What One Family's Heartbreaking Journey Reveals About Healthcare's Addiction to Fixing Problems Too Late
“The prevailing view of artificial intelligence (AI) in
medicine is that it will support physician-led care.” — Ezekiel
J. Emanuel, MD, PhD, and colleagues, JAMA, August 17, 2026
In Busselton, Western Australia, Sinead and Alex Forster
learned something no parent should ever have to learn: sometimes the phone call
you are praying for never comes.
Their son, Teddy Forster, had a rare genetic heart
condition. He needed a transplant.
The family waited.
And waited.
The heart never came.
Teddy died at just 20 months old.
Years later, Sinead and Alex faced an almost unbearable
twist of fate.
Their younger son, Noah Forster, was diagnosed with
the same condition only days after birth.
The family knew the road ahead.
Specialists.
Monitoring.
Hospitalizations.
Waiting.
At seven months old, Noah developed heart failure.
Eventually, he needed mechanical circulatory support while his family waited
for a donor.
Then the phone rang.
This time, the call came.
A donor heart was available.
Noah received a transplant.
One family experienced two radically different endings to
remarkably similar medical journeys: one child died waiting; another survived
because the right intervention arrived in time.
And that story made me think about something that happens
every day in physician practices.
Not transplantation.
Not cardiology.
Billing.
Yes, billing.
Because healthcare has developed a strange habit.
We wait for the problem to become expensive before we decide
it deserves our attention.
A claim gets denied.
Then we investigate the documentation.
A prior authorization fails.
Then someone starts making phone calls.
A payment disappears.
Then somebody opens a spreadsheet.
A/R gets old.
Then the revenue cycle meeting begins.
Everyone becomes a detective.
Nobody asks why we needed a detective in the first place.
The denial is not the problem.
The denial is the autopsy.
And this is where physician owners should pay attention.
The Healthcare Industry Has a Downstream Addiction
We love downstream solutions.
Denial management.
Appeals.
Coding audits.
A/R cleanup.
Prior-authorization teams.
Revenue-cycle outsourcing.
More dashboards.
More reports.
More people staring at yesterday's mistakes.
There is nothing inherently wrong with these functions.
They are necessary.
But they are often treated as the solution when they are
actually the last line of defense.
Think about a leaking roof.
If you keep buying better buckets, you may become very
efficient at collecting rainwater.
You have not fixed the roof.
Healthcare does this constantly.
We optimize the bucket.
We build software for the bucket.
We hire consultants for the bucket.
Then we congratulate ourselves because the bucket is now
AI-powered.
Meanwhile, the roof is still leaking.
The Physician Usually Sees the Problem First
This is where the conversation becomes uncomfortable.
Physicians often know exactly where the friction is.
You know when documentation requirements are becoming
absurd.
You know when an authorization process is slowing care.
You know when your staff is spending more time fighting the
payer than helping patients.
You know when your practice is losing revenue.
But physicians are rarely given the right operational tools
to correct the problem at the moment it begins.
Instead, the problem travels downstream.
Clinical encounter.
↓
Documentation.
↓
Coding.
↓
Claim creation.
↓
Submission.
↓
Denial.
↓
Appeal.
↓
Payment.
↓
Someone asks:
"What happened?"
That's backwards.
The better question is:
"What should have happened before the claim ever
existed?"
Noah's Story Contains a Lesson About Timing
The most powerful part of Noah Forster's story isn't simply
that he received a transplant.
It's timing.
Early recognition mattered.
Specialized care mattered.
Monitoring mattered.
Coordination mattered.
And ultimately, donor availability mattered.
Healthcare outcomes rarely depend on one magical
intervention.
They depend on a chain of decisions.
Revenue cycle performance works the same way.
A claim doesn't suddenly become bad when the payer denies
it.
The claim carries its history with it.
The clinical encounter.
The documentation.
The diagnosis.
The procedure.
The medical necessity.
The payer rules.
The authorization.
The coding.
The data.
By the time the denial arrives, the opportunity to prevent
the problem may already be gone.
Here's the Contrarian Part
Your billing department may be doing too good a job.
I know.
That sounds ridiculous.
But hear me out.
