A three-generation nursing legacy reveals a more uncomfortable truth: healthcare doesn’t just pass down values. It passes down systems, habits, workarounds—and everything we’ve quietly decided is “just how healthcare works.”
“Healthcare doesn’t lack information; it lacks connection.” — Demetri Giannikopoulos, Forbes contributor and health AI leader
That sentence is intentionally provocative.
It is also a useful place to begin.
Because healthcare has always been an industry where yesterday’s assumptions become tomorrow’s rules.
And eventually, those rules become so familiar that nobody remembers they were assumptions in the first place.
That brings me to Patricia “Pat” Pendleton.
In Lancaster, California, Patricia Pendleton became a nurse at 40 after her husband was diagnosed with cancer.
She raised three children.
She worked for 25 years at Antelope Valley Medical Center.
Her daughter, Terryl Call, watched her mother care for her father and saw what nursing looked like up close. Terryl eventually became a nurse herself and now serves as a house supervisor at the same hospital.
Then came the next generation.
Terryl’s daughters, Tayler Rodriguez and Amanda Starler, followed her into nursing.
They went through nursing school together.
Today, both are registered nurses working in the emergency department at Antelope Valley Medical Center.
Three generations.
Four women.
One hospital.
A family legacy that nobody had to manufacture.
That is the part of the story that matters.
Patricia didn’t need a PowerPoint presentation titled “The Strategic Importance of Nursing as a Multigenerational Career.”
She didn’t need a culture committee.
She didn’t need a motivational poster.
Her family simply watched her.
And then they decided what they saw was worth carrying forward.
Antelope Valley Medical Center describes the family as representing more than 40 years of combined nursing service. Recent reporting has highlighted the unusual three-generation path and the family’s connection to the Lancaster community.
And suddenly, a beautiful family story becomes a much more uncomfortable question for everyone else in healthcare:
What Are We Teaching the Next Generation Without Saying a Word?
Because people watch.
Medical students watch.
Residents watch.
Nurses watch.
Medical assistants watch.
Front-desk employees watch.
Young physicians watch.
Future practice owners watch.
And they learn healthcare less from what we say than from what we repeatedly do.
They see what happens when a patient waits.
They see what happens when an authorization is missing.
They see what happens when an insurance claim is denied.
They see whether leadership fixes the problem or blames the employee.
They see whether technology makes work easier or merely creates another login.
They see whether physicians spend their evenings practicing medicine or fighting administrative fires.
They see whether the organization solves problems—or becomes exceptionally good at working around them.
And eventually, they internalize the answer.
This is what healthcare is.
That may be our biggest legacy problem.
Not that healthcare is broken.
But that we have become very good at teaching people how to live with the broken parts.
The Most Dangerous Sentence in Healthcare
I have a candidate.
It isn’t:
“Your claim was denied.”
It isn’t:
“Authorization is required.”
It isn’t:
“Your insurance isn’t active.”
It isn’t even:
“Please fax it again.”
It is:
“That’s just how healthcare works.”
Those six words have probably protected more dysfunctional processes than any regulation ever could.
That’s just how prior authorization works.
That’s just how credentialing works.
That’s just how billing works.
That’s just how payers work.
That’s just how documentation works.
That’s just how the EHR works.
That’s just how physicians work.
That’s just how the front desk works.
Really?
Or is that simply how we have gotten used to working around the problem?
There is a difference.
A very expensive difference.
We Don’t Just Pass Down Values. We Pass Down Workarounds.
Patricia Pendleton passed down something valuable.
A sense of service.
A commitment to patients.
A belief that nursing mattered.
But healthcare organizations pass down other things too.
Workarounds.
If your practice requires a spreadsheet because the EHR cannot reliably tell staff something important, the spreadsheet becomes part of the culture.
If employees maintain private lists because nobody trusts the official system, those lists become part of the culture.
If staff know a payer’s secret phone number because the normal process doesn’t work, that phone number becomes institutional knowledge.
If one employee knows how to “get the claim through,” everyone starts relying on that person.
