without eliminating outliers (the very events you want to predict) Risk Navigator Clinical consistently yields R² values of .25 - .32. For members in the top 10 percent, Risk Navigator Clinical's sensitivity and specificity are 50% and 94%, respectively.

As good as those values are, we need to look beyond statistics. Risk Navigator Clinical excels in identifying the "Movers," i.e., members with current low utilization levels that will incur high cost in year two. The predictions are based on clinical factors, diagnoses, Symmetry Episode Treatment Groups® (ETGs®), Rx use-and other risk markers- timing and frequency of procedures, insurance type, etc. Our risk models are based on a powerful prediction engine, Multiple Intelligent Tasking Computer Heuristics "MITCH," that uses the right combination of non-linear models to correctly "fit" the characteristics of your population. The value of improved accuracy can be traced right to the bottom line through improved management, more focused programs and better resource allocation.

ABOUT MEDAI

PDF Version

For more information on this product and pricing contact us:
Phone: 800.446.3324 -or- email: sales@medai.com