AI in Patient Access: How UAMS Reduced Contact Center Burden and Improved Patient Experience

AI in Patient Access: How UAMS Reduced Contact Center Burden and Improved Patient Experience with Luma’s Navigator

Missed the webinar? Watch the full recording here.

Healthcare access centers are facing a perfect storm of increasing patient demand, high call volumes, and staffing shortages that make it difficult to provide timely, high-quality service. The University of Arkansas for Medical Sciences (UAMS) found itself in this exact situation. Despite a robust digital front door, phone lines remained the primary patient access point, leading to high hold times, staff burnout, and delayed care.

So, how can AI help? In a recent Patient Access Collaborative (PAC) webinar, Michelle Winfield-Hanrahan, Chief Clinical Access Officer at UAMS, and Aditya Bansod, Co-Founder & CTO of Luma Health, shared how Epic-integrated AI helped UAMS reduce contact center workload while improving patient experience.

Here’s what they learned—and how other Academic Medical Centers (AMCs) can safely and effectively implement AI in patient access.

The Challenge: High Call Volume, Staff Burnout, and Patient Frustration

Like many AMCs, UAMS faced critical challenges:

• Contact center overload: Despite digital self-service options, patients still relied heavily on phone support. “Efficiency was a major goal for us – to reduce our call volume,” said Winfield-Hanrahan.
• Staffing shortages & burnout: Contact center agents faced overwhelming workloads, leading to high turnover. UAMS wasn’t alone – research shows that contact center attrition rates are anywhere from 30% to 60%, with one poll of 400 contact center employees placing it at 42%.^[1]
• After-hours bottlenecks and operational inefficiencies: Patients calling after hours were sent to voicemail, creating a backlog of manual follow-ups. Staff spent 2-3 hours daily processing the previous day’s voicemails instead of handling real-time patient needs. 

Traditional solutions—like adding more staff or expanding contact center hours—weren’t sustainable.

Across the healthcare industry, experts agree that simply adding more staff to handle inefficiencies or expand access is not a long-term strategy. The United States is already experiencing a significant shortage of healthcare staff and clinicians, which the AHA only expects to grow in the coming years.^[2] “Traditional tactics have reached their limits,” said Sonia Singh, Chief Insights Officer at AVIA, in her report Radical Reinvention: Future Proofing Your Health System. “We have already tried reducing services, cutting costs, and expecting teams to achieve more with fewer resources.”^[3]

Faced with these trends and a significant drain on valuable contact center resources, UAMS needed a smarter way to manage after-hours calls and relieve the pressure on their staff. The solution needed to alleviate – not add to – staff burden and require minimal change management.

The Solution: AI-Powered Call Automation with Luma’s Navigator

To address these challenges, UAMS turned to Navigator, an AI-powered patient concierge created by Luma Health that integrates directly into Epic. Unlike standalone AI chatbots, Navigator enables real-time, voice-based self-service for patients calling in, without requiring them to log in to a portal.

“We wanted to take some of the administrative burden off of our agents,” said Winfield-Hanrahan. “Handling simple cancellations that would have otherwise gone to voicemail is the perfect use case for AI.”

The concierge is built on Luma’s real-time, API-based integration with Epic. It supports both patient interactions and staff efficiency by relying on:

• Deep EHR integration and patient context.
• Zero-retention generative AI used with healthcare privacy in mind, including OpenAI GPT-4, Claude Opus, deepgram, and others, fine-tuned for healthcare.

Luma also keeps information secure when UAMS’ patients use Navigator. Luma is ISO 27001:2022 Certified, HITRUST CSF r2 Certified, TX-RAMP Level 2 Certified and performs a SOC 2 Type II attestation annually, and Navigator is built to be ISO 42001:2023- and HIPAA-compliant.

The Results:

• 10,000+ calls handled annually, without staff intervention
• 800+ staff hours saved per year
• 95% of after-hours calls automated, reducing next-day backlog
• 100% Epic-integrated, ensuring seamless workflow adoption

"Implementation took just three weeks from start to finish. Navigator completely took manual work off our plates. It's so helpful." — Michelle Winfield-Hanrahan, UAMS

Want to see how it works? Try it out and learn more at https://www.lumahealth.io/patient-success-platform/ai-navigator/. 

How UAMS Designed a Seamless AI Rollout

Academic medical centers have complex workflows and often, significant oversight that might make implementing AI solutions feel out of reach – or at least, a goal for the longer term. But UAMS quickly implemented AI for their contact center and saw results within weeks.

Winfield-Hanrahan’s approach enabled a targeted application of AI that the organization was comfortable with. Now, the success is creating a foundation for future expansion of AI workflows. Here are her keys to success for implementing AI at an academic medical center, without getting caught in overly complex processes.

• Start small, scale strategically: UAMS began with after-hours cancellations. This workflow was an ideal fit for AI because patients were reaching a proverbial dead end when calling the access center, requiring staff follow-up and documentation the next day. Implementing an AI-based self-service option for these calls improved the patient experience, did not require significant change management, and still allowed patients who preferred leaving a direct voicemail to do so.
  • Why it worked: This lower-complexity, lower-stakes workflow allowed UAMS to test the accuracy of the Navigator concierge and served as a proof of concept before UAMS expanded AI to other workflows. UAMS also received white-glove implementation support from Luma, which minimized IT work and change management.
• Keep workflows Epic-first: A significant issue with the after-hours voicemail line was that it required staff to manually take action the next day. A staff member needed to manually go into Epic Cadence and cancel the appointment a patient called about after listening to the voicemail, which took up several hours each day. Winfield-Hanrahan selected Navigator specifically for its bi-directional, real-time integration with Epic to eliminate this manual work.
  • Why it worked: Navigator integrates directly into Epic Cadence, ensuring staff don’t have to manually document AI-handled calls. Winfield-Hanrahan also points to access and IT stakeholders working together as a bright spot.
• Focus on patient experience: Many organizations that are hesitant about using AI are worried about the impact to patient experience. Perhaps patients will feel that AI is impersonal, or they won’t be able to get their needs met and the diversion to an AI agent will waste the patient’s time. UAMS ensured that patients calling after-hours could still leave a voicemail for staff if they preferred a direct call back, but found that patients were happy to use AI to cancel their appointments in real time.
  • Why it worked: Patients are growing more familiar with using AI and self-service solutions. In fact, according to KLAS Research, patients have long been asking for more self-service options than their healthcare provider offers. At UAMS, 95% of patients calling after hours opted to stay on the line with Navigator instead of leaving a voicemail.

What’s Next: Expanding AI for More Patient Access Needs

Encouraged by Navigator’s early success, UAMS is now exploring additional AI applications, including:

• Automated referrals & fax routing to reduce administrative burden.
• More AI-powered self-service options so patients can manage appointments, request prescription refills, and more.
• Expanded multi-language support to ensure equitable access for diverse patient populations.

Want to Try Agentic AI in Your Contact Center? PAC Members Get 4 Months Free

For PAC members, Luma Health is offering 4 months free of Navigator AI, including:

• Up to four Navigator modules (e.g., cancellations, rescheduling, FAQs)
• White-glove implementation with Epic-experienced analysts
• Real-time analytics to measure AI’s impact on call volume and staff efficiency

Interested? Get in touch with the Luma team today to explore how EHR-integrated agentic AI can reduce manual work and improve patient access at your organization.