And eventually, the workaround becomes invisible.
It is no longer called a workaround.
It is called:
“Our process.”
That’s when dysfunction gets interesting.
Because once a workaround becomes a process, someone eventually builds software around it.
Then someone sells consulting around it.
Then someone creates a dashboard around it.
Then someone adds AI.
And 10 years later, we’re proudly automating the workaround we should have eliminated.
Healthcare has a remarkable talent for turning temporary fixes into permanent infrastructure.
We should probably stop congratulating ourselves for this.
The Great Healthcare Paradox
We have more healthcare technology than any previous generation.
We have electronic health records.
Cloud infrastructure.
Interoperability frameworks.
APIs.
Clearinghouses.
Patient portals.
Revenue-cycle platforms.
Decision-support systems.
AI scribes.
AI coding.
AI claims tools.
AI prior authorization.
AI everything.
And yet a remarkably simple question can still bring an office to its knees:
“What happened to this patient’s authorization?”
Someone checks the EHR.
Someone checks the payer portal.
Someone checks email.
Someone searches a fax.
Someone calls the payer.
Someone asks the front desk.
Someone checks a spreadsheet.
Someone eventually finds a PDF.
Everyone celebrates.
The authorization has been found.
This is not a technology problem in the conventional sense.
It’s a systems architecture problem.
The information existed.
The problem was that nobody could reliably connect it to the work that depended on it.
Healthcare Doesn’t Have an Information Problem
It has a connection problem.
A medical record can contain enormous amounts of information while still failing to answer the question someone needs answered right now.
The patient’s information exists.
The insurance information exists.
The referral exists.
The authorization exists.
The encounter exists.
The documentation exists.
The code exists.
The claim exists.
The payer response exists.
The payment exists.
The denial exists.
But if these things behave like isolated islands, humans become the bridge.
And humans are very expensive bridges.
Recent reporting on healthcare data fragmentation makes this problem increasingly difficult to ignore. One September 2026 analysis described the medical record as potentially scattered across organizations while the person it describes sits in the waiting room.
That is a perfect metaphor for healthcare technology.
The data is everywhere.
The patient is still waiting.
Now Let’s Talk About AI
This is where I want to be deliberately contrarian.
The healthcare industry doesn’t need more AI simply because AI exists.
It needs better problems for AI to solve.
We have reached a strange point where organizations sometimes ask:
“Where can we add AI?”
I think the better question is:
“Where are we wasting human judgment on repetitive, predictable work?”
Those are very different questions.
The first produces AI features.
The second produces useful systems.
And the difference matters.
Because AI applied to clean, connected information can be extraordinarily powerful.
AI applied to fragmented information can become an extraordinarily fast way of producing confident confusion.
That’s not intelligence.
That’s turbocharged ambiguity.
And healthcare already has enough ambiguity without giving it a GPU.
The AI Question Is Not “Can It?”
It’s:
“Should it?”
And:
“Based on what?”
And:
“Can we prove why it did that?”
Those questions are becoming more important as AI moves deeper into clinical and administrative workflows.
In a September 8 Atlantic essay, Ezekiel J. Emanuel and Vinod Khosla argued that medicine should be willing to experiment with increasingly autonomous AI rather than automatically assuming humans must retain the final decision in every circumstance. They also acknowledge that real-world evidence and long-term evaluation remain important.
That debate is fascinating.
But it exposes a deeper principle.
Healthcare should not be afraid of automation.
Healthcare should be afraid of unaccountable automation.
If an AI system recommends an action, can we see what information it used?
Can we understand what happened?
Can we reconstruct the decision?
Can a human intervene?
Can the organization audit the result?
Can the patient be protected if the system is wrong?
Those questions matter whether AI is deciding a diagnosis or prioritizing a billing exception.
The Most Valuable AI May Be the AI You Barely Notice
Forget the flashy demo.
Imagine a system that simply tells the staff:
“This claim has a high probability of denial. The problem appears to originate from the eligibility record. Fix it before submission.”
