Travis Hoyt is the Chief Analytics Officer (CAO) at MOBE. Before MOBE, he was vice president of analytics at Citra Health Solutions. He also held leadership roles at Rise Health and UnitedHealth Group.
“Lots of organizations say they're data-driven and learning insights-driven,” says Travis Hoyt, MOBE’s Chief Analytics Officer. “But, for MOBE, data is a critical element in everything we do. We use it in a number of unique ways to fuel our business and help people get better health and well-being outcomes.”
1. Identifying MOBE participants.
One of the most important parts of MOBE’s service is identifying the participants who will benefit the most from the program.
“Most vendors look for a condition, like diabetes. They find the people with diabetes within a data set and ‘manage’ them. But health care is way more complex than that,” says Hoyt. “Just because people have a condition doesn't necessarily mean they’re well managed. And it also doesn't mean that diabetes is what they care about right now. Typically, people with diabetes have other chronic conditions, and diabetes might not be at the top of their list of worries. They might have housing insecurity or food insecurity. They might have concerns with work or family that are driving their day-to-day mindset. As a result, we stack several data models to create our identification process.”
Hoyt says, “The point is you can’t think of a population as a vertical silo. You have to look at the whole person and what’s going on in their lives. That's what we do differently. We use a combination of claims data and data that doesn’t come from typical sources. Rather than identifying participants based primarily on the presence of conditions, we focus on patterns in behavior and utilization. And that’s where the value is.”
2. Motivating participants to try MOBE.
While solutions that only look at one condition at a time have low engagement rates (around 4-6% of the target population), MOBE’s average engagement is about 30%. That’s because MOBE’s data team works with our marketing team to help increase engagement by finding ways to get people to join the program.
Hoyt says, “Once prospective participants are identified, we combine marketing strategies and data in unique ways to segment which people to reach out to. We provide messages that are timely and relevant for them based on what’s going on in their lives—reinforcing MOBE’s whole-person focus. We see how they respond to our initial messages and use data about their digital interactions to deduce their interests. And we use that information to keep them motivated and interested in learning more about what MOBE has to offer them.”
3. Helping participants get the most out of the program.
Once participants join MOBE, the data team captures a wide variety of data points to keep the program engaging and useful for each individual.
For example, Hoyt says, “When participants interact with their MOBE Guide or Pharmacist through the MOBE Health Guide app, we use data to help with retention and content curation—making sure that they’re getting the right information to keep them motivated and continue on their health journey.” He continues, “We also use the data to measure their outcomes. We do surveys. We get information from wearable devices—like fitness trackers and watches—to monitor important things like sleep, steps, and BMI. We measure those improvements for them to help track progress.”
4. Refining and improving MOBE’s services.
MOBE also uses data to refine and make changes that help the program become more effective.
“When we see participants improving, we look at what’s the causal relationship or the average treatment effect for those members on their cost and utilization,” Hoyt says. “That gives us the full picture of the population that needs help. We can see the cost reductions and start understanding where they come from. As a result, we can continually update our program to get even better outcomes for participants.”
According to Hoyt, “The ultimate goal is to move the line further and further upstream, where prediction and prevention come together. So, instead of playing whack-a-mole as a condition pops up, we’re trying to understand why the mole is in your yard in the first place. And, when we can do that, we can help prevent the issues from ever starting, instead of only addressing conditions people already have.”
“Our models are getting more and more sophisticated and smarter,” says Hoyt. “We already have a few predictive models at MOBE that we’re very proud of. We have a spinal surgery predictive model that looks for risks of imminent or high-risk participants for spine surgery. We also have one for large joint replacement. We continue to get more and more predictive about things, so we can move that predictive line closer to prevention.”
5. Proving MOBE’s value to health plans and employers.
Hoyt says, “When we’re talking to a prospective client—before we even start a contract—we offer to do a business case for them. We look at their existing health care programs, disease management, care management, wellness programs, and everything else they have going on. There’s never much overlap between MOBE and their existing programs.”
He continues, “Prospective clients can clearly see how MOBE targets a population that doesn’t usually engage in other programs. Other programs find people with chronic condition-based needs. So does MOBE. The difference is MOBE’s target participants continuously seek health care through the health care delivery system instead of engaging in disease/care management programs. But, when those same participants work with MOBE, we have industry-leading engagement levels.”
MOBE also provides compelling data around cost predictions. Hoyt explains, “We have a stacked set of machine learning algorithms that evaluate and predict cost risk. However, we don’t use cost as an input feature in our risk prediction models. We use behaviors, such as how our prospective participants engage with the health care system and the frequency and intensity of the services that they receive. We also look at probability risks, episodic triggers, and engagement with our program. Our algorithms are wildly unique.”
According to Hoyt, “We always have a ‘triple aim’ in mind. We focus on doing good for lots of people: our clients, the MOBE participants we’re supporting, and our own company. Our clients—health plans and larger employers—see a financial return and more productive employees. Our participants get help on their path to health and happiness. And MOBE gets to be a successful business, too. Usually, in a three-way exchange, somebody is a loser and that’s not the case here. MOBE’s business model creates positive incentives that are aligned across the board.”
Find out more about MOBE, including ways you can leverage data for better outcomes.
