What Every Practice Should Fix First
By Dr. Charles Accurso and Praveen Suthrum
Most GI practices do not have a demand problem.
They have a workflow problem.
Phones are ringing. Referrals are coming in. Procedure schedules are full. Patients are waiting weeks—or months—for appointments in many markets.
And yet, despite all this demand:
- Physicians are exhausted
- Staff are overwhelmed
- Patients struggle to access care
- Revenue leaks silently through inefficient systems
Before gastroenterology thinks about entirely new business models, most practices are sitting on something far more immediate:
Hidden capacity.
Hidden margin.
And hidden human intelligence trapped inside broken workflows.
The Prior Authorization Problem
If there is one workflow that perfectly captures the absurdity of modern healthcare operations, it may be prior authorization.
A patient needs care.
The physician already knows the patient needs care.
The staff knows the patient needs care.
And then the system pauses while everyone attempts to convince the insurance company that the care is necessary.
What follows is familiar to nearly every GI practice:
- portal logins
- uploads
- documentation requests
- hold times
- peer-to-peer scheduling
- repeated follow-ups
Highly trained people spend enormous amounts of time navigating administrative friction instead of caring for patients.
Everyone loses:
- Patients get delayed care
- Physicians become frustrated
- Staff burnout
- Practices lose efficiency
And despite technologies already existing to automate portions of this process, much
of GI still handles it manually.
This is not primarily an AI problem.
It is a workflow design problem.
The Greatest Waste in GI May Be Human Intelligence
Across GI groups, one pattern appears repeatedly.
Highly trained people are spending enormous amounts of time doing work that
should not require highly trained people.
Consider the modern revenue cycle:
- eligibility verification
- prior authorization
- charge entry
- claim status checks
- AR follow-up
- denial management
- appeal generation
Much of this is still manual.
Not because the technology does not exist.
Because the workflows were never redesigned.
The same pattern exists throughout the practice:
- scheduling calls
- recall management
- patient messaging
- intake coordination
- documentation routing
And increasingly, even physicians themselves are pulled into administrative work that
has little to do with actual medical decision-making.
This creates a strange paradox inside GI.
The industry is filled with highly intelligent people.
But enormous amounts of that intelligence are trapped inside repetitive systems.
Demand Is Not the Problem
One GI leader shared an example from the Midwest.
Patients were waiting months for appointments.
Primary care physicians were frustrated because they could not get patients seen quickly enough.
At first glance, this looked like a physician shortage problem.
But when the workflows were examined more closely, the issue was broader:
- fragmented scheduling systems
- inconsistent access pathways
- poor recall management
- operational bottlenecks
- physicians spending time on low-value tasks
It was less like a lack of demand.
And more like: having a line outside a restaurant but being unable to serve the customers efficiently.
That distinction matters.
Because many practices already possess more capacity than they realize.
And with GI demand continuing to rise — driven by an aging population, lower screening ages, increasing GI disease burden, and physician shortages — workflow redesign is no longer optional.
The practices that redesign intelligently will expand access.
The practices that do not will increasingly experience operational strain.
The Most Important Workflow in GI May Be the Recall System
One of the most practical insights from this discussion was surprisingly simple: most GI practices underestimate the power of the recall system.
At first glance, recalls appear administrative.
They are not.
They are continuity.
They are patient retention.
They are procedural volume.
They are longitudinal relationships.
They are operational stability.
One physician reflected that after decades in practice, nearly 50% of his procedures are no longer coming from new patients.
They are coming from patients he has cared for over 20 or 30 years.
Patients returning for:
- surveillance colonoscopy
- follow-up care
- repeat procedures
- continuity
And many are still coming back because the system remembered them.
Sometimes through something as simple as a letter.
Not sophisticated AI.
Not futuristic technology.
A functioning recall process.
That realization changes the conversation.
Because before GI practices chase entirely new revenue models, many are still leaking enormous value through weak continuity systems.
The future will matter.
But the practices that master operational fundamentals first will be in the strongest position to shape that future.
The Problem Is Often Not AI. It’s the Process.
Recently, one of us was speaking with a healthcare executive who felt overwhelmed
by the number of emails she was receiving.
Her instinct was immediate:
“We need AI to automate this.”
But after a few minutes of discussion, a more important question emerged:
Why were all these emails reaching her personally in the first place?
This was not primarily an AI problem.
It was a workflow problem.
And this is where many healthcare organizations make a critical mistake.
They try to automate broken systems instead of redesigning the systems themselves.
Sometimes the most important operational question is not:
“How do we automate this?”
It is:
“Why are we doing this at all?”