If your organization has become exceptionally good at:
- fixing
claims
- appealing
denials
- correcting
codes
- chasing
documentation
- calling
payers
- working
aging A/R
you may have accidentally created a system that tolerates
bad upstream data.
The organization learns:
"Don't worry. Billing will fix it."
That's dangerous.
Because billing eventually becomes the shock absorber for
everyone else's workflow problems.
And shock absorbers eventually wear out.
The Real Product Isn't Billing
This is the idea behind OnnX.
The traditional revenue-cycle mindset asks:
How do we process the claim better?
The upstream mindset asks:
How do we create better claim-ready data in the first
place?
That distinction matters.
Because a cleaner claim is not created in the billing
office.
It begins with the encounter.
The right information must be captured.
The right documentation must exist.
The right clinical context must be connected to the right
billing logic.
The right information should move through the workflow
without being repeatedly re-created by humans.
That is where AI can become genuinely useful.
Not as a fancy replacement for a biller.
Not as another chatbot.
Not as a machine that generates impressive summaries nobody
asked for.
But as an early-warning system for revenue risk.
AI Should Not Be the Hero
Healthcare doesn't need another AI superhero story.
The industry has enough of those.
AI should be boring.
Boring is good.
If AI can quietly identify:
"This encounter appears incomplete before
submission."
That's useful.
If it can recognize:
"This procedure may require documentation that isn't
present."
Useful.
If it can flag:
"This claim has characteristics associated with a
high likelihood of denial."
Very useful.
The best healthcare AI may not look like science fiction.
It may look like fewer problems at 4:47 p.m. on Friday.
Physicians understand that kind of innovation.
Three Experts, Three Lessons
1. Atul Gawande: Systems Matter
Gawande's work repeatedly demonstrates that healthcare
outcomes are shaped not only by clinical knowledge but by systems,
processes, and human behavior.
The lesson for physician owners:
Don't simply ask who made the mistake.
Ask what system made the mistake predictable.
2. Eric Topol: Better Data Can Change Medicine
Topol has consistently emphasized the importance of data,
earlier detection, and technology that supports clinicians rather than simply
adding another layer of administration.
The operational lesson is similar:
Data is most valuable before the decision, not after the
failure.
3. Don Berwick: Fix the System, Not the Individual
Berwick's quality-improvement philosophy emphasizes reducing
waste and improving systems rather than blaming individuals.
That matters enormously in medical billing.
When the same denial happens repeatedly, the answer probably
isn't:
"Why aren't our billers working harder?"
It may be:
"Why does our workflow keep producing this
problem?"
That's a very different question.
The Five Questions Every Physician Owner Should Ask
1. Where does our revenue leakage actually begin?
Not where you discover it.
Where does it begin?
2. Which denials repeat?
A one-time error is an error.
A recurring error is a system problem.
3. How much staff time is spent fixing preventable
mistakes?
Calculate it.
The number may be uncomfortable.
4. Which information is captured too late?
If billing has to reconstruct information that should have
existed during the encounter, your workflow has an upstream problem.
5. What could we prevent instead of repair?
This may be the most important question of all.
The "Best Practice" We Should Question
Here's an industry assumption worth challenging:
"Every practice needs a strong denial-management
process."
Yes.
But why stop there?
A better practice should ask:
Why are we producing so many denials that we need such a
strong denial-management process?
That's the difference between optimization and
transformation.
Optimization makes the broken process faster.
Transformation asks whether the process should exist in its
current form at all.
A Simple Exercise for Your Practice
Take your top five denial categories.
Don't start with software.
Start with paper.
For each denial, write down:
Where did the information needed to prevent this denial
originate?
Then ask:
- Was it
available?
- Was it
documented?
- Was it
structured?
- Was it
transmitted correctly?
- Did
someone have to interpret it manually?
- Could
the problem have been identified earlier?
You may discover something surprising.
Your billing department isn't necessarily where your billing
problem starts.
Three Metrics That Matter Upstream
Physician owners tend to watch:
Days in A/R
Collection rate
Denial rate
Those matter.
But they're lagging indicators.
Consider adding:
Documentation Completeness
How often is required information present before submission?
First-Pass Clean Claim Rate
How many claims move through without intervention?
