TLRT: Responsible AI and the Evolving Role of the Healthcare Contact Center

To gain a first-hand perspective on best practices for bringing AI into the contact center, CHIME Senior Strategist Frank Cutitta and Patty Hayward, GM of Healthcare and Life Sciences at Talkdesk, convened a panel of experienced healthcare executives who are exploring the potential benefits of AI-driven technology in the contact center.

The roundtable included:

• Ganesh Persad, Director of IT Applications and Digital Health Memorial Healthcare System (FL)
• Mikkel Finsen, CTO and CISO, Open Door Family Medical Centers
• Alexander Koster, Director, AVP of Value Transformation Nemours Children’s Health System
• Bernard Rice, CIO, Nemours Children’s Health System
• Timothy Calahan, CTO, Michigan Medicine
• Emily Krohn, Information Technology Director, Faith Regional Health Services
• Paul Williams, AVP Infrastructure Technology, Penn Medicine
• Mona Baset, VP of Digital Services, Intermountain Health
• Praveen Chopra, Chief Digital and Information, Officer Emplify Health (formerly Bellin and Gundersen Health System)

SUMMARY

Artificial intelligence (AI) is quickly becoming a mainstay of the healthcare technology stack, with rapidly growing interest across the clinical, administrative, and operational spectrum. While developers are still in the early phases of experimentation with many of the potential use cases, some areas are already well on the road to maturity, with the contact center being one of them.

AI-supported customer interactions are already a mainstay in a variety of other industries, giving healthcare an edge as leaders seek to streamline and simplify patient experiences throughout the consumer journey.

But there are still challenges ahead, especially in an ecosystem where trust, privacy, and accuracy are not just good for an organization’s reputation, but are also increasingly tied to compliance rules and regulatory mandates.

Deploying AI in a responsible and effective manner within the healthcare contact center will require careful planning, strong governance, and a clear perspective on what consumers really want and need out of their engagements.

“Driving value from AI in healthcare contact centers is not as straightforward as deploying a new bot,” said Cutitta. “There are still major considerations around how to responsibly deploy consumer-facing AI and when it should be used to resolve queries autonomously or to support a human agent. Perhaps most importantly, AI also poses questions about how the role of contact centers will change in the future and the potential for driving new health and business results. We need to carefully consider AI’s impact across the different dimensions of healthcare consumer experience: people, processes, and technology.”

ESTABLISHING CLEAR GOALS FOR OPTIMIZING CONTACT CENTER OPERATIONS

Contact centers aren’t just the digital front door for new patients. They also provide critical navigation services for existing patients, making them central to the overall consumer experience. But they don’t always get the attention they deserve from a business strategy standpoint, the executives noted, making it difficult to appropriately bring new technologies on board.

Ganesh Persad, the Director of IT Applications and Digital Health at Memorial Healthcare System in Florida, emphasizes the importance of a holistic approach to a contact center’s ability to gain comprehensive insights into each individual patient/consumer they interact with. “Agents now require a preliminary understanding of the patient/consumer profile and the possible needs of a patient before initiating a conversation, such as risk scores, social determinants of health, care gaps, and personal preferences” he explained. “However, creating detailed profiles and personas has been a complex task with outdated technology, particularly when the contact center has not fully accounted for interoperability and integration.”

To prime the ground for a successful AI deployment, contact center leaders need to be at the table during conversations about technology implementation to ensure that priorities remain aligned, added Alexander Koster, Director, AVP of Value Transformation at Nemours Children’s Health System.

“The factors that influence technology and staffing decisions are very different depending on the perspective of the decision maker,” he noted. “Strategic alignment is crucial, even before you start talking about things like AI.”

Call center centralization and optimization have endless benefits that enhance productivity and increase revenue. Organizations that implement these technologies well will see measurable improvements including reduced leakage of unanswered calls, improved patient access and experience, and valuable insights to inform decisions and strategies around call disposition and routing, according to Paul Williams, AVP of Infrastructure Technology for Penn Medicine.

“We realized a nearly 20% increase in new patient visits over a two-year period, resulting in significant increased revenue while integrating practices and reducing operational resource requirements,” he noted.

USING AI TO IMPROVE EXPERIENCES IN A TIME OF WIDESPREAD LABOR CHALLENGES

Adding further complexity to the issue are fears around AI’s potential role as a “people replacer” instead of a tool for augmenting and enhancing the human touch, the attendees continued. In a time when staff is hard to find, yet consumers are increasingly demanding personalization, balancing the roles of AI and human agents is one of the trickiest tasks.

