A man in Pontiac found help when he needed it. His story raises a much bigger question: why does healthcare make people work so hard to get something they already know they need?
“Health and research data are often fragmented across
institutions, databases, and systems.” — Venus
Oliva Cloma-Rosales, MD, physician, public-health scientist and
health-data researcher
There is something almost embarrassingly simple about what
saved Tim Phillips.
He walked through a door.
Not into a futuristic hospital.
Not into a $500 million digital-health platform.
Not into an AI-powered clinical command center with
seventeen dashboards.
He walked into a mental-health urgent-care facility in
Pontiac, Michigan.
And someone helped him.
That sounds almost too ordinary to be news.
But perhaps that is exactly why it is worth paying attention
to.
The Moment Before the Door
Tim Phillips is 48.
He was working as a live-in nanny caring for three children
when anxiety, loneliness and the pressures of life became overwhelming.
About two years ago, he reached a moment when he was
contemplating ending his life.
Then he thought about the children.
Instead, he went to CNS Healthcare's mental-health
urgent-care facility in Pontiac.
Karla Heinig, the facility's mental-health urgent-care
supervisor, described what happens after someone walks through the door:
check-in, assessment, a nurse, vital signs, and a prescriber when needed.
Jennifer Shumaker, chief clinical officer of adult services
at CNS Healthcare, identified something remarkably important.
The problem is not always the absence of care.
Sometimes the problem is the distance between needing care
and actually getting it.
Wait times matter.
Access matters.
Timing matters.
Shumaker said walk-in and same-day treatment can help people
get care earlier, before a crisis develops.
When asked how soon someone could be seen, her answer was
strikingly simple:
Immediately.
Phillips later said:
“I’m happy.”
And then:
“I probably wouldn’t be here. I probably would have took my
life.”
His message to other people is equally simple:
“Don’t be afraid to get the help you need. You’re not
alone.”
That is a mental-health story.
It is also a healthcare operations story.
And that second part deserves more attention.
Healthcare Has a Friction Problem
We talk constantly about healthcare innovation.
AI.
Automation.
Interoperability.
Digital transformation.
Ambient documentation.
Remote monitoring.
Predictive analytics.
Agentic workflows.
The vocabulary gets more impressive every year.
But there is a less glamorous word that may explain more
about healthcare than all of them:
Friction.
How much effort exists between a person needing something
and actually getting it?
For Tim Phillips, the question was painfully personal:
Can I get help when I need it?
For a physician:
Can I get the authorization before the patient waits
another week?
For a nurse:
Can I find the information I need without opening six
systems?
For a front-desk employee:
Can I determine whether this patient's insurance is
actually active?
For a coder:
Do I have the documentation required to submit this claim
correctly?
For a practice owner:
Can I understand why revenue is leaking before another
month closes?
Different problems.
Same architecture.
Friction.
And here is the uncomfortable part:
Healthcare has become very good at teaching people to
tolerate it.
We call it workflow.
We call it administrative burden.
We call it payer complexity.
We call it documentation requirements.
We call it operational overhead.
Sometimes those descriptions are accurate.
But they can also become euphemisms.
A badly designed process is still a badly designed process
even after we give it a sophisticated name.
The Strange Economics of Healthcare
Consider one of the strangest economic arrangements in
modern healthcare.
We take highly trained professionals and ask them to
compensate for systems that were poorly designed.
Physicians chase authorizations.
Nurses reconcile information.
Staff call insurance companies.
Coders correct incomplete documentation.
Billers appeal denials.
Practice managers build spreadsheets.
Someone inevitably becomes the unofficial expert in a
process nobody else understands.
And every organization seems to have one.
You know the person.
They have been there for 19 years.
They know which payer portal actually works.
They know which phone number bypasses the automated system.
They know which modifier causes trouble.
They know which fax machine still receives something
important.
Nobody remembers who gave them this knowledge.
Nobody has documented it.
And everyone becomes nervous when they take a vacation.
