Level up your learning with Luma experts
X

How Epic-Integrated Call Center AI Saves Staff Time and Improves Patient Access at an Academic Medical Center

Kara Bodner

The University of Arkansas for Medical Sciences handles 95% of after-hours calls with AI automation for healthcare call centers

Healthcare call centers are overwhelmed. Agentic AI can help – right now.

Like many access leaders, University of Arkansas for Medical Sciences’ Michelle Winfield-Hanrahan had a legacy workflow in her contact center – and it was hurting patient experience and staff efficiency. 

When a patient called after the contact center – also commonly known as an access center or call center – had closed for the day, they left a voicemail on a dedicated after-hours phone line. The workflow had been created to avoid costly overnight staffing. “Staff shortages don’t only affect nursing and clinical roles, but their effect spans into the call center as well,” said Winfield-Hanrahan.

UAMS’ Epic-integated Luma AI solves healthcare call center challenges

But the after-hours line presented a challenge for staff the next day. Many patients used it to let UAMS know that they needed to cancel an appointment – and staff needed to take action on those requests right away. 

“The team was using three hours’ worth of time every day just listening to voicemails from patients who called in after hours. Then, they had to manually cancel appointments,” said Winfield-Hanrahan. “We had to try to backfill those appointments or lose the revenue.” 

UAMS needed healthcare call center automation to improve the patient experience, save time for staff, and avoid this lost revenue.

The AI-powered healthcare call center agent that cleared 800+ hours of work overnight

Luma’s Navigator AI concierge was the solution. With HIPAA-compliant, zero-retention agentic AI that integrated with UAMS’ Epic EHR, patients easily cancel their appointments after hours. The next day, staff simply see an up-to-date schedule – no manual follow-up required.

“We were looking for efficiency — and we found it with Navigator,” said Winfield-Hanrahan. “Navigator completely took that manual work off our plates.”

Winfield-Hanrahan cites Epic integration, a quick implementation time, and minimal change management as benefits that encouraged her to use Navigator. “Implementation took just three weeks from start to finish,” she said.

UAMS also chose Luma’s AI concierge because it complemented and expanded on the patient access options UAMS already offered with the contact center and MyChart. “Navigator sounds and acts like a human, and it’s so helpful,” said Winfield-Hanrahan.

With Navigator, UAMS has seen results including:

  • 95% of inbound calls to after-hours line are automated
  • 88% callers verified as current patients
  • 800+ fewer access center hours spent processing cancellations annually
  • ~10,000 calls handled annually without staff intervention
  • 4% of total call center inbound calls handled by AI vs. staff
  • Reduced call abandonment rate in the call center

Results at University of Arkansas for Medical Sciences (UAMS) using Luma’s Navigator

To hear more about UAMS’ story and their results from Navigator, check out the following resources:

To learn more about Navigator and how it could help your organization, check out these resources:

In this webinar, Michelle Winfield-Hanrahan dives deep into how she chose and implemented Luma’s AI concierge in UAMS’ call center.

AI that works with Epic (not against it)

Today, your staff might be dealing with overwhelming call volumes – and many of these calls are for simple needs, such as a cancellation or an FAQ, that don’t require the experience and knowledge of a dedicated staff member.

But calling a business or health system and reaching an AI agent is no longer in the far-off future. Organizations like UAMS are using them every day to provide a better patient and staff experience, and ultimately serve more patients with fewer resources.

An AI concierge allows you to decant, or deflect, simple calls to an AI agent while your staff handle higher-complexity calls and patient needs. Worried that patients won’t want to use AI, or that AI lacks a personal touch? This strategy gives you the best of both worlds. Patients who are comfortable with self-service can quickly meet their own needs using AI, while those who need to reach a staff member can stay on the line. 

