Could AI Present Solutions to Improve Plan Member Experience?

At the America's Health Insurance Plans (AHIP) Institute & Expo, a presentation focused on how artificial intelligence (AI) solutions can help payers deliver the seamless experience that consumers have come to expect.

Bobbi Coluni, vice president of payer offering management at Truven HealthAnalytics, part of the IBM Watson Health, acknowledged that the traditional business model for insurers is under increasing pressure in a world of near-constant acquisitions and consolidation. Adding to this pressure is the fact that consumers have become accustomed to a smoother experience and now have higher expectations for convenience in every interaction. Together, these demands create an opportunity to “leverage data and cognitive services to reinvent the customer experience at every touchpoint,” she explained, which is one of the aims of IBM Watson.

To illustrate how consumers expect more information but are often disappointed when it comes to buying healthcare, Coluni recalled that she had recently had a surgery that was performed in an operating room instead of the outpatient setting she expected, and she was shocked when she received a bill of $19,000. If that situation could happen to someone who had spent years developing a cost transparency system, she pointed out, it can happen to other consumers who are making care decisions without having all the information they need.

Another barrier for insurance plan members is finding the answers for seemingly simple questions like which drugs are covered, how to submit a claim, or what coinsurance is. When they reach out to their insurer’s call center, the agents themselves struggle to find the right information, as it could reside in any one of a plethora of documents and locations. Thus, the system’s complexity creates a “perfect storm” of frustration for both the insurer’s members and employees.

That complexity is why IBM Watson thinks that its AI could be used to improve the member experience, answer questions consistently, and support health outcomes through every touchpoint. Coluni provided a sneak peek of the new Watson assistant for health benefits, which can be used for self-service chats. For instance, a member wondering if a knee replacement surgery is covered would “ask” Watson, which can eventually provide an estimate of the total out-of-pocket cost to the member. By accessing and weaving together all of that information, AI can ideally “inform that individual and answer the questions they didn’t ask but we know they need to understand,” she explained.

Coluni outlined several capabilities that payers should look for in an AI vendor:

• A cognitive platform for conversations with natural language
• Annotation of clinical terms with explanations members can understand
• Dynamic conversation that uses logic to ask clarifying questions and give specific responses
• A cost estimation tool so members know their responsibility before they get a bill
• Integration of multiple data sources to help personalize interactions.

While these might seem like a lot to ask of an AI system, Coluni agreed with fellow AHIP speaker Atul Gawande, MD, that it’s always possible to do better.

“At the time we’re living,” she concluded, “we have more technology and tools to change this member experience and change the way healthcare is delivered and will feel in the healthcare system than ever before.”