That is not institutional knowledge.
That is institutional dependency.
We have built healthcare systems where human beings become
the middleware.
We Keep Fixing the Wrong End of the Problem
This is particularly obvious in medical billing and
revenue cycle management.
A claim gets denied.
So we measure the denial.
We calculate the denial rate.
We create a denial work queue.
We assign somebody to investigate it.
Someone writes an appeal.
Someone submits the appeal.
Someone follows up.
Someone escalates it.
Someone updates the spreadsheet.
Then six weeks later, somebody asks:
Why did this happen?
That question should have come first.
A denial is rarely born a denial.
It usually has a history.
Maybe eligibility was never verified correctly.
Maybe the authorization was missing.
Maybe the authorization was obtained for the wrong service.
Maybe documentation did not support the billed service.
Maybe a modifier was missing.
Maybe information was entered differently in two systems.
Maybe the claim was submitted with a mismatch that should
have been caught before submission.
The denial is the visible event.
The cause may have occurred much earlier.
That distinction matters.
Because if you only optimize the denial department, you can
become extraordinarily efficient at processing problems you should have
prevented.
That is not efficiency.
That is high-speed inefficiency.
The AI Paradox
And now we are adding AI.
This could be enormously valuable.
It could also create a spectacular new version of the same
problem.
Imagine this:
AI generates the claim.
Another AI reviews the claim.
A payer AI denies the claim.
Your AI identifies the denial.
Another AI drafts the appeal.
The payer's AI reviews the appeal.
Your AI escalates it.
At some point, two artificial intelligences may be arguing
about a modifier while a human employee sits nearby wondering whether lunch is
still happening.
We should not automatically assume that adding intelligence
to every step creates an intelligent system.
Sometimes it simply creates faster interaction between
broken steps.
The important question is not:
Where can we put AI?
It is:
Which friction should disappear?
That is a very different question.
Fragmented Data Is Not Just a Technology Problem
This is where Dr. Venus Oliva Cloma-Rosales' comment this
week is particularly relevant.
She described health and research data as fragmented across
institutions, databases and systems, and argued that AI can help connect those
sources and turn fragmented information into evidence for better decisions.
That idea extends well beyond research.
The same problem exists inside everyday medical practices.
A practice may have:
- Patient
demographics in one system.
- Eligibility
information somewhere else.
- Authorization
information in a payer portal.
- Clinical
documentation in the EHR.
- Claims
in a practice-management system.
- Denials
in an RCM platform.
- Payments
in another financial workflow.
- Staff
knowledge inside someone's head.
The data exists.
The problem is that the relationship between the data is
often weak.
That is a crucial distinction.
The future of healthcare may not depend simply on collecting
more data.
It may depend on making existing data understandable as a
connected system.
The Rear-View-Mirror Problem
Most billing dashboards tell you what happened.
Claims.
Denials.
A/R.
Payments.
Collections.
Days in A/R.
Useful numbers.
But they are mostly rear-view mirrors.
The more interesting question is:
Why did it happen?
And then:
What is changing upstream?
Imagine a three-provider family practice.
1,200 visits.
1,150 claims.
14% initial denial rate.
42 days in A/R.
$7,500 a month in billing expense.
The dashboard can tell the owner that something is wrong.
But suppose the real problem is not "billing."
Suppose 40% of the denials trace back to three upstream
issues:
Eligibility verification is inconsistent.
Authorization information is not consistently captured.
Documentation requirements are being interpreted differently
by different clinicians.
Now the denial rate is not the diagnosis.
It is the symptom.
The practice doesn't need another report saying:
You have denials.
The owner already knows.
The owner needs the system to say:
Here is where the pattern begins.
That is a fundamentally different kind of intelligence.
A Practice Is a Chain, Not a Collection of Departments
Revenue cycle management is often treated as a department.
That is a mistake.
It is a chain.
Registration → Eligibility → Authorization →
Documentation → Coding → Charge Capture → Claims → Denials → A/R → Payments
The financial outcome at the end of the chain is heavily
influenced by decisions made near the beginning.