That’s not sexy.
No robot doctor.
No hologram.
No digital avatar.
No AI-generated inspirational speech.
Just:
“Hey. This looks wrong.”
That’s useful.
Now imagine it can also tell you:
- what looks wrong;
- why it looks wrong;
- who owns the next action;
- what information is missing;
- what deadline matters;
- what happened last time;
- what payer rule applies;
- and what evidence supports the recommendation.
Now AI becomes more than prediction.
It becomes operational intelligence.
Medical Billing Has a Dirty Secret
Here is my contrarian thesis:
Medical billing is not primarily a billing problem.
It is often a data-quality and workflow-state problem that becomes visible in billing.
By the time a claim denies, the original mistake may be ancient history.
The patient’s insurance changed.
Nobody captured it.
The referral expired.
Nobody knew.
The authorization existed.
Nobody connected it to the encounter.
The documentation was incomplete.
Nobody caught it before coding.
The payer changed its rule.
Nobody updated the workflow.
Then the claim denied.
And suddenly the billing department gets the blame.
That’s convenient.
It is also frequently wrong.
Billing Is Often the Crime Scene, Not the Crime
Think about that for a moment.
A denial is an event.
But it isn’t necessarily the origin of the problem.
The denial is often where the organization discovers the problem.
That is different.
Imagine a water leak.
You notice the ceiling stain.
You can spend all afternoon painting the ceiling.
The stain disappears.
Congratulations.
The roof is still leaking.
That is how some revenue-cycle strategies work.
Denial occurs.
Research denial.
Appeal denial.
Correct denial.
Resubmit claim.
Collect payment.
Celebrate recovery.
Then do it again next month.
We call this revenue-cycle management.
Sometimes it is closer to revenue-cycle archaeology.
We’re digging through the remains of decisions made weeks earlier.
What If the Denial Is the System’s Teacher?
This is where healthcare could become much smarter.
A denial shouldn’t merely create work.
It should create learning.
If the same payer repeatedly denies the same scenario, the system should notice.
If the same authorization mistake occurs repeatedly, the system should notice.
If the same eligibility error appears repeatedly, the system should notice.
If a particular workflow creates predictable rework, the system should notice.
The question isn’t:
“How many denials did we resolve?”
It is:
“How many future denials did we prevent because we learned from the last one?”
That is a completely different measurement philosophy.
The Front Desk May Be Part of Your Revenue Cycle
Here’s another idea that makes people uncomfortable:
Your revenue cycle may begin before the patient ever sees the physician.
At registration.
At scheduling.
At eligibility.
At referral.
At authorization.
At intake.
That’s where information enters the system.
And information has consequences.
If the wrong payer is entered, downstream teams inherit the error.
If eligibility is misunderstood, downstream teams inherit the error.
If the referral isn’t captured, downstream teams inherit the error.
If authorization isn’t connected to the correct encounter, downstream teams inherit the error.
Then someone in billing gets the message:
“Can you fix this?”
Why?
The billing person didn’t create the information.
They simply discovered its consequences.
Stop Asking Billing to Be the World’s Most Expensive Detective Agency
Billing professionals are extraordinarily valuable.
But they should not have to reconstruct the entire patient journey every time a claim goes wrong.
They should not need forensic skills to answer basic operational questions.
They shouldn’t need:
- five browser tabs;
- three passwords;
- a spreadsheet;
- an old fax;
- a sticky note;
- and Linda.
You know Linda.
Every practice has a Linda.
Linda knows everything.
Linda remembers which payer changed its portal.
Linda knows which fax number works.
Linda knows which physician forgot to sign something.
Linda knows that the “official” process is not actually the process.
Linda is wonderful.
Linda is also a single point of institutional failure.
If your revenue cycle collapses when Linda takes a vacation, you don’t have a resilient system.
You have a very experienced employee holding your architecture together.
Hero Employees Are Not Infrastructure
Every organization loves the hero employee.
The person who can solve anything.