Preventable Denial Rate
Not every denial is preventable.
Separate the preventable from the unavoidable.
That's where the operational opportunity lives.
The Biggest Pitfall: Measuring Activity Instead of
Outcomes
Healthcare loves activity.
How many claims were processed?
How many calls were made?
How many appeals were submitted?
How many accounts were touched?
But activity isn't the same as progress.
A billing team can work incredibly hard while the underlying
system gets worse.
The better question is:
How many problems did we prevent?
That's a very different metric.
The Human Cost of Revenue Cycle Friction
We sometimes talk about denials as if they're merely
numbers.
They aren't.
A denied claim creates work.
That work lands somewhere.
Usually on a human being.
A physician.
A biller.
A practice manager.
A medical assistant.
A front-desk employee.
Someone stays late.
Someone makes another phone call.
Someone sends another fax.
Someone opens another payer portal.
And eventually someone asks:
"Why are we doing this again?"
That's not just an efficiency problem.
It's a workforce problem.
And Here's the Irony
Physicians entered medicine to care for patients.
Then healthcare built an elaborate maze around them.
Now we're telling physicians:
"We need you to spend more time documenting so
someone else can determine whether you deserve to be paid."
Then we wonder why physicians are frustrated.
Maybe the problem isn't that physicians need to work harder.
Maybe the system needs to require less rework.
Myth Buster
Myth: More billing staff means more revenue.
Reality: More people can sometimes compensate for
poor processes. They can also make poor processes more expensive.
Myth: AI replaces billers.
Reality: Good AI should reduce repetitive work while
keeping humans responsible for judgment and oversight.
Myth: Every denial should be appealed.
Reality: Some denials are not worth pursuing.
Prevention and prioritization matter.
Myth: Documentation is the physician's problem.
Reality: Documentation is part of a larger
clinical-operational workflow.
Myth: RCM starts after the patient leaves.
Reality: Revenue cycle begins much earlier.
Legal and Ethical Considerations
There is an important boundary here.
The objective should never be to game reimbursement.
It should be to accurately represent the clinical encounter.
That means:
No upcoding.
No manufactured documentation.
No manipulation of medical necessity.
No automation without appropriate oversight.
AI should help practices identify missing or inconsistent
information.
It should not invent information.
That distinction isn't cosmetic.
It is foundational.
The Ethical Question for Healthcare AI
We should stop asking only:
"Can AI automate this?"
Ask:
"Should AI automate this?"
And then:
"What happens when it is wrong?"
Healthcare AI needs guardrails.
Human oversight.
Audit trails.
Clear accountability.
Privacy protections.
Appropriate security.
And transparency about what the system actually did.
The goal is not autonomous billing at any cost.
The goal is safer, cleaner, more predictable workflows.
A Practical 30-Day Upstream RCM Reset
Week 1: Diagnose
Pull your denial data.
Find your five biggest recurring categories.
Don't blame people.
Find patterns.
Week 2: Trace
Follow each problem backward to the point of origin.
Where did the failure begin?
Week 3: Fix
Change the workflow at the earliest practical point.
Add validation.
Improve documentation prompts.
Remove unnecessary handoffs.
Week 4: Measure
Compare:
- Denial
rate
- Clean
claim rate
- Days
in A/R
- Staff
rework
- Preventable
denials
Then repeat.
Where OnnX Fits
This is the problem I am working on with OnnX.
Not:
"How do we build another billing platform?"
There are already plenty.
The bigger question is:
Can we move revenue-cycle intelligence upstream?
Can technology help a physician practice identify revenue
risk while the clinical and operational information is still being created?
Can we reduce the amount of reconstruction that happens
after the encounter?
Can we make the revenue cycle less dependent on humans
catching yesterday's mistakes?
That's the opportunity.
Recent News: The Bigger Pattern
The Forster family's story provides a powerful human example
of why timing matters in healthcare.
Their experience reminds us that earlier recognition,
specialized care, coordination, and timely intervention can fundamentally
change outcomes.
The same philosophy applies to healthcare operations.
The question isn't whether problems will occur.
They will.
The question is:
How early can we see them?