Yet organizations often manage each segment separately.
That creates a strange situation.
Everyone optimizes their own box.
Nobody owns the entire journey.
The registration team says:
"We entered the patient correctly."
The authorization team says:
"The authorization was obtained."
The clinician says:
"I documented the encounter."
The coder says:
"I coded what was documented."
The biller says:
"I submitted the claim."
The payer says:
"The claim does not meet requirements."
Technically, everyone may be right.
Financially, the system can still be wrong.
That is the problem with fragmented optimization.
The Hidden Cost Isn't Always the Denial
Imagine a $200 claim that gets denied.
People often focus on the $200.
But the real cost may be much larger.
Someone investigates it.
Someone calls.
Someone searches the chart.
Someone checks the payer portal.
Someone sends a message.
Someone corrects the claim.
Someone resubmits it.
Someone tracks the result.
Someone follows up again.
The organization has now consumed human time to recover
money that should not have required recovery.
So perhaps the more useful metric is not simply:
How many claims were denied?
It is:
How much unnecessary work did those claims create?
That metric changes the conversation.
Because a healthcare organization can have acceptable
financial performance while quietly exhausting its workforce.
The Physician Is Not the Backup System
This problem extends beyond billing.
Physicians are often the final workaround for broken
systems.
A task does not get completed?
Ask the physician.
A form is confusing?
Ask the physician.
A prior authorization needs clarification?
Ask the physician.
A patient cannot navigate the portal?
Ask the physician.
An alert fires?
Ask the physician.
Another message arrives?
Ask the physician.
Eventually, the physician becomes the human equivalent of a
software patch.
That is dangerous.
Not because physicians cannot do these things.
They can.
But because every workaround consumes attention.
And attention is one of the scarcest resources in
healthcare.
A physician who spends ten minutes solving a systems problem
is not simply spending ten minutes.
That is ten minutes removed from clinical reasoning, patient
communication, teaching, recovery, or another task that actually requires a
physician.
We should stop treating physician availability as infinite
infrastructure.
The Physician Burnout Conversation Needs a New Question
We talk about physician burnout as if the solution is mostly
personal resilience.
Sleep.
Mindfulness.
Boundaries.
Wellness programs.
Those things can matter.
But there is another question:
How much unnecessary work are we asking physicians to
absorb?
A September 2026 article on healthcare AI adoption noted
that physician skepticism toward new technology is understandable after years
of systems that promised efficiency while adding clicks, alerts and
administrative responsibilities.
That is an important observation.
Physicians are not necessarily resistant to technology.
They may be resistant to more work disguised as
technology.
There is a difference.
Prior Authorization Is the Perfect Example
The American Medical Association's latest physician survey
found that physicians reported an average of 40 prior authorizations per week.
Nearly one-third said requests are often or always denied. The survey also
found that 94% said prior authorization contributes to burnout, while physician
and staff time devoted to prior authorization averaged 13 hours per week.
Those numbers are not merely an argument for faster
authorization.
They are an argument for asking why so much human labor is
required in the first place.
And there is an important distinction:
Automating a burden is not the same as removing the
burden.
If a bad process takes 13 hours and automation reduces it to
six, that is an improvement.
But the better question is:
Why are we spending six hours?
That question is harder.
It requires redesign.
It requires standardization.
It requires better information.
It requires changing incentives.
And sometimes it requires admitting that a process everyone
has accepted for years is simply not very good.
The New Healthcare Arms Race May Be AI Versus AI
There is already growing use of AI in administrative
healthcare workflows.
Recent CAQH findings reported by AJMC indicate that AI
adoption is increasingly concentrated in high-burden administrative functions,
including prior authorization, fraud detection and documentation.
That trend makes sense.
Administrative work is repetitive.
It generates data.
It follows patterns.
It is expensive.
It is therefore attractive territory for automation.
But there is a trap.