But heroics are often evidence of weak systems.
A mature organization should not require extraordinary memory to perform ordinary work.
The knowledge should be:
structured, visible, connected, reproducible and auditable.
Otherwise, the organization is renting its operational memory from an employee.
And employees eventually retire.
Or move.
Or get promoted.
Or take two weeks off and discover the office has forgotten how to breathe.
The Next Generation Is Watching This Too
This brings us back to Patricia Pendleton.
Her family watched her work.
But imagine a different story.
Imagine a young physician watching their mentor spend every evening fighting denials.
Imagine a resident watching nurses spend hours on administrative work.
Imagine a medical assistant watching staff repeatedly enter the same information.
Imagine a young practice owner learning that the secret to survival is knowing which payer representative to call.
What do they conclude?
Maybe:
“This is medicine.”
That’s dangerous.
Because medicine isn’t supposed to be a professional endurance contest.
Healthcare should not require extraordinary resilience to survive ordinary operations.
We Keep Calling Burnout a People Problem
Maybe sometimes it is.
But we should be suspicious of any system that repeatedly exhausts good people and then tells them to become more resilient.
A useful system should not depend on heroic endurance.
The current healthcare conversation increasingly recognizes that workforce problems are tied to organizational design, staffing, administrative burden and operational conditions.
Even current debates around AI implementation are exposing a related problem: technology can be introduced without adequately involving the people who actually use it.
For example, recent reporting on the integration of ChatGPT capabilities into Epic has raised questions from nurses about whether frontline nursing voices were adequately represented in safety evaluation and implementation.
That’s an important warning.
You cannot redesign healthcare from the conference room alone.
The people doing the work have to be part of designing the work.
Don’t Automate the Workflow Until You Understand the Workflow
This should be printed on every healthcare technology pitch deck.
Before automating something, ask:
Why does this workflow exist?
Then:
Does it actually need to exist?
Then:
What information does it require?
Then:
Where does that information originate?
Then:
Who owns the next step?
Then:
What happens when something goes wrong?
Then:
Does the system learn from that failure?
Only after answering those questions should someone ask:
“Where should AI go?”
Otherwise, you’re decorating the problem.
The Five-Layer Test
I like to evaluate healthcare workflows through five questions.
1. Data
Do we have the right information?
2. Context
Do we know what that information means?
3. State
Do we know what has happened and what is currently waiting?
4. Ownership
Do we know who is responsible for the next action?
5. Feedback
Does the system learn when the outcome is wrong?
Most healthcare technology focuses heavily on the first question.
Some addresses the second.
Far fewer systems properly address all five.
And that’s where the opportunity becomes interesting.
Healthcare Data Is Not the Same Thing as Healthcare Intelligence
A database can tell you:
Patient has insurance.
A useful system should tell you:
Patient’s insurance is active, but this procedure requires authorization, the authorization expires Friday, the current encounter is scheduled for Monday, and nobody owns the renewal.
That’s the difference between data and operational intelligence.
One is a record.
The other is a decision.
The Connected Data Graph
This is why I believe healthcare workflows need to move toward a connected data model.
Think about the patient journey as a graph:
PATIENT
↕
INSURANCE
↕
ELIGIBILITY
↕
REFERRAL
↕
AUTHORIZATION
↕
ENCOUNTER
↕
DOCUMENTATION
↕
CODING
↕
CLAIM
↕
PAYER
↕
PAYMENT
↕
DENIAL / EXCEPTION
↕
RESOLUTION
↺
FEEDBACK LOOP
The important concept isn’t the boxes.
It’s the relationships.
A referral should mean something in relation to a patient.
An authorization should mean something in relation to a service.
A claim should reflect what actually happened.
A denial should connect back to the event that caused it.
A resolution should feed the next decision.
That’s how a system develops memory.
The Revenue Cycle Should Have a Memory
Not just document storage.
Operational memory.
It should know:
What happened?
When did it happen?
Who handled it?
What information was available?
What policy applied?
What was missing?
What changed?
