Help

BI’s Article search uses Boolean search capabilities. If you are not familiar with these principles, here are some quick tips.

To search specifically for more than one word, put the search term in quotation marks. For example, “workers compensation”. This will limit your search to that combination of words.

To search for a combination of terms, use quotations and the & symbol. For example, “hurricane” & “loss”.

Login Register Subscribe

Predictive models should be used to identify effective intervention strategies

Look for effective intervention strategies to hold down cost of workers comp claims

Reprints
Predictive models should be used to identify effective intervention strategies

While third-party administrators tout predictive models' ability to reduce workers comp costs, experts say employers should weigh whether such models can go beyond predicting claim outcomes and identify effective intervention strategies.

Models have become more adept at spotting conditions that typically increase claim duration or severity, experts say. They include comorbid health conditions, obesity, depression, taking opioid painkillers or attorney involvement in a claimant's case.

Now that TPAs have become better at predicting the correlation of such factors with higher-cost claims, experts say the next step for them is using the models to intervene in workers comp cases before trouble develops.

“Ultimately, prediction is of limited value,” said Russell Pass, chief information officer and executive vice president of product development for Gallagher Bassett Services Inc. in Itasca, Ill. “What we're really looking for is influence. We want to drive claims toward the best possible outcomes.”

Middle-market employers, in particular, should weigh whether predictive workers comp models are worth the expense, said Joe Picone, Richmond, Va.-based claim consulting practice leader for Willis North America Inc.

“If you're a middle-market employer (and) you have 100 claims or less per year, you probably don't need a state-of-the-art predictive modeling program,” Mr. Picone said.

%%BREAK%%

Experts recommend that employers use several strategies to evaluate whether a TPA's predictive model can truly reduce workers comp costs.

Nearly all recommend asking what interventions the TPA would use or encourage the employer to use to change the course of troublesome comp claims.

“Ask the party who's providing the data to give you some actionable things that they would recommend you do,” said Paul Braun, managing director of casualty claims for Aon Global Risk Consulting in Los Angeles.

Mr. Picone recommends providing TPAs with a year's worth of workers comp claims data to see if their models correctly identify workers comp claims that developed high costs.

If a model is successful at pinpointing troubled claims, he said, employers should ask what strategies the TPA would use to help improve claim outcomes. He also said a TPA should have several strategies, rather than using only certain tactics.

“If nurse case management is always the answer, it becomes very subjective what the true savings will be,” Mr. Picone said.

Predictive models used by Atlanta-based Broadspire Services Inc. use a combination of early discussions with claimants and data mining during a claim to determine if interventions are needed to help an injured worker get well, said Joel Raedeke, the TPA's Chicago-based vice president of consultative analytics.

%%BREAK%%

While both methods are useful, Mr. Raedeke said the firm has found information provided early in a case helped it customize return-to-work strategies for claimants.

“If you don't do the upfront direct capture, you're really missing out on a lot of visibility within the first few days, where setting the course of the claim is critical,” he said.

Cases pinpointed for intervention can see cost savings, Mr. Raedeke said. Claims that are given pain management help because of the predictive model get an average 27% reduction in medical costs, while claims that are assigned a nurse case manager gain an average 6.5% reduction in overall costs.

Early predictive models used by Gallagher Bassett were able to detect workers comp cases that could be fraudulent or develop high-cost severity, Mr. Pass said.

While such results helped Gallagher Bassett recommend appropriate reserves for such cases, he said they were less effective in recommending ways to change the course of a claim.

The TPA has worked to evaluate and refine its models to identify claims that have a wide range of potential outcomes that could be affected by certain interventions.

%%BREAK%%

As a result, Gallagher Bassett's claims adjusters now see a selected group of flagged claims that have the highest chance of benefitting from extra assistance, said Michael Hessling, Gallagher Bassett's chief client officer.

“We've got a much more narrowed, focused approach with much greater ability to direct activity and make a difference on claims,” Mr. Hessling said.

Experts agree that as workers comp TPAs continue to improve their predictive models, employers should continue to consider if they are effective in positively influencing the trajectory of comp claims.

“It's not just about being predictive,” Broadspire's Mr. Raedeke said. “That's not the end of the game.”

Read Next

  • Predictive models that can make cost-saving recommendations are future goal

    Predictive models used by third-party administrators typically capture a variety of data from workers comp claims, ranging from questionnaires injured workers fill out at the start of their claims to systems that comb through a case file for keywords indicating a claimant may have trouble getting back to work.