If providers automate against payer automation, and payers
automate against provider automation, we could create an administrative arms
race.
The humans may disappear from the middle.
The complexity will not.
That is not the future we should aim for.
The objective should be fewer unnecessary interactions, not
simply faster interactions.
What Tim Phillips Teaches Us About Healthcare Design
This is where his story becomes unexpectedly powerful.
Tim Phillips did not need healthcare to become more
technologically sophisticated in that moment.
He needed healthcare to become accessible.
The difference between:
"I need help"
and
"I can actually get help"
is friction.
That same principle applies across healthcare.
A patient experiences:
Need → access
A physician experiences:
Clinical decision → action
A nurse experiences:
Problem → information
A practice experiences:
Care delivered → payment received
Whenever unnecessary steps sit between those points,
friction grows.
And friction has consequences.
Sometimes the consequence is frustration.
Sometimes it is lost productivity.
Sometimes it is delayed care.
Sometimes it is burnout.
Sometimes it is lost revenue.
Sometimes, as Phillips' story reminds us, the stakes can be
much more serious.
The Contrarian Idea
Here is the contrarian idea I keep coming back to:
Healthcare may not need more efficiency nearly as much as
it needs less unnecessary work.
Those sound similar.
They are not.
Efficiency asks:
How can we do this faster?
Elimination asks:
Why are we doing this at all?
Healthcare technology has spent decades getting better at
the first question.
The next generation should become much better at the second.
What Should We Actually Measure?
If I were auditing a medical practice today, I would look
beyond traditional revenue-cycle metrics.
I would still measure:
- Initial
denial rate
- Final
denial rate
- Days
in A/R
- Clean
claim rate
- Net
collection rate
- Authorization
turnaround
- Eligibility
failure rate
- Claim
correction rate
- Payment
variance
- Revenue
leakage
But I would add another category:
Friction Metrics
1. Rework rate
How often does the same case have to be touched again?
2. Manual intervention rate
How many cases require a human workaround?
3. Information retrieval time
How long does staff spend finding information that should
already be available?
4. Exception rate
How often does the standard workflow break?
5. Repeat-contact rate
How many times must someone call, message or resubmit?
6. Unnecessary work hours
How much staff time is consumed correcting preventable
problems?
7. Upstream error rate
How many downstream failures can be traced to an earlier
stage?
Those measurements tell a different story.
They begin to reveal the cost of complexity.
A 30-Day Friction Audit
You do not need a massive transformation project to begin.
Start with 30 days.
Pick five recurring problems.
For each one, ask:
Where did the problem first appear?
Not where was it discovered.
Where did it begin?
Then ask:
What information was missing?
Who had to compensate?
How many times was the case touched?
What could have prevented it?
What could be standardized?
What should never require human intervention again?
That last question is particularly important.
Because the goal of automation should not be to create a
robot that performs unnecessary work.
The goal should be to make the unnecessary work disappear.
What Not to Automate
This may be the most important technology lesson.
Do not automatically automate a broken process.
First understand it.
Then simplify it.
Then standardize it.
Then automate what remains.
Otherwise you risk building:
Garbage in → faster garbage out.
Or, in the healthcare version:
Ambiguity in → automated ambiguity out.
AI is extraordinarily good at scaling processes.
That is precisely why we should be careful about what we ask
it to scale.
If the underlying process is undefined, AI can scale
confusion.
The OnnX Question
This is where my thinking about OnnX has changed.
The opportunity is not simply to build another billing tool.
There are already plenty of tools that tell practices what
happened.
The more interesting opportunity is to understand why it
happened.
And then connect:
Problems → Causes → Actions → Outcomes
That means looking across the entire revenue cycle.
Eligibility.
Authorization.
Documentation.
Coding.
Claims.
Denials.
A/R.
Payments.
Not as eight isolated modules.
As one connected system.
The goal is a living Practice Intelligence layer.
Something that can say:
"Your denial rate increased."
Useful.
But incomplete.