Workflow Redesign Comes First
This may be the single most important operational principle in this book.
Many GI practices are layering:
- AI
- automation
- portals
- messaging systems
- software tools
on top of operational structures that evolved over decades without intentional redesign.
The result:
- more alerts
- more fragmentation
- more administrative burden
- more physician fatigue
Before applying AI aggressively, practices should first:
- identify choke points
- simplify workflows
- remove unnecessary steps
- standardize repeatable processes
This exercise does not even need to begin with technology.
It can begin with pen and paper.
Take a clean sheet and ask:
- What should only physicians be doing?
- What should staff be doing?
- What should software be doing?
- What should not be happening at all?
Only then does automation become truly powerful.
Because AI alone will not fix GI.
Workflow redesign will.
The Importance of Open Systems
One lesson increasingly emerging from larger GI organizations is that operational redesign depends heavily on interoperability.
Many older healthcare systems were built as closed environments.
But the future of GI operations may depend on flexible systems that allow:
- recalls
- scheduling
- AI agents
- patient communication
- documentation
- revenue cycle
- analytics
to work together seamlessly.
This is why technology decisions are no longer simply IT decisions.
They are operational strategy decisions.
The practices that build flexible operational infrastructure today will likely adapt far faster than those trapped inside rigid workflows tomorrow.
The GI Friction Audit
Before investing heavily in AI or new services, leadership teams should conduct a
simple operational review.
Step 1 — Identify Physician Misallocation
Ask:
- Which tasks are physicians doing that do not require physician-level expertise?
- How much time is spent on documentation, messaging, refill management, or
scheduling interruptions?
Step 2 — Identify Repeatable Manual Work
List every repetitive process:
- prior auth
- eligibility
- recalls
- claim follow-up
- intake
- denial management
Then ask:
Why is this still manual?
Step 3 — Identify Revenue Leakage
Track:
- denials
- delayed payments
- incomplete recalls
- scheduling gaps
- no-shows
- referral leakage
Step 4 — Identify Workflow Complexity
Ask:
What processes should not even exist anymore?
This question alone can transform operations.
Step 5 — Prioritize High-ROI Fixes
Do not begin with massive transformation projects.
Start with:
- high-volume
- repetitive
- measurable
- low-risk operational fixes
Momentum matters.
The Low-Hanging Fruit Is Still Massive
Ironically, many practices do not need advanced AI first.
They simply need better automation.
There is enormous low-hanging fruit sitting inside:
- scheduling
- recalls
- insurance workflows
- claim follow-up
- documentation routing
- repetitive administrative tasks
Much of this can already be automated using technologies that have existed for years.
Not futuristic AI.
Basic workflow automation.
The challenge is not technological possibility.
The challenge is implementation.
Implementation is tedious.
- mapping workflows
- handling payer variation
- testing processes
- managing exceptions
- standardizing operations
It is not glamorous work.
But once implemented, the gains compound for years.
One large GI organization described how operational redesign began not with AI, but
with rethinking the entire patient journey:
- scheduling
- intake
- recall management
- revenue cycle
- patient communication
- payment workflows
The result was not simply efficiency.
It was scalability.
As the organization grew from dozens of physicians to hundreds, the redesigned operational infrastructure allowed the practice to expand without collapsing under administrative complexity.
That is the real power of workflow redesign.
What We Learned the Hard Way About AI
One of the most important lessons we learned during an AI recall pilot was
surprisingly simple:
AI does not need to solve the hardest problems first.
Initially, the ambition was full end-to-end workflow automation.
And in many areas, the AI performed extremely well.
But then subtle problems emerged.
For example, when patients discussed medications during voice interactions, the AI occasionally misunderstood words or documented the wrong medication.
Even if accuracy reached 95%, the remaining 5% became clinically unacceptable.
The result was unexpected: staff now had to perform quality control on nearly every AI interaction.
Instead of eliminating work entirely, we had partially shifted the work.
That experience changed how we think about implementation.
The lesson was not:
“AI doesn’t work.”
The lesson was:
Start with narrow, reliable tasks where AI can consistently achieve near-perfect accuracy.
Then expand gradually.
That approach is slower.
But it is safer, more realistic, and ultimately more scalable.
The Incentive Problem
One of the most important realities in GI is that operational redesign cannot be
separated from physician incentives.
A physician operating in a volume-based compensation system may be financially
rewarded for seeing:
- multiple lower-acuity follow-up patients
instead of: - fewer complex patients requiring deeper longitudinal management.
That does not make physicians resistant to better care.
It simply means the economics of the system shape behavior.
This is why many operational conversations fail.
