How to leverage data to predict the 5 percent who drive the most health cost

Many employers receive an annual report from health plans listing high-cost claims. Most find themselves looking in the rearview mirror wondering what they could have done to prevent such costly episodes or at least reduce their impact.
This impact goes far beyond the rising costs of health care. A HCMS Group study found the top 5 percent of cases drove over 50 percent of overall spending when absenteeism, disability and workers’ compensation were included. Fortunately, it’s possible to leverage integrated data to predict which individuals are headed toward this top 5 percent, allowing employers to offer uniquely targeted clinical prevention.
But the stakes are high and rising. The HCMS Group’s Research Reference Database, covering 3.8 million members, revealed that the number of cases costing at least $100,000 each year has climbed 16 percent since 2015 and now averages five per 1,000 members. Cases costing $500,000 or more per year have increased 22 percent over the same time period.
High-cost cases happen to high-risk individuals — those with multiple conditions who often deal with family and work issues. They typically consume health care, lost time and disability benefits at a much higher rate. This comes at a price to their health and to their employer’s bottom line. Many individuals stay in the top 5 percent cost range year after year with multiple family members in the top 5 percent at the same time.
Smart Business spoke with Justin Schaneman, MS, vice president of Data Analytics, and Rene Sims, MSN, vice president of Clinical Services at HCMS Group, a WorkPartners affiliate, about how data can predict your top 5 percent, which is the first of two articles on leveraging integrated data and applying it to clinical interventions.
How is data being used to better predict this risk?
Many employers have gained access to data warehouses, hoping to generate actionable insights from mining big data. The value, however, can be limited when only medical and pharmaceutical expenses are included.
Integrating a broader array of data produces more actionable findings. Expanded databases incorporate disability and absence data, which reflect the policies that drive people’s behavior and consumption of benefits. They bring into focus the full cost of burden of disease, illnesses and injuries, allowing employers to better harness the predictive power of their data. Sophisticated employers also include the results of employee performance evaluations, compensation levels, safety violations, training course completions and other person-centric data types.
Once the right data is being collected, how can it be applied to the health plan?
The next step is to fully leverage the value of that broader set of data inputs. An integrated risk index that takes into account compensation and lost time data will dramatically increase the predictive power of the index. This can measure risk related to human capital health and job performance, not just disease. It also provides a leading indicator to flag emerging risk cases early in their trajectory, so interventions can shorten or prevent the period of time when they are both high-risk and high-cost.
Leveraging integrated data on workers’ compensation, disability and incidental absence — in addition to medical and pharmaceutical claims — allows predictive, individual risk indices that enable timely intervention for the high-risk top 5 percent.
This kind of risk index can provide early identification of individuals who will soon be both high-risk and high-cost. Clinical intervention can then be offered to those who need it most.

As big data opportunities continue to expand in the workforce health arena, advanced analytic and predictive modeling approaches are becoming more relevant and can improve outcomes for both the employee and employer. The key is to partner this predictive power with a clinical service that can leverage that intelligence effectively, targeting individuals who will benefit from the service. The next article will discuss how to set up that clinical intervention.

Insights Health Care is brought to you by UPMC Health Plan