A more valuable system would say:
"Your denial rate increased 3.2 percentage points over
the last 60 days. Most of the increase traces to eligibility mismatches in two
payer categories. Those errors began upstream at registration. Here is the
recurring pattern. Here is the action. Here is whether the action worked."
Now we are moving from reporting to intelligence.
From rear-view mirror to windshield.
The Difference Between a Dashboard and Intelligence
A dashboard displays information.
Intelligence creates context.
A dashboard says:
Denials: 14%
Intelligence asks:
Why?
A dashboard says:
A/R: 42 days
Intelligence asks:
Which patients, payers, services or upstream failures are
driving the increase?
A dashboard says:
Authorization delays: 9 days
Intelligence asks:
Which step creates the delay, how often does it occur,
and what can prevent it?
The distinction sounds subtle.
It is not.
One tells you what happened.
The other helps you understand what to do next.
The Ethical Question
There is also an ethical dimension to all of this.
Whenever someone says:
"We made the process more efficient," I want to
ask:
Efficient for whom?
The payer?
The physician?
The practice?
The patient?
The billing company?
The employee?
The algorithm?
Those interests do not always align.
A system that reduces payer administrative cost by creating
more work for a physician has not necessarily created healthcare efficiency.
A system that reduces staff time but makes patients navigate
three additional portals has moved friction.
It has not eliminated it.
Healthcare needs to measure friction from the perspective of
the entire system.
Especially the person at the receiving end.
The Legal Question
There is also a legal and operational reality.
Automation does not eliminate accountability.
If software makes a mistake involving eligibility,
authorization, coding, documentation or claims, "the algorithm did
it" is unlikely to be a satisfying operational answer.
Healthcare organizations still need:
- Appropriate
human oversight
- Clear
responsibility
- HIPAA-compliant
handling of protected health information
- Appropriate
business associate agreements where required
- Audit
trails
- Access
controls
- Validation
of automated outputs
- Defined
escalation procedures
- Coding
and billing oversight
- Clear
documentation of system limitations
The more consequential the decision, the more important the
human accountability layer becomes.
The objective is not to remove humans from healthcare.
It is to stop wasting human intelligence on work machines
and better-designed processes can handle.
Three Experts. Three Lessons.
Venus Oliva Cloma-Rosales, MD
Her observation this week about fragmented health and
research data points toward a larger opportunity: connect fragmented
information so it becomes useful for decision-making.
Lesson: Data becomes more valuable when relationships
between data become visible.
Ryan Sadeghian, MD
In a recent Healthcare IT News discussion about clinical AI
adoption, Sadeghian argued that physician skepticism toward technology is
understandable after years of tools that promised efficiency while adding
clicks and administrative work.
Lesson: The best technology does not ask physicians
to tolerate another layer of technology.
Stephen J. Morgan, MD
Morgan emphasized involving clinicians in the design and
implementation of new tools and listening to the actual problems users
encounter.
Lesson: If you want adoption, start with the
problem—not the technology.
Put those three lessons together and something interesting
emerges:
Connect the data. Start with the real problem. Keep
humans involved in judgment.
That is a much more useful AI strategy than simply asking
where the next chatbot belongs.
The Future of Medical Billing May Be Less Billing
That sounds strange coming from someone building a medical
billing company.
But I believe it.
The future of revenue cycle management should not be about
becoming better and better at recovering preventable mistakes.
It should be about preventing the mistakes.
The best denial is not the denial that gets appealed
successfully.
It is the denial that never happens.
The best A/R intervention is not the one that collects a
120-day-old balance.
It is the one that prevents the balance from becoming 120
days old.
The best automation is not the robot that works all night.
It is the process that no longer needs the work.
That is a very different definition of innovation.
The Real AI Opportunity
The healthcare AI conversation often starts with:
"What can AI do?"
I think we should start somewhere else.
What should no longer require human effort?
That question is harder.
It forces us to understand workflows.
It forces us to understand incentives.
