How can using analytics help an employer manage health care costs?
In the past, employers have managed health benefit plans by looking at a limited set of aggregate financial measures, such as average cost per employee, utilization rates for inpatient, ER and other services, and prescription drug patterns. These measures obscure the fact that health care costs are a very skewed distribution of expenses.
Typically, 3 percent of covered individuals will account for 55 to 60 percent of total medical expenses in a year. In order to manage wellness and health management programs, employers must look past measures based on average costs and instead identify the costs and use patterns of specific subgroups, based on their health status and behaviors.
For example, diabetics will typically have substantially higher costs than average, but the variation of cost among diabetics is also significant. In 2009, an Aon Consulting study showed that the cost of diabetic care could vary by a factor of five times, depending on the severity of a diabetic’s condition and the additional diseases.
Effectively managing diabetics requires that employers look past average measures and identify those members who are at greatest risk and who could benefit most from care management services.
For employers seeking to improve the performance of their disease management vendors and programs, it is essential that they have access to the analytics needed to identify the level of health risks among diabetics.
Predictive modeling analytics allow employers to identify those with the greatest health risks and match disease management program resources to individual needs. Predictive modeling provides a way to measure the severity of health risks across the group and track the outcomes of wellness and disease management programs. By providing a clear measure of health risk, predictive models allow employers to evaluate the performance of their health management vendors and strategies over time.
The more sophisticated predictive models also incorporate measures to track an individual’s compliance with appropriate clinical treatment guidelines. For example, the recent study of diabetics showed that costs vary not just by the severity of an individual’s condition but also by the level of compliance with clinical care guidelines. For individuals with comparable health risk, costs can vary by 100 percent or more depending on compliance.
From an employer perspective, analytic tools that can track both health risk and compliance are a critical means of managing long-term health costs.
How can an employer tailor its approach to individual employees?
Perhaps the most significant difference of the new data analytics is the recognition that the data needed to manage health programs goes far beyond the traditional analysis of paid claims. While claims data will always be an essential part of any health analysis, it is limited in several ways.
Employers need to expand the data used to distinguish different groups of employees. For example, studies show that compliance with clinical care guidelines can be highly correlated with pay level. Likewise, studies indicate that health care behaviors are significantly different across ethnic groups.
Wellness and disease management programs need to acknowledge the differences and tailor their approach in ways that will most effectively reach individuals across various ethnic groups. For employers, being able to use analytics that clarify the impact of wellness programs on specific groups of employees is critical to being able to manage these vendors and programs.
In short, as employers move to more sophisticated programs to manage health care costs, they need to expand the range and sophistication of their data analytics.
John Boss is an executive vice president in Aon Risk Services’ Health and Benefits Practice. Reach him at (317) 237 2411.
