Predictive analytics enables workers' comp insurers to predict future outcomes with unprecedented accuracy. Insurers and TPAs in all lines, however, can apply predictive analytics to mitigate risk more effectively.
The term "predictive analytics" is relatively new, but the concept behind it is as old as insurance itself: using data that describe past results to predict future outcomes. Historically, only the largest insurers have had the resources to apply advanced predictive models to their actuarial, underwriting and claims processes. Today, however, insurers of all sizes are using highly accurate, high-return predictive analytics applications to foster profitable growth, lower loss ratios and reduce expenses.
Consider, for example, the value of predictive modeling in mitigating risk in workers' compensation (WC) claims. Medical costs for WC claims are rising at an alarming rate, which gives insurers a strong incentive to identify as early and accurately as possible those claims with the highest potential for loss–and then manage them proactively.
Problems can arise, however, when basic, often imprecise business rules allocate senior adjusters and medical specialists to claims better suited for junior, less-expensive adjusters. Shoulders, back, and knee claims, for instance, are often targeted simply because, historically, they have the highest severities. Although such claims do tend to have the highest severity, not all of them are severe.
For this article, Valen Technologies analyzed more than 100,000 WC claims in the Valen Networks data warehouse. "Our analysis showed that nearly 70 percent of back injuries incur losses under $2,500," says Dan Bankson, senior vice president of analytics with Valen. "On the flip side, a mere three percent of the most-severe back-injury claims accounted for more than 55 percent of the total losses incurred from back injuries."
The back-injury phenomenon is replicated to a lesser but still significant degree by other types of injuries. These patterns raise important questions:
1] How can insurers identify high-severity claims early and then mitigate the risk effectively?
2] Which treatments and procedures are ineffective and only prolong the worker's return to work?
Predictive analytics is the answer. It helps insurers and TPAs identify these high-severity claims with unprecedented accuracy. By developing predictive models from the insurer's own data as well as data from vast warehouses of medical and claims information, insurers are scoring claims early in their lifecycle and better managing the ultimate outcome of claims that have the potential for higher-than-expected medical costs.
Naturally, insurers and TPAs are always looking for ways to improve their claim handling. By mitigating the risk on the small handful of claims that have the potential to become severe, an investment in a predictive analytic solution is a win-win for the insurer and the insured. A predictive analytic solution reduces ultimate losses, thus generating a return on investment measured in multiples; returns the claimant to work more quickly and with a lower incidence of recidivism; and, ultimately, enables the insurer to price more competitively in the market.
Taken as a whole, the resulting analyses empower insurers to identify the common characteristics of high-severity claims and allocate their senior loss adjusters, medical specialists and other resources to mitigate those claims that truly have the highest potential severity. At the same time, insurers reduce loss-adjustment expenses by shifting expensive resources away from less-severe risks.
"Medical claims are one of the fastest growing areas of focus for predictive analytic solutions in the insurance market. The opportunity to reduce settlement times, decrease the risk of litigation and improve the customer experience is driving market leaders to look for new predictive characteristics, streamline claims data capture processes and improve adjuster decision criteria," according to Mark Gorman of The Gorman Group, a research and consulting firm specializing in the adoption of business analytic solutions in the insurance market. "The possibilities are exciting."
For readers seeking further information concerning workers' comp claims predictive analytics benchmarking, please contact either of the article's authors.
(Becky Clegg, AIT, AAM, CPIW) is product director (iVOS), for Aon eSolutions. She can be reached via e-mail at becky.clegg@aon.com. Richard Vlasimsky, is vice president, market development for Valen Technologies, Inc. He can be reached at richard.vlasimsky@valen.com.)
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.