What was attempted?
What failed?
What worked?
What should happen next?
And if the same situation happens again, the system shouldn’t start from zero.
That’s the absurdity of many administrative workflows.
We make the same mistake.
Then we solve it.
Then three weeks later we make the same mistake again.
And somehow call this experience.
A 30-Day Experiment for Any Independent Practice
If you are a physician owner, don’t start by buying software.
Start with observation.
Days 1–5: Follow One Patient
Choose one patient journey.
Follow:
Scheduling → Registration → Eligibility → Referral → Authorization → Encounter → Documentation → Coding → Claim → Payment
Write down every handoff.
Don’t fix anything yet.
Just observe.
You will probably discover something uncomfortable.
Days 6–10: Circle Every Handoff
Every time information moves from:
- person to person;
- system to system;
- spreadsheet to EHR;
- fax to staff;
- email to billing;
- payer portal to practice software;
circle it.
Those are your friction points.
Days 11–15: Find the Rework
Ask:
What did someone have to do twice?
Then:
What did someone have to verify because they didn’t trust the first answer?
Then:
What did someone have to search for?
Rework is a gold mine.
It tells you where the system is creating unnecessary labor.
Days 16–20: Find the Earliest Failure
For every denial or exception, ask:
When could we first have known?
Before scheduling?
At registration?
At eligibility?
Before the encounter?
Before coding?
Before claim submission?
At adjudication?
This is where the concept of upstream prevention becomes practical.
Days 21–25: Assign Ownership
Every recurring exception needs:
Owner.
Trigger.
Deadline.
Next action.
Escalation path.
If nobody owns it, it will eventually become everyone’s problem.
Which usually means it becomes nobody’s problem until the patient complains.
Days 26–30: Measure Recurrence
Don’t just count how many problems were fixed.
Count how many returned.
That’s the number that tells you whether the system learned anything.
The Metrics That Matter
A practice should know more than total collections.
Track:
Clean claim rate
First-pass acceptance rate
Preventable denial rate
Eligibility error rate
Authorization failure rate
Denial resolution time
A/R over 90 days
Staff rework hours
Revenue leakage
Exception recurrence
The last one may be the most revealing.
Because a system that repeatedly creates the same exception is not improving.
It is simply becoming faster at cleaning up after itself.
The Most Expensive Resource in Healthcare Isn’t Always Money
It’s attention.
A physician’s attention.
A nurse’s attention.
A biller’s attention.
A practice manager’s attention.
A patient’s attention.
When a physician spends 20 minutes figuring out why an authorization disappeared, that is not merely 20 minutes.
It is 20 minutes removed from clinical work.
When a nurse spends an hour hunting for information, that’s not merely an administrative expense.
It’s attention that could have gone toward patients.
When a patient spends an afternoon trying to understand a bill that shouldn’t have been confusing, that’s not simply a customer-service problem.
It’s trust being consumed.
Administrative friction has a human cost.
We simply don’t put that cost on the spreadsheet.
The Problem With “Efficiency”
Healthcare loves the word.
But we often define efficiency incorrectly.
If a team used to process 1,000 exceptions and now processes 2,000, did we become more efficient?
Maybe.
Or maybe we became more efficient at creating and processing exceptions.
The better question is:
Did the amount of unnecessary work decrease?
That’s real efficiency.
The best workflow may be the workflow you no longer need.
The best notification may be the notification that never has to be sent.
The best denial may be the denial that never occurs.
The best billing intervention may be the one that happens before billing.
That is the uncomfortable logic of prevention.
Three Myths Worth Killing
Myth #1: “More Staff Will Fix It”
Sometimes you need more staff.
But if staff are repeatedly fixing the same preventable problem, hiring more people can simply increase the cost of dysfunction.
You are adding more people to the bucket instead of fixing the hole.
Myth #2: “Our EHR Has the Data”
Having data is not the same as having usable, connected, trusted data.
A system can contain 10,000 fields and still force someone to call Linda.
