PBM uses predictive modeling to identify early intervention needs

The marriage of art and science is helping companies identify and intervene in claims involving opioids (and other prescriptions) that are likely to turn into high-cost, chronic cases with the potential for medication misuse and/or abuse.

Developers believe it could be a major game changer in the effort to curb prescription overutilization, misuse, and abuse, and with it the unnecessarily high pharmacy costs of some workers' compensation claims.

"We've built a model that can predict the pharmacy costs associated with a claim," said Joe Anderson, director of analytic services for Progressive Medical. "It's interesting to see how often we can change the future path of a claim's pharmacy spend."

Program measures effects of interventions (11/12/12)

CWCI: Compound drugs are more potent, costly (03/11/13)
State legislators to address controversial meds in workers' comp (03/04/13)
NCCI: Physician dispensing on the rise in most states (10/06/11)
Florida Legislature targets repackaged drug costs (01/30/12)

Armed with the knowledge that early narcotic use is a driver of workers' compensation costs, the company has built an analytical model Anderson says can accurately identify creeping claims before they become catastrophic that would otherwise remain hidden. Claims professionals are notified, and in some cases, Progressive Medical's clinical staff intervenes. "And that's where it appears to really get interesting, as early results come in from a pilot program currently under way to determine the effectiveness of various interventions," he said.

How it works. The model involves sophisticated analyses of clients' data. But the developers say it's not just the numbers that are important -- it's making them meaningful.

"It's becoming a very important tool for our staff," said Tron Emptage, chief clinical officer for Progressive Medical. "Don't treat the numbers, treat the patient. Use the data to help you."

The company's clinical team gets involved after the data has recognized a particular case that would likely involve prescription misuse. "By using the model's output the clinicians are able to recommend interventions earlier than previously possible," Emptage said. "Because we are able to gain control earlier, we achieve a better outcome."

Analyzing pharmacy transaction data helps the company find claims with both high cost and high narcotic use before they actually occur. "I think as we move from identifying which claims have had high pharmacy cost to identifying which claims will have high costs in the future, it changes our relationship with our customers," Anderson said. "It changes the payer/ PBM relationship."

Looking through the company's book of business over several years has revealed many claims that did not turn into high-cost, high opioid use claims. Looking at basic triggers alone can be misleading, Anderson explained.

"From a data perspective some claims contain elements indicative of high-cost future pharmacy spend; focusing on those elements individually can lead to the wrong course of action," Anderson said. "The key is finding those claims that can be effectively impacted by early intervention."

As Emptage says, it's finding the needle in the haystack of claims that can be affected by a change in direction.

Use of the model. The program has been under development for more than a year and a half. The analysts say they are learning more each day, especially in reviewing claims over the past several years. "When we look at the top 100 pharmacy claims, the majority of the time there is something we can do to change the outcome of the claim," Anderson said.

Progressive Medical's clients are using their programs to varying extents -- some partnering with the company for its interventions, others simply relying on the information.

"Each client uses data and data analytics differently," said Sarah Berger, vice president of marketing. "We always keep that in mind. The more data we can get, the stronger the analysis and the better outcome we achieve."

In addition to information from clients, the company is using internal data. The information increases the ability to find which claims will be high cost for pharmacy.

The company believes such in-depth data analyses bode well for the workers' comp system. "Pharmacy is an early trigger for overall workers' compensation claims costs," Anderson says. "We're in a great position to identify claims earlier as a result of our network penetration that captures the pharmacy transaction from the first fill. We also understand which claims will get pharmacy fills early on. And we have a clinical team positioned to act sooner. "

The company also believes the model may ultimately help curb the rate of medication misuse in the system -- and save costs.

"It will definitely make us smarter, better, faster," Emptage said. "Our goal with learning more is not just to send an alert to an adjuster, it's to help us [act] more appropriately and find the right time to intervene. It's been proving itself already. We learn more everyday and get better every day. Will it help overall? I certainly believe so."

Read more at the WorkersComp Forum homepage.

November 12, 2012

Newsletter Sign-up