Here are some of the benefits of AI in healthcare call centers:

  • Convenience and quality for patients: Needs are solved easily and naturally — even after business hours.
  • Less manual work for staff: Fewer calls and portal messages to handle, plus visibility into every patient interaction in real time.
  • More efficiency for your access center: Staff spend less time on straightforward calls and administrative tasks.

Legacy workflows and the potential for problems in healthcare call centers

Maybe your organization doesn’t struggle with an overloaded after-hours phone line, like UAMS did. But don’t discount the value that AI might bring to other challenges your call center is facing. Winfield-Hanrahan, an experienced access leader who has consulted with many health systems to improve their call center workflows, encourages fellow leaders to consider other legacy workflows that might create problems in healthcare call centers and impact your patients and staff.

Call centers are costly to staff, and agents can be difficult to retain, said Winfield-Hanrahan. “It’s a challenging job, and we want to make sure that our agents are spending time on ‘true-to-task’ work helping patients – not on hours of administrative tasks,” she said. 

Research backs up this challenge. Contact center attrition rates are anywhere from 30% to 60%, with one poll of 400 contact center employees placing it at 42%. And the attrition rate for agents is about 1.3x higher than the average annual attrition rate in the US. Another recent study reported that more than half of contact center agents are on the verge of burnout.

If your organization is experiencing any of the following challenges, you might have legacy or manual workflows that AI could help automate with minimal process changes or change management:

  • High call volume and long hold times
  • Low enrollment rates for MyChart or your patient portal – or low patient portal engagement
  • High attrition rates for your call center staff or agents
  • Too many calls to handle – both outbound calls and inbound calls
  • Manual follow-ups to compensate for inefficient or overly-manual workflows (such as manually canceling appointments) 
  • Lengthy turnaround times for referrals

Think AI isn’t for you? Here are an access leader’s tips for evaluating and choosing an AI concierge

Not sure whether AI is the right fit for you, or how to go about evaluating and selecting the right AI concierge? The benefits of AI in healthcare call centers don’t require you to reinvent all of your workflows – a smart application of AI automation, like UAMS’, can make a big difference overnight.

Winfield-Hanrahan offers these tips for fellow access leaders:

  • Clearly define the problem you’re solving and the intended outcome. Winfield-Hanrahan is open to new technologies like AI, but emphasizes that the specific problem, goals, and metrics need to be clearly defined before introducing them.
  • Avoid falling for the hype cycle, but use AI if it’s the right fit. According to Winfield-Hanrahan, Navigator was a good fit because UAMS already had experience with Luma’s deep Epic integration and ability to create customized workflows. “I look for vendors that can be a partner,” said Winfield-Hanrahan. Navigator was secure, connected to UAMS’ other workflows, and could expand to more use cases in the future – all justifying using AI beyond just industry hype. 
  • Have the right people at the table to evaluate the function AI will be solving. In the call center example, Winfield-Hanrahan points to her deep experience with call center workflows as key to the evaluation process – and she recommends that technology or digital transformation leaders include relevant stakeholders if they’re looking to implement AI. 
  • Start with a low-risk use case that doesn’t require significant change management. After-hours cancellations were an ideal initial use case for AI at UAMS, because they didn’t significantly change workflows for staff or patients.

Ready to see Navigator in action?

We’re here if you want to chat about Navigator, how UAMS is using it, or creative ideas for how it could solve inefficiencies in your call center. See Navigator in action here. 

Ready to learn more? Schedule a meeting today!

Want to join the conversation with other healthcare leaders talking about the latest technology, challenges and opportunities, and creative ways to improve healthcare delivery? Tune in to Digital Health: On Air, our podcast discussing pressing healthcare topics with experts and leaders like you. You can find it on Spotify or YouTube, too!


You might be especially interested in episodes featuring CHIME’s Keith Fraidenburg (AI in Action: How Health Systems are Approaching the AI Boom) or Ardent Health Services’ Anika Gardenhire (Innovating with Purpose: Strategies for Meaningful Investment in AI).