Myth #3: “AI Will Replace Billing”
That’s the wrong question.
The better question is:
Which work should humans never have been doing manually in the first place?
AI should help people spend less time hunting, sorting, checking and repeating—and more time exercising judgment where judgment actually matters.
The AI Revolution May Be Less Revolutionary Than We Think
There is a funny irony happening in healthcare.
We are talking about autonomous AI doctors while many practices still struggle to reliably determine whether a patient is eligible for coverage.
We are debating whether AI can outperform physicians while staff are still faxing documents.
We are discussing artificial general intelligence while someone is maintaining a spreadsheet called:
FINAL_FINAL_INSURANCE_LIST_v7.xlsx
Maybe the future isn’t waiting for us.
Maybe the future is simply waiting for healthcare to fix its plumbing.
Because sophisticated intelligence sitting on top of disconnected infrastructure is still disconnected intelligence.
The New Competitive Advantage: Predictability
Independent practices don’t necessarily need more complexity.
They need predictability.
Predictable eligibility.
Predictable authorization.
Predictable documentation.
Predictable claims.
Predictable follow-up.
Predictable cash flow.
Predictable accountability.
Predictable exception handling.
That doesn’t mean every payer behaves predictably.
It means the practice should be able to distinguish:
What we can control.
What we cannot control.
What we should have known earlier.
What we need to do now.
That distinction is enormously valuable.
This Is Where OnnX Comes In
This is the problem I believe deserves a different architecture.
OnnX is built around the idea that healthcare billing is fundamentally a data-quality and workflow problem that begins upstream.
The objective isn’t simply to create another billing dashboard.
It is to connect the information and workflow states that determine whether the work gets paid.
Patient.
Insurance.
Eligibility.
Referral.
Authorization.
Encounter.
Documentation.
Coding.
Claim.
Payer.
Payment.
Denial.
Resolution.
The system should be able to identify risk before the claim becomes a problem.
It should understand ownership.
It should understand dependencies.
It should understand deadlines.
It should preserve traceability.
And when something goes wrong, it should feed the lesson back into the workflow.
Not just recover revenue.
Prevent the next failure.
That distinction is the whole game.
Don’t Build a Smarter Dashboard
Build a Smarter System.
Dashboards tell you what happened.
Useful systems help determine:
What is happening?
Why is it happening?
What matters?
Who owns it?
What should happen next?
When does it become urgent?
What can we learn from it?
That’s the difference between reporting and orchestration.
And healthcare needs more orchestration.
Less archaeology.
The Legal and Ethical Question
As healthcare systems become more automated, accountability becomes more important—not less.
If an AI system flags a claim, the practice should understand why.
If an automated workflow prioritizes an authorization, there should be a record of what information informed that decision.
If a system changes a workflow, someone should know.
If an automated recommendation is wrong, there needs to be a way to investigate.
“AI decided” cannot become the healthcare equivalent of:
“The computer made me do it.”
Technology should create more traceability, not less.
That is especially important in healthcare because administrative decisions can affect access, payment, compliance, patient experience and clinical operations.
The future cannot simply be automated.
It must be auditable.
The Ethics of Automation
There is another principle worth remembering:
Don’t automate a bad process just because you can.
If a workflow is wrong manually, automating it makes the wrong workflow faster.
If an approval rule is poorly designed, AI can apply the poor rule at scale.
If the underlying data is wrong, automation can amplify the mistake.
Before automating:
Understand the process.
Identify the decision.
Validate the data.
Define accountability.
Test the exception cases.
Then automate.
Not the other way around.
What Patricia’s Family Really Teaches Us
Patricia Pendleton didn’t need to tell her daughter:
“Terryl, here is the five-year plan for becoming a nurse.”
She lived the example.
Terryl didn’t need to write a strategy memo for Tayler and Amanda.
They watched.
That’s how culture works.
And that’s how organizational culture works too.
People watch what leaders do when things become difficult.
Do leaders blame?
Do they hide?
Do they improvise?
Do they fix?
Do they listen?
