Weekly, if not daily, articles are published and posted across the internet hailing the potential of artificial intelligence in healthcare. Much of this content focuses on two primary aspects of the overall healthcare ecosystem: clinical and administrative.

However, one healthcare tech company thinks it has found a unique way to use the power of AI to enhance care access to an often overlooked and underserved group.

Since big data and machine learning became ubiquitous terms, forecasts have extolled the potential of artificial intelligence in healthcare. There are exciting and futuristic concepts from a clinical and diagnostic perspective, like earlier diagnosis and treatment.

Some of those are starting to culminate, such as Google’s DeepMind, which was recently used for groundbreaking research at London’s Moorfields Eye Hospital.

AI is also expected to dramatically increase efficiency and accuracy on the administrative side of healthcare by automating core administrative functions. Things like anticipating and answering patient questions, speeding up prior authorization and claims reviews — these routine tasks are ideal for data-hungry machine learning.

But Change Healthcare, a tech company that works with providers and payers to build a more collaborative and efficient healthcare system, has identified a different kind of use case. It has implemented artificial intelligence to help Medicare Advantage precisely identify and increase the chances that people who qualify for dual enrollment in Medicare and Medicaid get the care and benefits to which they’re entitled.

According to Change Healthcare, 34 percent of the 58.5 million Medicare beneficiaries live at or below the federal poverty level, indicating many of those could also be eligible for Medicaid. Given its wealth of data from healthcare transactions across its business activities, Change Healthcare saw an opportunity to target and engage those most likely to qualify for dual enrollment and help guide them through the process.

"When the business model aligns and the technology aligns, the use case aligns, and you have sufficient data, then it’s really applicable and a good use of AI not just to improve the quality of care but improve the access to care," Change Healthcare Chief AI Officer Nick Giannasi said.

Prior to big data, health plans had to identify dual-eligible candidates through manual programming. The use of demographics and psychographics could identify potential candidate groups, but the process was inefficient and likely to suffer from unintentional bias. Furthermore, limited resources meant it was possible to reach out to only a fraction of these candidates.

"It was a bit of a crapshoot," Giannasi explained, pointing out the labor-intensive and inefficient use of calling, emails, ad campaigns to reach this market.

But thanks to an increase in compute power, Giannasi said it’s now possible to take a list of 20 million people and, with a much higher degree of certainty, identify only the most likely to qualify.

Giannasi said that increased compute power also will help fuel the second part of this concept, which is to add varied content delivery that matches all the variations of individuals identified as candidates.

In other words, getting the right pieces of information to the right people at the right time to help inform them and guide them through the process of getting the correct coverage.

"It basically becomes a recommendation engine," Giannasi said, comparing it to the way Netflix or digital music services understand a user’s preferences to make recommendations on what he or she might want to watch or listen to.

Only about a month in, it’s still too early to determine just how successful the program, called Dual Enrollment Advocate, has been. The application process itself can take several months. Still, Giannasi said some leading indicators have increased dramatically and hint at a "pretty phenomenal result."

Change Healthcare hopes its AI program can help get insurance coverage for many older, poorer and sicker people, which benefits the patients as well as the health plans.

"We're applying data science, behavioral science, and consumer-experience design to solve one of the trickiest problems for Medicare Advantage plans, which is how to accurately identify people for Medicaid eligibility," Keith Roberts, VP of engagement at Change Healthcare, said.

"AI and machine learning can't do it alone. Behavioral science can't do it alone. And health plans and legacy systems can't do it alone. The time has come to bring these healthcare IT and scientific disciplines together to help solve a critical business challenge for our customers."

Dual Enrollment Advocate may not make headline news for a technological breakthrough, or become the poster child for AI in healthcare, but it’s an example of using today’s powerful computing potential for practical and purposeful use.

"It doesn’t sound nearly as sexy as doing a diagnosis of cancer with AI," Giannasi admitted. But it does illustrate the many different ways artificial intelligence capabilities can benefit the healthcare ecosystem.