As Value-Based Care Scales, Healthcare Contact Centers Pivot to Close Clinical Gaps

Most contact center AI is built to deflect calls and save money. But in healthcare, some operations leaders are using it for the impactful purpose of spotting gaps in patient care. When AI scans claims databases for overdue checkups and prompts agents to schedule appointments during routine calls, contact centers can directly support clinical outcomes rather than just reducing handle times.

Patrice Chance is a healthcare operations leader and the current Manager of Member & Provider Operations at Ascension. Having managed multimillion-dollar budgets and built high-volume contact centers at organizations like Scott & White Health Plan, Memorial Hermann, and Universal American, she works closely with day-to-day Medicare and Medicaid operations. Chance sees AI as a way to link routine service interactions with preventive medicine.

"AI can search through our claims database, look for whether a member has had a wellness visit, put up an alert, and tell the CSR this member needs to be scheduled. That helps improve our HEDIS scores." By prompting these appointments, contact centers can better connect their daily work with clinical measures and help improve the quality scores that are the industry's ultimate measure of success.

• Knowing when to fold: Effective AI orchestration relies on a seamless handoff that triggers as soon as an interaction shifts from a data request to a more nuanced need. While bots excel at retrieving objective facts, Chance says they lack the lateral thinking required when a member disputes a coverage decision or an out-of-pocket cost. "If a member is calling to find out how much they have met towards their deductible, AI should be able to handle that and provide the exact balance. But the moment a member states they thought they already met their deductible, that is when AI needs to stop and hand it over to a customer service representative because now they have to dig a little deeper," she explains.

• Math versus nuance: To understand the system's success, Chance looks at how many calls AI has successfully handled versus how many it has transferred. She sees a 60/40 split as a good benchmark. "If AI handles over 60 percent of our calls, it leaves 40 percent for our agents to assist because those calls are more complex."

Automating routine inquiries can change how agents spend their time. As AI takes on straightforward questions, agents often find themselves spending more time navigating intricate disputes. To support staff in that environment, Chance looks toward using automated quality monitoring to review every interaction and surface targeted coaching opportunities. "AI tools can audit every single one of our calls and give us real-time feedback that a specific person may require additional training," she says. It's a model that mirrors wider industry trends in human-centric AI coaching, framing automation as a workforce empowerment tool rather than a replacement strategy. It allows us to upskill or up-train our agents because we have real-life data where AI is showing us exactly where our deficiencies are."

• The scaling maturity gap: Chance acknowledges that realizing these operational gains depends heavily on an organization's budget and readiness. Many large insurers are investing heavily in AI to support everything from customer service to back-office operations. For some smaller TPAs, the timeline moves much slower. She notes that leaders in smaller environments are sometimes wary of even basic tools, highlighting the institutional hesitancy that operations directors must overcome to build a business case for new technology. "I have noticed some of my leaders don't even want to utilize Google Meet to record meetings because they're afraid the recordings might get shared externally. Sometimes they're a little hesitant to implement basic functions before we even start exploring tools."

• Peer-reviewed procurement: To navigate the hesitancy, Chance focuses on building her own technical literacy and grounding decisions in data from peers. She evaluates tools like knowledge management systems or generative AI by studying how other organizations have deployed them and what results they report. "I take those lists and reach out to those organizations to gather data to fully understand what we are trying to solve and how we want to utilize AI."

The same research-driven approach also dictates how Chance manages risk. Because false information regarding claim statuses or clinical inquiries can present compliance challenges, she favors limited pilots and careful testing. It's a practical safeguard to ensure that the technology is rigorously vetted before it interacts with live patient data. "Because the healthcare industry is so complex, you have to ensure you are using AI responsibly. One of the major risks is the hallucination gap. We want to ensure that whatever tools we are utilizing are implemented correctly and tested before they are put out to mass production." Ultimately, she believes the future of healthcare operations lies in moving past the fear of the unknown to embrace data-backed automation safely. By grounding tech adoption in peer research and high-fidelity testing, leaders can build the infrastructure necessary to protect member privacy while maximizing operational reach. "A lot of smaller health plans are afraid to implement AI because they don't fully understand the return on investment. We shouldn't be afraid to utilize it. We should just do our research thoroughly."