It forces us to understand clinical reality.
It forces us to understand data quality.
It forces us to understand where errors originate.
And it forces founders to spend time with the people
actually doing the work.
That is less glamorous than launching another AI demo.
It is also where some of the most valuable healthcare
companies may be built.
One Open Door
Go back to Tim Phillips.
He was overwhelmed.
He needed help.
The healthcare system could have told him to call someone
else.
It could have given him a number.
It could have asked him to complete another form.
It could have told him the next appointment was three weeks
away.
Instead, he walked through a door.
Someone assessed him.
Someone listened.
Someone connected him with care.
That is not a small thing.
It is a design principle.
When someone is ready to take the next step, the system
should not make that step unnecessarily difficult.
That principle applies to mental healthcare.
Primary care.
Emergency care.
Specialty care.
Prior authorization.
Medical records.
Patient access.
Revenue cycle management.
Everything.
The Question I Would Ask Every Practice Owner
If Tim Phillips walked through the door of your practice
today, what would happen next?
Not just clinically.
Operationally.
Could you find his information?
Could you verify his coverage?
Could you determine what he needed?
Could you document it correctly?
Could you obtain anything required before care?
Could you submit the resulting claim cleanly?
Could you explain what happened if payment was delayed?
Could your team do all of that without relying on one person
who "just knows how things work"?
That is the real test.
Because a healthcare system is not defined only by what it
can do.
It is defined by how much unnecessary friction stands
between a person and what they need.
Final Thoughts
We do not have a shortage of healthcare technology. We have
a shortage of healthcare systems designed around the elimination of unnecessary
work.
The next generation of healthcare AI should not merely make
broken workflows faster; it should make some workflows unnecessary.
And perhaps the simplest measure of innovation is this: How
much easier did we make it for the next person to get what they need?
Your Turn
What is the most unnecessary piece of work your medical
practice performs every week?
Not the biggest problem.
Not the most expensive problem.
The most unnecessary one.
The task everyone accepts because "that's just how
healthcare works."
I would genuinely like to hear it.
Leave it in the comments.
And if this made you think differently about healthcare
friction, medical billing, revenue cycle management, or AI, share it with
someone who works inside the system.
Sometimes the most valuable healthcare innovation begins
with one uncomfortable question:
Why are we still doing this?
A Practical Resource for Independent Practices
If you run an independent medical practice, start with one
exercise:
Write down the five tasks your staff complain about most.
Then ask one question about each:
Could we prevent this problem instead of becoming better
at fixing it?
That question may reveal more about your practice than
another dashboard.
Recent Healthcare News Behind the Discussion
The Tim Phillips story was reported by WXYZ Detroit on
September 21, 2026. Phillips described how accessing walk-in mental-health care
at CNS Healthcare in Pontiac became a turning point, while CNS Healthcare's
Jennifer Shumaker emphasized the importance of same-day access before a crisis
develops.
At the same time, healthcare AI continues moving deeper into
administrative workflows. Current reporting shows increasing use of AI in areas
such as prior authorization, documentation and other high-burden administrative
processes.
That makes the central question even more important:
Are we eliminating friction—or simply automating it?
FAQ
Is healthcare's biggest problem really technology?
No single explanation captures healthcare's complexity.
Technology can solve specific problems, but poorly designed workflows,
fragmented information, conflicting incentives and administrative requirements
can create friction that technology alone does not eliminate.
Why focus on upstream problems?
Because downstream problems are often symptoms.
A denial is discovered at billing, but its cause may have
occurred during registration, eligibility, authorization, documentation or
coding.
Is AI the solution?
AI can be part of the solution. But AI is not automatically
a solution simply because it is automated.
The quality of the underlying process and data still
matters.
Does automation eliminate administrative work?
Sometimes.
But automation can also move work from one person or
organization to another. The right question is whether total unnecessary work
has actually decreased.
Why does this matter to independent medical practices?