“Contact center work can be very difficult,” acknowledged Praveen Chopra, Chief Digital and Information Officer at Emplify Health (formerly Bellin and Gundersen Health System). “About 85% of the time, the people calling in are stressed or dissatisfied, which can contribute to high turnover at the contact center. We need to identify how to use AI to relieve those stresses on both the consumer and the employee so that our staff feel like they have a supportive and viable environment to further their careers.”

The key is finding the right blend between leveraging AI for customer self-service and making human agents available to handle complex concerns, added Mona Baset, VP of Digital Services at Intermountain Health.

“We do want to make it clear that consumers have the option to self-serve if they want to,” she said. “And if they don’t, they can opt out and speak with an agent. If they choose that path, we also want to use AI to equip those staff members with the information they need to understand the patient’s issues and solve the problem quickly.”

“Did they call in earlier that day and didn’t get the answer they needed? What were their last interactions, and how can we use that information to predict what they might need? How can we use AI to reduce the avoidable volume of calls to our human staff while completing the necessary calls more seamlessly and with greater satisfaction on both sides?”

Developing a better understanding of customers and anticipating their needs will be essential for improving experiences with both human and AI agents, said Timothy Calahan, CTO at Michigan Medicine.

“We’re playing around with generative AI as it relates to summarizing patient data or pulling unstructured information out of incoming faxes and other documents,” he said. “If we had better access to longitudinal records, we could respond to incoming requests in a more targeted and personal way.”

“There’s huge promise for using generative AI in chatbots and other tools, but we’re just in the early stages of understanding how to best deploy those technologies in a way that improves the patient experience without depersonalizing it.”

One way to improve patient experience with AI is to improve its linguistic capabilities. At Open Door Family Medical Centers, which serves part of the NY metropolitan area, 80% of patients speak Spanish. “We encounter so many different dialects of Spanish from various regions of Central and South America, and some of the AI tools are unable to pick up the nuances of each dialect,” Explained Mikkel Finsen, the health system’s CTO and CISO. “When these patients are introduced to AI, many will hang up and not interact further.”

DEVELOPING BEST PRACTICES FOR RESPONSIBLE AI USE IN THE CONTACT CENTER

As with any new technology, AI is likely to require some trial and error before it truly shows its value in the contact center.

The most important lesson to consider as this process unfolds is not to assume what patients want, cautioned Patty Hayward, GM of Healthcare and Life Sciences at Talkdesk.

“It may seem more efficient to use a phone tree that asks patients to press one or two to get to the right place, but how many times have we seen them just press one because they’re impatient or can’t remember nine different options?” she said. “Think about the most natural way of interacting with technology. Conversational AI might be much more effective for steering people in the right direction, even if it seems like a more ‘unstructured’ approach to a technologist.”

Taking a persona-based approach to developing an understanding of what users want will be important for making the most of AI tools, especially when contact centers handle a number of different interactions, from donors to internal clinical staff to patients to caregivers or parents, said Chopra.

“Start with figuring out what experience you want to provide to each type of person and then design the technology around turning that vision into a reality,” he advised. “Get as granular and tactical as you can so you are able to equip your AI and your human agents with the right information to meet the needs of each persona.”

Maximizing the value of every interaction can also improve experiences while relieving pressure on staff in other areas of the organization, added Emily Krohn, Information Technology Director at Faith Regional Health Services.

“Determining and reminding patients of missing vaccines, screenings, or that it is time for their annual wellness exam in the contact center can actually produce a net good,” she said. “We need to think about how to use AI to gather information on open gaps in care and surface options for getting those services completed while we have the patient engaged. Building those capabilities into the infrastructure can improve satisfaction while ensuring high-quality care delivery.”

Last but not least, be certain to keep the human element at the forefront of decision-making, especially when it comes to staffing the flesh-and-blood portion of the contact center, said Baset.

“We are always looking to hire people who are curious, adaptable, and flexible, because they are going to need those skills to keep working with technology that’s changing almost every week,” she said. “Solid technology skills matter, but don’t get too hung up on specific platforms — those tools might not be around in a year or two. But having people who are comfortable with change and able to make correlations between different processes is priceless.”

As artificial intelligence makes its way into the healthcare contact center, leaders will need to think strategically about how to leverage these tools to augment their digital engagement processes. Whether AI is being used to enhance self-service capabilities or empower more effective person-to person interactions, it will be imperative to find the right mix of modalities to make certain that the patient experience is personalized and effective.

Doing so will create an environment where AI is being used responsibly and appropriately to improve interactions while maintaining high quality of care and services, easing burdens on human staff, and meeting the changing expectations of the next-generation consumer.