Transforming healthcare contact centers through generative AI

ChatGPT and other generative artificial intelligence (AI) platforms are poised to have a transformative impact on healthcare and other industries. The ability of generative AI to create text, images, audio, video, and other forms of content is already being tested for use cases in marketing, finance, education, and healthcare. 

The current inherent limitations of generative AI, however, introduce compliance and ethical risks that healthcare organizations must consider before they integrate this technology into clinical processes. Specifically, ChatGPT and other generative AI tools suffer from accuracy problems that frequently lead them to invent facts that don’t exist in the real world. (This phenomenon is called “hallucinating.”)

I think we can all agree that generative AI’s propensity for inventing facts renders it poorly suited for physicians to rely upon at the point of care. Further, patient privacy concerns and lack of governance increase the challenges of using generative AI for clinical use cases. 

While these problems are solvable, these clinical use cases likely will take time to develop and deploy safely. Fortunately, one area of healthcare can benefit today from generative AI: the contact center. Generative AI can be deployed by providers and health plans to assist customer support agents and enhance patient/member support and self-service.

Enhancing the customer experience

Contact centers are integral to the customer experience (CX). Patients and members rely on healthcare contact centers to set up appointments, request information, check on prescriptions, pay bills, and more. Delivering a superior CX enables healthcare providers and payers to build the type of enduring customer loyalty that translates into future revenue.

Many healthcare consumers will have more interactions with a contact center than their individual clinicians or plan representatives. Though some of these interactions are relatively simple and transactional (such as setting an appointment or making a payment), a patient/member’s contact center experiences can determine whether that person remains with a provider group or health plan or takes his business elsewhere.

Despite these high stakes, too many healthcare organizations fail to address factors that undermine CX stability and performance. High levels of turnover continue to plague contact center workers across all industries. Results of a 2022 survey by the International Customer Management Institute (ICMI) showed that 57% of contact center professionals said turnover had increased since the prior year.

Agent burnout is a constant challenge for contact centers. A combination of chronic understaffing and the already intense pressure of dealing with high volumes of calls from patients and plan members is driving more contact center workers from their jobs. 

Many of the people reaching out to healthcare contact centers may be seeking important test results, have questions or complaints about insurance coverage, or are trying to schedule an immediate appointment with an oncologist or cardiologist. In addition, some of them have serious health issues and may be anxious, frustrated, and possibly in pain. 

That’s a lot to handle for relatively low-paid contact center workers. Making matters worse is that many contact center agents must rely on technologies which make their jobs even more stressful, such as user-unfriendly interfaces as well as inefficient workflows. Customer service representatives can scroll through multiple screens to find medical or insurance information for a distraught patient/member waiting on the line – with the call backlog growing by the minute – only so many times before they reach their limits and seek other employment.

Limited technologies also have an impact on the CX through a lack of channel options. Many patients/members prefer talking directly to a support agent on the phone. Others may prefer email, text, or even a chatbot. Lack of omnichannel options for consumers is a deal-killer in the digital age. Even if a healthcare organization does offer multichannel support, information found in other channels may be inaccessible or difficult to access by a contact center agent on a call with a patient or member.

Where generative AI fits in 

Successfully assisting patients and plan members can be highly fulfilling to contact center workers because they know they are providing a valuable service to people in need. ChatGPT and other generative AI platforms can be deployed today to make healthcare contact centers more efficient and effective for patients, members, and support agents.

Generative AI combined with automation can ease the agent burden by performing tasks the agent otherwise would do. For example, generative AI could be used to craft email replies to patients/members who have questions or need specific information, to summarize discussions, and to set dispositions. This creates a frictionless experience for patients and plan members by making communication more efficient and by freeing up agents to focus on interacting with patients/plan members.

Today’s consumers want to feel empowered, so it’s critical that a healthcare organization’s CX strategy includes self-service components. ChatGPT and other generative AI algorithms are well-suited to enabling customer self-service. A generative bot could help patients and plan members make appointments, clarify post-op instructions, or provide advice on medication schedules. 

Deeper self-service is about more than just reducing friction and wait times for consumers. When bots can handle a wider variety of topics and patient inputs – a key selling point of generative AI chatbots – it means more calls and chats are deflected from contact centers. Improved call deflection has cost benefits for the organization, and it means live staff are focused on only the most complex conversations that require a human touch.

Should patients or plan members ask to speak to a human, generative AI could then route them to the appropriate agent or department. And when contact center agents are conversing with customers, generative AI can feed them concise, well-organized summaries of relevant information so they won’t have to waste time searching for answers in a knowledge center.

If a caller wants to schedule an appointment with a specialist, ChatGPT can provide agents with information on facility availability, patient transportation limitations, and other variables to facilitate the process and ensure a specific time slot meets the patient’s needs. For newer agents lacking experience in the nuances of scheduling, generative AI can be enormously helpful.

Conclusion

Until generative AI’s accuracy and hallucination problems are eliminated or controlled, it will remain a poor fit for broad clinical use. But generative AI platforms can be used by providers and payers today to improve CX by allowing agents to be more helpful and productive and offering better self-help features for patients and members. This will have the dual benefit of improving patient satisfaction and reducing agent burnout and turnover.