Arrow Back
Resources Hub
Calendar
November 16, 2026
Time To Read
7 minutes

Where AI can transform patient access, according to health system CEOs and CFOs

Health system CEOs and CFOs share how AI agents can close patient access gaps, earn staff trust, and deliver measurable ROI.

By
Share

summary

At a Notable-hosted roundtable during Becker's CEO + CFO Roundtable, health system leaders said patient access breaks down in the handoffs between systems and teams, not just from a lack of capacity. They agreed agentic AI can close those gaps, but only with clinical guardrails, front-line buy-in, and a clear plan for the capacity it frees.

Content

Share this post

Health systems are under pressure from every direction. Hospital expenses grew 7.5% in 2025 against an average operating margin of 1.3%, and the average wait for a new patient appointment is 31 days. Often the capacity exists, but the patient access work around it, like referral management, prior authorizations, and patient scheduling, stalls between fragmented systems and the teams chasing it.

That challenge framed the breakfast roundtable Notable hosted at Becker's CEO + CFO Roundtable in Chicago this November, where health system leaders joined Dr. Nick Desai, Notable's Chief Medical Officer. 

The discussion centered on how agentic AI and healthcare automation can close those gaps. Four themes stood out.

1. The work still has to be done by someone

When leaders shared where work stalls, the answers were remarkably consistent. Behind every patient request is a staff member chasing something: outside records, referral paperwork, insurance checks, or an open slot on the schedule.

One attendee described a staff member whose job is essentially to "gopher" records from multiple sources so providers have what they need. At one practice, missing records still cancel visits. Others pointed to the effort of booking back-to-back PT, OT, and speech therapy as a single block, and to the challenge of knowing in advance when a visit needs more time, such as a life care planning appointment.

This is the gap many healthcare operators are familiar with. A worklist or queue surfaces work, but it doesn't finish it. Leaders described teams "making the system work" through workarounds, which keeps patients moving but quietly consumes capacity.

The common answer to these gaps has been adding another point solution. Dr. Desai described the operator's dilemma from his time in the health system C-suite: 19 teams each asking for five point solutions would have meant 95 disconnected tools running on top of the EHR. That's more than any organization can monitor, trust, or manage. An agentic AI platform that works across the EHR and existing systems, and completes the work instead of queuing it, offers a different approach.

2. Access is a growth strategy, and reaching patients is the hard part

Patient experience came up as a core pillar of growth and retention, and several leaders said their access teams are actively looking at newer approaches. The pain points were familiar to anyone who runs operations:

  • Long waits for a provider appointment
  • Long call center hold times
  • Difficulty reaching patients for follow-up care and annual exams
  • Gaps in the end-to-end referral process
  • Care gaps that, left open, turn into ED visits
  • Late cancellations, including patients who aren't medically optimized and cancel anywhere from clinic through the OR
  • Rising administrative costs

One hospital COO described exactly what proactive access could look like: triggers on lab and clinical data that flag patients who aren't ready ahead of their OR or clinic dates, and prompts to book longer visits when a patient needs them. The goal isn't just answering the phone faster, it's reaching the right patient before the gap opens.

That's the shift health systems using Notable's AI agents are already seeing. Montage Health cut referral turnaround from 23 days to 1.5 days, with 80% of its 10,500 referrals now touchless. Catholic Health's Voice AI authenticates callers in 36 seconds. Marshall Health saw a 20% increase in telehealth visits once patients could schedule themselves.

3. Adoption depends on trust

There’s one question that’s always top of mind when discussing AI: almost every gap discussed in the room has a person on the other side of it, chasing the work. If agents start closing those gaps, what happens to those people? At Notable, we believe that jobs will change rather than disappear, but that dropping agents into today's workflows without redesigning them just produces a faster version of the same broken process.

Leaders named the other concerns that slow adoption:

  • Clinical safety. Administrative teams tend to be more receptive. Once agents cross into clinical work, the scrutiny rises sharply, and clinicians want clear guardrails.
  • Auditability. No one wants to trust software blindly. Leaders look for human-in-the-loop checks, audit trails, feedback loops, and visibility into what agents do and where their information comes from.
  • Fear of job loss. Staff need to see proof of how the technology makes their work better, not just hear that it will.
  • Physician workflow. Tools that add steps get dropped, and physicians fall back on what they already use unless someone helps them in person.
  • Governance. Layers of governance and vetting slow every rollout.
  • Affordability. Agents deliver a return but cost money, and smaller organizations feel that most, especially while still paying for legacy systems.
  • Workforce development. If AI absorbs entry-level work, where does the next generation build judgment?

However, the leaders in the room were just as clear about what works. Standardize processes and break down silos before introducing agents. Let clinical leaders explain the "why." Bring in peer organizations to share their experience, and keep internal champions with the technology from start to finish. Most of all, earn buy-in from front-line staff and operators, not only the executive team.

4. Impact has to be measured and put to use

Dr. Desai closed on a point that resonated with the CFOs in the room: unlocking capacity accomplishes nothing if you haven't decided what to do with it. If agents gave access teams back 30% of their time tomorrow, would it absorb growth, fill open positions, reduce overtime and agency spend, or something else? And how would you explain that choice to staff?

Some staffing gaps are flow and visibility problems rather than headcount problems. Measured that way, impact can be measured by more than just cost savings; it’s patients retained, revenue recovered, capacity used well, and time given back to teams. Southwest General, for example, saves $4 on every insurance FAQ call its AI assistant handles.

Five questions to take back to your executive team

The biggest takeaway from the room was that the hardest part of AI adoption is redesigning the work and bringing people along with it. Here are five questions to bring back to your executive team to address these challenges:

  1. Find the gaps: Where does work wait between your systems, and what is it costing you?
  2. Redesign the work: Are you rethinking the process, or just automating today's version of it?
  3. Connect the system: Will this work across the EHR and everything a patient touches, or is it one more point solution?
  4. Prepare your people: How will roles change, and where will the freed capacity go? This is the question the room spent the most time on, and the one that's easiest to skip.
  5. Measure impact: Track patients retained, revenue recovered, capacity used well, and time given back to your teams.

The measure of automation won't be how intelligent it is. It will be the impact: more patients getting care, fewer missed follow-ups, and less time chasing work.

Want to see how an agentic platform can address the access gaps in your organization? Get in touch today. 

‍

Heading

Button Arrow 
Button Arrow

Frequently asked questions

Health systems are using AI agents to automate referral management, patient scheduling, prior authorization, and call center work, so tasks are completed end to end instead of waiting in a queue. Montage Health, for example, cut referral turnaround from 23 days to 1.5 days with Notable, and Marshall Health saw a 20% increase in telehealth visits after enabling patient self-scheduling.

Leaders at Becker's CEO + CFO Roundtable said adoption depends on trust: clear guardrails for clinical use, human-in-the-loop checks, audit trails, and buy-in from front-line staff. They also recommended standardizing processes before introducing agents and deciding in advance how the freed capacity will be used.