MANCHESTER -- Standing in her Commercial Street office, Heather Lavoie looks out across the Merrimack River and chuckles.
From her desk at Geneia, the Manchester native can see the West Side neighborhood where she grew up. She attended Queen City schools before embarking on a 30-year journey to her current post in life -- president of a growing health-care technology company that uses artificial intelligence to predict future health ailments and disease-related costs.
“I worked incredibly hard to move across the river,” jokes Lavoie.
Geneia is headquartered in Pennsylvania, with 50 employees in the Millyard -- about half of whom reverse-commute out of Boston.
In 2017, U.S. health-care costs were nearly $3.5 trillion. By 2026, costs are expected to increase to $5.7 trillion, representing nearly 20 percent of the economy.
Traditionally, health plans have used actuarial models that determine risk and future costs at the population level.
The Geneia Data Intelligence Lab (GDI Lab) is using machine learning techniques to create a "risk score model" that enables health plans to act on the insights to prevent health deterioration and future costs.
Lavoie said the growing use of artificial intelligence represents the biggest change she has seen in her industry over the last 30 years.
“We’ve created a factory, a pipeline, for development and deployment of artificial intelligence models, predictive models,” Lavoie said. “Before you may not have had the infrastructure to process through such large volumes of data, and now the systems are such that it’s easy to contain, manage, process through that much data. The tooling is better in terms of being able to process through that data, and it creates a more rapid cycle time for learning. There’s no way that a human with standard algorithms can make sense of so many different factors in the way that artificial intelligence can, so it allows us to be much more predictive.”
Lavoie said her firm is developing methods to use genomic data to predict the onset of disease and other medical issues.
“We have ones that predict diabetes,” Lavoie said. “One in three people in our country are pre-diabetic, and one out of every $7 we spend is on diabetes, and so it’s a significant issue that is bubbling up. Self-funded employers, commercial employers, they think about oncology as one of the most significant cost-drivers for their organizations. But diabetes is about to hit, because it’s getting into younger and younger populations -- not just the elderly -- and because of the prevalence. So we’re spending a lot of time trying to predict that and predict complication rates, so that we can intervene sooner.”
Geneia also has models that predict opioid addiction, Lavoie said.
“The team has a model that they’ve been testing out," she said. "We prove them out for a long period of time before we commercialize it, but they’re ready to commercialize it, and they’re thinking by Q1 of next year that it will be ready.”
Lavoie said Geneia’s models are proving to be highly correlated with actual addiction.
“It’s obviously a sensitive area, but it’s a critically important area to understanding based on prescribing patterns and other patterns who has a stronger likelihood to be addicted if then prescribed, so then being able to alert prescribers in advance of them providing medication. Not necessarily looking right now at the genetic side of addiction, but rather on the behavioral side and being able to provide it as a tool for prescribers.”
Lavoie has led startups, health plans and provider organizations during her 30-plus year tenure in health care. She has directed initiatives for private and public organizations on strategy, transparency, diversification, product innovation, operations and analytics. Previous to Geneia, Lavoie co-founded and served as vice president of product development, delivery and engineering for Choicelinx Corp., through its successful exit to CIGNA Health Care. Lavoie is a graduate of Notre Dame College and received a master of business administration degree from Southern New Hampshire University.
Sitting in her third floor office, Lavoie says statistics suggest she shouldn’t be where she is today.
“I grew up on the West Side, over by the Workmen’s Club, so not on the fancy part of the West Side,” Lavoie said. “My mom was 17, my dad was 18 when they had me. They worked in the shoe factory. My dad was a leather cutter, and my mom worked in the factories as well until they left, and then they had to reinvent themselves from there.”
The Adverse Childhood Experiences (ACE) test is a tally of different types of abuse, neglect, and other hallmarks of a rough childhood. According to the ACE study, the rougher your childhood, the higher your score will likely be, and the higher your risk to develop later health problems.
“I score pretty high,” Lavoie said. “I had a difficult childhood. My grandfather committed suicide when I was 12, and I was there. I can answer about eight out of 10 (on the ACE score), so that puts me on the rate of really high correlation with chronic disease. What you see with kids like myself, you kind of go one of two ways -- you can really have a significantly challenged life or you may go in the other direction and just really work like heck and focus on high performance. Thankfully I went in the other direction, but it’s a bit of a flip of the coin. I fortunately had great teachers at West, great teachers coming up and a great community there.”
Lavoie said the best advice she can give anyone, in their careers or in life, is embrace the zigzag in life.
“It (life) just doesn’t follow a linear path,” Lavoie said. “There’s times I made a lateral move instead of just an advancement. Sometimes that makes you more well-rounded.”