Practices discuss:
- access
- continuity
- AI
- longitudinal care
- patient experience
Without discussing:
- compensation
- productivity expectations
- scheduling models
- physician incentives
But these issues are deeply interconnected.
If operational redesign creates:
- more complexity
- more time-intensive care
- lower throughput
Without aligning incentives appropriately, adoption will remain difficult regardless of the technology involved.
Healthcare systems often ask physicians to behave differently while continuing to reward the same behaviors financially.
The Hidden Opportunity Inside Existing GI Infrastructure
Before practices aggressively pursue entirely new business models, many still have significant unrealized opportunity inside the infrastructure they already own.
This includes:
- pathology
- infusion services
- ambulatory surgery centers
- anesthesia
- chronic disease follow-up systems
In many cases, the issue is not a lack of opportunity.
It is a lack of operational integration and patient continuity across the existing
ecosystem of care.
The first stage of GI 2.0 may not be building entirely new businesses.
It may be fully optimizing and integrating the assets already sitting inside the practice.
The Hidden Opportunity Beyond the Procedure
One of the clearest examples of unrealized value may be chronic GI care itself.
Take IBS.
Many practices:
- diagnose
- provide dietary guidance
- prescribe therapy
And then the patient largely disappears from the system.
But what if GI practices owned more of the longitudinal journey?
- follow-up
- monitoring
- behavioral support
- chronic care management
- digital engagement
The opportunity here is not simply new revenue.
It is:
- stronger continuity
- better patient experience
- improved outcomes
- and deeper ownership of digestive care over time
Most practices do not realize how much margin is hiding inside managing the patient journey more intentionally.
Why Practices Don’t Fix These Problems
If the inefficiencies are so obvious, why do they persist?
Because the current system still works well enough.
Practices remain busy.
Revenue remains strong.
Demand remains high.
That creates comfort.
Most GI groups are not operating from crisis.
They are operating from momentum.
At the same time:
- physicians are overloaded
- leadership bandwidth is limited
- workflows evolved incrementally over decades
- many physicians have not had the time or opportunity to deeply educate themselves on AI and operational redesign
This is not an intelligence problem.
It is a bandwidth problem.
And perhaps most importantly:
- many leaders still think linearly while the underlying technologies are advancing exponentially.
Most organizations redesign only when the old system becomes impossible to
maintain.
But the GI groups that move early will likely gain disproportionate advantages in:
- access
- scalability
- physician retention
- patient experience
- and operational flexibility.
The Goal Is Not Less Human Medicine
Ironically, the best use of AI may ultimately make medicine feel more human again.
One physician described his ideal future this way:
“I want to walk into the exam room, take a history, examine the patient, give follow-up orders, and never touch a laptop.”
That vision matters.
Because the future of GI should not be physicians competing with machines.
It should be physicians operating at a higher level because machines are absorbing the repetitive operational burden around them.
If physicians are freed from:
- endless clicks
- repetitive documentation
- administrative burden
- fragmented workflows
They regain:
- attention
- cognitive bandwidth
- presence
And patients feel that difference immediately.
The First 90 Days of GI 2.0
Do not begin with a massive transformation.
Begin with momentum.
Over the first 90 days:
Fix recalls
Make sure patients are not leaking out of the system unnecessarily.
Improve answering and access systems
If demand exists, patients should be able to enter the system efficiently.
Identify low-hanging automation opportunities
Focus first on:
- high-volume payers
- repetitive workflows
- high-dollar operational bottlenecks
Find physician champions
Do not force organization-wide transformation immediately.
Pilot operational changes with physicians and teams willing to experiment.
Measure ROI early
Visible wins create organizational trust.
Align physicians, APPs, and operational teams
Operational redesign only works if people understand why the changes matter.
Generate momentum through small wins
Momentum changes culture faster than presentations do.
Most GI practices do not need to reinvent themselves overnight.
But many do need to redesign the systems they already depend on every single day.
The future of GI will not be built by working harder inside broken workflows.
It will be built by redesigning those workflows around the intelligent use of human capacity.
And in many practices, that transformation may begin with something far less glamorous than AI.
It may begin with finally getting the basics right.
And once operational clarity begins to emerge, a larger question naturally follows:
If GI practices become more efficient…
if physicians reclaim cognitive bandwidth…
if operations become increasingly automated…
Then where does future growth actually come from?
That is where the discussion goes next.
Because fixing friction is only the beginning.
Chapter 3 explores the next layer:
- longitudinal care models
- digital GI
- diagnostics
- chronic disease ownership
- patient journey expansion
- and how GI practices can create value far beyond the procedure room.
The future of GI belongs to practices that redesign workflows around the intelligent use
of human capacity.