Smaller practices often have fewer people available to
absorb operational complexity. One broken workflow can therefore consume a
disproportionate amount of staff time and directly affect cash flow.
What should a practice measure beyond denial rates?
Consider rework, manual intervention, exception rates,
information-retrieval time, repeat contacts and upstream error rates.
Those metrics help expose the hidden cost of complexity.
Myth Buster
Myth: A lower denial rate automatically means a healthier
revenue cycle.
Not necessarily.
A practice may have a reasonable denial rate while losing
significant time to eligibility problems, authorization work, documentation
corrections, claim edits or payment follow-up.
Myth: More automation automatically means more
efficiency.
No.
Automation can accelerate a bad process.
Myth: Physician resistance means physicians dislike
technology.
Not necessarily.
Physicians may be rejecting technology because previous
technology increased clicks, alerts and administrative work.
Myth: The billing department owns the revenue cycle.
The billing department owns an important part of the revenue
cycle.
But revenue-cycle performance begins much earlier.
The Bigger Idea
Healthcare has spent years building systems that answer:
What happened?
The next generation should become much better at answering:
Why did it happen?
And eventually:
How do we prevent it from happening again?
That is the difference between reporting and intelligence.
Between reaction and prevention.
Between treating symptoms and redesigning systems.
And perhaps between adding another healthcare tool and
actually building a better healthcare system.
One Last Thing
Tim Phillips did something incredibly difficult.
He asked for help.
Healthcare's job should be to make that decision easier—not
harder.
Whether the person is a patient walking into a mental-health
clinic, a physician trying to obtain authorization, a nurse searching for
information, or a practice owner trying to understand missing revenue, the
principle is the same:
Remove the unnecessary obstacle.
Structure the information.
Let people spend their time on the work that actually
requires them.
That is what better healthcare should feel like.
About the Author
Dr. Daniel Cham, MD is a physician, healthcare
entrepreneur and medical consultant focused on the intersection of healthcare
operations, medical billing, technology and practical AI. His work explores how
better information structure can reduce administrative friction and help
independent medical practices operate more intelligently.
His current work with OnnX focuses on a simple
premise:
Most revenue-cycle problems begin before the billing
department ever sees them.
Disclaimer
This article is for educational and informational
purposes only. It does not constitute medical, legal, financial, coding or
billing advice. Healthcare organizations should evaluate technology, workflow
changes and automation in the context of their own clinical, regulatory,
contractual and operational requirements.
The discussion of Tim Phillips' experience is based on
publicly reported information and is intended to highlight healthcare-access
and systems-design lessons, not to provide clinical commentary on his
individual circumstances.
Continue the Conversation
LinkedIn: Connect with Dr.
Daniel Cham
Website: Dr.
Daniel Cham
Spotify: Listen to the
podcast
YouTube: Watch
on YouTube
X: Follow on X
Facebook: Follow on
Facebook
References
- WXYZ
Detroit — Tim Phillips and CNS Healthcare, September 21, 2026
The primary source for the human story about Tim Phillips, Karla Heinig, Jennifer Shumaker, and same-day mental-health access. WXYZ/BHNet — Tim Phillips’ story - Dr.
Venus Oliva Cloma-Rosales — Voice of Asia, September 21, 2026
Current physician/public-health scientist discussing fragmented health data and the role of AI in connecting information for better decisions. Voice of Asia — Dr. Venus Oliva Cloma-Rosales - American
Medical Association — 2026 Prior Authorization Physician Survey
Survey of 1,000 practicing physicians: 40 prior authorizations per week on average, 13 hours of physician/staff time, 94% reporting that prior authorization contributes to burnout, and only 33% believing the latest insurer pledge will make a meaningful difference. AMA — Prior Authorization Physician Survey
#Healthcare #HealthcareAI #MedicalBilling
#RevenueCycleManagement #HealthcareOperations #PhysicianBurnout
#PriorAuthorization #HealthTech #IndependentPractice #PracticeManagement
#HealthcareInnovation #AIinHealthcare
