Newsletter Sign-up

As with any emerging technology-based process, of course, predictive analytics keeps improving over time and with use. And for risk managers around the country, that's good news. Best of all, predictive analytics applications are starting to show up in some interesting new places, including being embedded within Risk Management Information Systems (RMIS) -- a new twist on the predictive analytics front.

At the Dean Foods Company, a $12 billion Dallas-based global food and beverage provider, using predictive analytics tied to workers' compensation claims, for example, has resulted in a major success in just a little over a year. According to Jo Harris, vice president, risk management for Dean Foods, the basic concept was to find a way to predict which claims would benefit most from early intervention -- a major objective for employers, especially those who self-fund their workers' compensation risks.

"It all started when Marsh and I had a conversation two years ago. We were looking to drive down our cost structure but also help employees get healthy and back to work as quickly as possible," Harris said. "I told them we had all this data, but we needed a better way to take advantage of it."

The Dean Foods Risk Management department has finite resources and needs to do more with less (a common denominator across risk management today with economic pressures mounting). So the department sought a cost-effective solution that gave it quick, easy access to solving its pressing need to get injured workers back on the job.

In March 2010, Dean Foods began working with CS STARS and Oliver Wyman to do a complete evaluation of workers' compensation claims. The goal: to uncover potential high-cost claims that would be candidates for early intervention. In this case, the predictive analytics modeling tools were baked right into STARS?, the CS STARS risk management platform Dean Foods currently uses.

"As we applied updated claims to our database, STARS added a layer of intelligence," Harris explained. "This way, we have a listing of those claims that typically, without early intervention, would have higher costs. With that data, we could improve our decision making."

Ron Fowler, a principal at Oliver Wyman, a global management consulting firm who has worked with Harris and her team, says initially they thought it could solve the problem by identifying high-cost claims, but quickly learned from Harris that Dean Foods had something else in mind.

As noted, Harris wanted to give its claim intake managers an improved opportunity to get injured workers back to work as quickly as possible. While Dean Foods was averaging 4,000 claims annually, its two intake managers were each only able to review 600 claims over that time period, clearly a problem in that potential savings were slipping through the cracks.

"We used predictive models to figure out the 1,200 claims they should be handling, not the 2,800 they shouldn't be handling" Fowler said. "Instead of predicting large claims per se, Dean Foods wanted us to predict which claims should not be scrutinized - in terms of potential high-cost, early return-to-work situations. Their claims handlers received the information and set up the action items."

Fowler noted that the CS STARS/Oliver Wyman predictive analytics platform also can ferret out the high-cost claims as well, but in this case, that wasn't the client's main objective.

"We use STARS analytics to predict which workers' compensation claims don't appear to be high risk," he said. "That way, clients know early in the life of the claim what claims to target and what actions to take. It really is an optimal approach."

Fowler won't get an argument from Harris. Instead of sifting through 100 percent of its comp claims, Dean Foods, the nation's largest diary producer with 25,000 employees worldwide and 40 different brands, now needs to closely examine just 30 percent of the claims singled out by the CS STARS platform to get improved results.

"Even with that, every case isn't a candidate for early intervention," she said. "We try to make a difference within the company by being smarter with the way we execute our processes."

"Predictive analytics is well established these days in risk management," says Scott Robinson, principal and CS STARS' in-house analytics expert. "But the potential for predictive analytics is just coming into its own. CS STARS really focuses at the granular level within claims, providing the ability to predict with good accuracy a high-cost claim the second it comes in the door."

Robinson says the primary goal is keeping it simple -- to embed predictive analytics into the equation so clients can make better front-line decisions and direct attention appropriately to high-cost claims.

"In most cases, risk managers are working with scarce resources," he said. "With predictive analytics, we narrow our client's focus on to a smaller percentage of claims that are most likely to grow out of control. This allows them to review fewer and impact more."

Robinson added that while carriers and TPAs have been using predictive analytics for some time now, by integrating these concepts into its RMIS offering, CS STARS is bringing the benefits to the organizational level for employers who are self-administering claims, or for those who want to keep an eye on potential problems for claims managed by their TPA.

Whether a risk department is using CS STARS to administer their claims or to consolidate data from multiple sources, he explained, you can still apply these analytics to get a complete organizational picture of claims activity.

"By embedding predictive analytics into a RMIS, the organization is able to apply a consistent methodology to all of its losses," he said. "And being the largest RMIS provider allows us to focus our predictions on an individual client, or to aggregate data to provide macro trends."

At Dean Foods, there is little doubt that using predictive analytics as part of its RMIS is having a positive effect.

"It has resulted in an amazing ability to connect the dots with these early intervention cases, and has been critical to our success," Harris said.

How critical? Harris reported that since using the CS STARS/Oliver Wyman solution a little more than a year ago, Dean Foods has experienced a 20 percent reduction of loss costs in workers' compensation.

"This is a substantial cost savings and a result of collaborative efforts among leadership, operations, safety, continuous improvement and risk management," Harris said. "Predictive analytics is a major factor in helping risk management contribute to the overall success and engage our business units to do the right thing for our employees."

(The above piece is part of our continuing Perspectives series designed to highlight key products and services to our readers. This paid-for Perspective was written and edited by Risk & Insurance® on behalf of our marketing partner. Additional Perspectives can be found on our Web site at www.riskandinsurance.com/.)

March 1, 2012

Copyright 2012© LRP Publications