What Healthcare Leaders May Be Missing
The next major healthcare efficiency gain may not come from
squeezing another 2% out of collections.
It may come from preventing the errors that require
collection work in the first place.
That sounds less exciting.
It also sounds less like a billion-dollar AI pitch.
But it may be much more valuable.
Because the best workflow is often the one that doesn't
require fixing.
The Future of Medical Billing
I don't believe the future is:
Human billers vs. AI.
That's the wrong fight.
The future is:
Bad data vs. good data.
Reactive workflows vs. predictive workflows.
Rework vs. prevention.
Complexity vs. simplicity.
And ultimately:
Downstream correction vs. upstream intelligence.
The technology will change.
The principle won't.
The Question I Want Physician Owners to Consider
What if your practice doesn't actually have a billing
problem?
What if it has a data problem that becomes visible as a
billing problem?
That distinction changes everything.
Because if you're treating the symptom, you'll keep paying
for the symptom.
If you fix the source, the downstream system gets quieter.
And sometimes quiet is the best sign that healthcare
technology is actually working.
Final Thoughts: The Call We Should Want
Teddy Forster's story ended with a call that never came.
No physician could manufacture a donor heart.
No technology could guarantee the right outcome.
But Noah Forster's story reminds us why early
recognition, coordinated care, timely intervention, and human generosity
matter.
There is a parallel lesson for physician owners.
Don't wait for the denial.
Don't wait for A/R to explode.
Don't wait for your staff to burn out.
Don't wait until the financial problem becomes impossible to
ignore.
Find the signal upstream.
Fix the workflow before the failure.
Build a practice that prevents problems instead of
becoming exceptionally good at repairing them.
That is where I believe the next chapter of medical billing
begins.
Get Involved
Here is the question I want to leave with you:
What is one recurring problem in your practice that
everyone has learned to "work around" instead of actually fixing?
Tell me in the comments.
I'd genuinely like to hear what physician owners are seeing
on the ground.
And if this perspective makes you rethink where your
revenue-cycle problems actually begin, share this article with another
physician or clinic owner.
The healthcare industry doesn't need another conversation
about working harder downstream.
Maybe it's time we started moving upstream.
About the Author
Dr. Daniel Cham is a physician, medical consultant,
healthcare entrepreneur, and founder of OnnX, an AI-powered medical
billing SaaS focused on helping small and medium-sized physician practices
reduce administrative friction and improve revenue-cycle performance.
His work focuses on the intersection of clinical
medicine, healthcare operations, medical billing, technology, and practice
sustainability.
Connect with Dr. Cham on LinkedIn and explore his work in
healthcare innovation.
Continue the Conversation
Healthcare improves when clinicians, operators,
entrepreneurs, and patients challenge assumptions and share what they are
learning.
For more practical perspectives on healthcare operations,
medical billing, innovation, physician leadership, and the future of healthcare,
continue the conversation through Dr. Cham's platforms.
Explore the insights:
Dr. Daniel Cham's website
Listen to the podcast:
Dr.
Cham on Spotify
Watch and subscribe on YouTube:
Dr. Cham on YouTube
Follow the conversation on X:
Dr.
Cham on X
Connect on Facebook:
Dr. Cham on Facebook
Knowledge creates leverage. Start with one problem, trace
it upstream, and change the system that created it.
Your next improvement may not require another tool. It
may require asking a better question.
Start upstream.
Free Resource
Check the Featured section of my LinkedIn profile for a
free resource designed for independent physician practices. No signup required.
If this article resonated with you, consider reposting it
so another physician owner can see the problem differently.
#Healthcare #HealthcareLeadership #PhysicianOwner
#MedicalBilling #RevenueCycleManagement #RCM #HealthcareAI #HealthTech
#HealthcareInnovation #PracticeManagement #PhysicianLeadership #MedicalPractice
#HealthcareOperations #MedicalEconomics #DigitalHealth #HealthcareTechnology
#IndependentPractice #OnnX #HealthcareTransformation #PhysicianEntrepreneur
Disclaimer: This article is intended for
general educational and informational purposes and does not constitute medical,
legal, financial, coding, billing, or regulatory advice. Specific circumstances
should be reviewed with appropriately qualified professionals.
