Weekly Wrapup: Bringing AI and predictive analytics to health insurance

Register now

The health insurance sector tends to move at a different rate than other insurance lines, due to the peculiarities of its dominant distribution model and status as a political football. But like all other insurers, health payers can reap the rewards of digital transformation and emerging technologies.

With healthcare costs perpetually rising, anticipating headwinds and challenges is crucial to the sector. The Society of Actuaries recently released a report on predictive analytics in the healthcare sector that finds more than four in five health insurers planning to use the technology within the next five years; about half of those respondents already do.

But health insurers also are less bullish on the total amount of savings that could be realized with effective predictive analytics rollouts – they were half as likely as providers to expect budget savings of 15% or more. And, headwinds to expanding use of the technology remain; most notably talent issues.

“We may be seeing a decline in payers’ use of predictive analytics this year because they don’t think they have staff with the right skills to analyze large data sets to spot trends and draw conclusions,” writes Lilian Dittrick, fellow of the Society of Actuaries. “However, actuaries have the training and education to unearth insights that solve healthcare industry challenges, and can help payers become more comfortable with predictive modeling.”

About 15% of SOA survey respondents see machine-learning techniques as an eventual outgrowth of predictive analytics work. That dovetails with the assertions of a second paper, in which Accenture says that if payers effectively leverage AI, the industry could unlock as much as $7 billion in costs.

“Technologies such as robotic process automation (RPA), intelligent automation, virtual agents and machine learning can be incredibly transformative, freeing up resource capacity to redeploy toward more strategic functions under a potentially self-funding model,” Accenture writes in the The “Intelligent Payer: A Survival Guide” report, citing three places for health insurers to begin realizing benefits from AI:

  • Anticipating customer questions: “Applying advanced call analytics to deploy proactive outreach combined with automated communications can deflect the potential influx of avoidable calls, while improving overall satisfaction by anticipating the needs of customers and members.”
  • Improving benefits loading: “A benefits-capture utility can be developed using natural language processing (NLP) and RPA to simplify and validate benefits data entry in the field that then informs the structure required in the claims system.”
  • Accelerating authorization and claims review: “Intelligent automation and virtual agents can streamline the intake of information associated with initial steps of eligibility/prior authorization requests, allowing agents to focus on more complex cases. Machine learning can be applied based on utilization management trends to enable efficient clinical reviews.”
For reprint and licensing requests for this article, click here.
Health insurance Predictive analytics Predictive modeling Machine learning Artificial intelligence RPA