Currently Viewing:
In Focus Blog

Deploying Data to Prevent Opioid Abuse in At-Risk Patients

Mary Caffrey
Developing a predictive model can identify which patients could be at risk of developing an opioid use disorder and allow health systems to prevent this from happening.
Nearly 50,000 Americans died from opioid use in 2017, and CDC data estimate that 40% of these deaths involved prescription drugs. The nation’s crisis arose after prescription opioids became more widely used in the 1990s, and the problem has exploded over the past decade as many users moved over to heroin when they lost access to medication.

The challenge for health plans is 2-fold: How can payers work with providers to manage patients already dependent on opioids? And, more importantly, how can payers identify those at risk of becoming addicted and keep these patients safe?

Data can point the way, according to Elizabeth Ann Stringer, PhD, the chief science and clinical officer for Axial Healthcare, which offers solutions for payers and health systems that want to avoid the mistakes of the past. In an interview with The American Journal of Managed Care®, Stringer explained how the company is gaining insights from healthcare claims that can guide better care and prevention.

As Stringer explained, Axial Healthcare was founded in Nashville, Tennessee, in 2012 by a pair of pain management specialists who saw that too many patients being treated with opioids were not getting better. “Their health was deteriorating over time,” she said.

Axial Healthcare partners with major aggregators of patient information—from health plans to state agencies—to gather data and learn more about what puts patients at risk for opioid addiction, Stringer said. In doing so, the company has generated data to combat myths about opioid use and developed an algorithm that can predict which patients are at risk of becoming addicted, which can allow payers and providers to target services to this population.

“So often, providers think that their patients are very stable on their long-term opioid use, and they diagnose these patients with opioid dependence,” Stringer said. “And they think there is a very large difference in this patient population from those that are in the active throes of addiction,” which includes seeking out medication and misusing it.

However, Stringer said, when Axial Healthcare scientists examine the levels of healthcare utilization by those diagnosed with opioid dependence, compared with those diagnosed with opioid abuse (based on International Classification of Diseases codes), they find the levels of healthcare use are similar. “We don’t know what drives those patterns of high utilization,” she said. “We need to take the next step to dive deeper into the analysis to understand why these patients are looking similar.”

But the key takeaway for providers is that long-term opioid users may not be as stable as their providers believe.

Predicting Opioid Abuse

Using data gathered from multiple populations covering 5 million lives—including Medicare and Medicaid beneficiaries, as well as those with commercial coverage—Stringer said Axial Healthcare has been able to develop an algorithm that can predict which characteristics make a person likely to be diagnosed with an opioid use disorder in the next 12 months.

“We looked a number of different variables that might drive a patient to opioid use disorder,” Stringer said. “There’s a little less than a 1000 different variables that we were able to bring into our large algorithm, and we used 80% of that information to start to train the model… We found around 400 of those variables were very strong at helping us predict who was going to traverse into opioid use disorder.”

Axial then used the remaining 20% of the patient population data to validate the model, she said. “This lets us not only understand not only patients who are currently using opioids who might be high risk (of developing opioid use disorder), but even those patients who are not currently receiving opioids,” who are at risk of developing a disorder.

“We all know that once a patient moves into opioid use disorder, it’s obviously a lot of healthcare utilization and high expense, but opioid use disorder is a chronically relapsing disorder that a patient is going to have to manage for the rest of their life.” If that condition can be prevented, “We want to be able to do that.”

Help for Veterans

Stringer praised the Veterans Health Administration for its partnership in helping understand how opioid use disorder affects the military population. She said that higher rates of abuse among current service personnel and veterans are quite simply due to the injuries and stress many have experienced in the line of duty. “One of the reason for this is the stress and exposure that most every day Americans are not exposed to,” she said. Axial’s model shows that behavioral health conditions are the best predictor of who will develop a use disorder, so it follows that veterans would be at risk. But Stringer said an additional risk factor are the social challenges veterans face.

“As a society, we don’t always do a great job of socially supporting our veterans when they come back,” she said, from tours of duty.

As the opioid crisis evolves, Stringer said the way physicians are educated in pain management is evolving, too. “Pain is a symptom that something else is wrong with the body,” she said, and today medicine is moving away from the idea that just treating the pain is always safe and low risk.

Cost-effectiveness in Care

Opioids are often generic, and so they seem cost-effective, Stringer said, until one considers all the downstream costs that can follow from overutilization of healthcare.

Health plans have sometimes been reluctant to pay for medication assisted treatment (MAT), including buprenorphine in different formulations or combinations with naloxone. MAT can be expensive alongside behavioral health care, but payers are wise to invest. “When patients are treated in this comprehensive manner, this actually decreases costs, both in the long run and in the short run,” Stringer said. “When patients do not receive this comprehensive treatment, this increase rates of utilization at a hospital, which is completely unnecessary.”

In the current public health crisis, it’s essential that payers and providers not try to scapegoat one another, because this doesn’t serve patients. A key to Axial’s success at the practice level is support to providers to help individualize care. “We make sure we have a patient-centered approach, where we’re understanding that patient’s history and we’re understanding the current treatment to make suggestions for an improved care path in the future.”

 
Copyright AJMC 2006-2019 Clinical Care Targeted Communications Group, LLC. All Rights Reserved.
x
Welcome the the new and improved AJMC.com, the premier managed market network. Tell us about yourself so that we can serve you better.
Sign Up