Do they redesign?
Do they say:
“That’s just how it works.”
Or do they say:
“Why does it have to work this way?”
That sentence may be the beginning of innovation.
The Legacy Test
Every physician owner should run this test.
Imagine a medical student joins your practice tomorrow.
They stay for five years.
They watch everything.
They see the technology.
They see the workflows.
They see the meetings.
They see the billing.
They see the patient complaints.
They see the staff frustrations.
They see how leadership behaves.
At the end of five years, what do they believe healthcare is supposed to look like?
Would you be proud of that answer?
Or would you quietly say:
“Well, you know… healthcare is complicated.”
That’s where the conversation gets interesting.
Because “healthcare is complicated” can be true.
It can also be an excellent excuse.
We Don’t Need More Heroes
Healthcare already has enough heroes.
Patricia Pendleton is a hero.
Terryl Call is a hero.
Tayler Rodriguez is a hero.
Amanda Starler is a hero.
They show up.
They care.
They serve.
They represent something worth preserving.
But we should be careful about designing systems that require heroic people to compensate for ordinary failures.
We don’t need physicians to become more heroic.
We need systems to become less hostile to the people using them.
We don’t need nurses to become infinitely resilient.
We need organizations to stop manufacturing unnecessary friction.
We don’t need billing professionals to become better detectives.
We need systems that stop losing the evidence.
The Future of Healthcare May Be Less About Doing More
And more about making less necessary.
Less rework.
Less duplication.
Less searching.
Less faxing.
Less manual verification.
Less waiting.
Less guessing.
Less chasing.
Less “Who has this?”
Less “Did anyone call them?”
Less “Can you send that again?”
Less “I thought you had it.”
Less “The system says something different.”
Less “That’s just how healthcare works.”
That may not sound revolutionary.
It is.
Because every unnecessary step removed from healthcare returns something valuable:
time.
And time is what clinicians never seem to have enough of.
Final Thoughts: Systems Become Legacy
Patricia Pendleton started nursing at 40.
Her daughter watched.
Her granddaughters watched.
And three generations later, they are still caring for patients.
That is a beautiful legacy.
But there is another kind of legacy being created inside every medical practice.
The workflows.
The habits.
The shortcuts.
The workarounds.
The tolerances.
The technologies.
The assumptions.
The things nobody questions anymore.
Those are being passed down too.
Systems create habits.
Habits become culture.
Culture becomes legacy.
So perhaps the question isn’t simply:
What kind of physician do I want to be remembered as?
Perhaps it is:
What kind of healthcare system will people remember me for building?
A system where staff spend their days fighting administrative fires?
Or one where the fire prevention system actually works?
A practice where billing is an endless archaeological expedition?
Or one where the path from care to payment is visible?
A workplace where technology adds work?
Or one where technology quietly removes it?
A culture where people say:
“That’s just how healthcare works.”
Or one where somebody has the courage to ask:
“Why?”
And then actually fix it.
Three Actions Worth Taking
Stop measuring how efficiently your practice cleans up problems. Start measuring how many problems never happen.
Stop asking where to add AI. Start asking where unnecessary human work is being created.
Stop thinking only about the legacy you leave in patients. Think about the system you leave in the hands of the people who come after you.
Get Involved: The Question I Want Physicians to Answer
Here’s the question I would genuinely like to hear from physicians and practice owners:
What is the most ridiculous administrative process in your practice that everyone has simply accepted as normal?
Not the polished answer.
The real answer.
The spreadsheet nobody trusts.
The fax nobody likes.
The payer portal everyone hates.
The authorization nobody can find.
The report nobody reads.
The claim everyone knows will deny.
The task everybody does twice because nobody trusts the first result.
Tell me in the comments.
Then ask yourself:
If the next generation copied the way your practice operates today, would healthcare actually get better?
If the answer is no, that’s not a reason for embarrassment.
It’s a starting point.
Step into the conversation.
Challenge the assumption.
Raise your hand.
Be willing to redesign what everyone else has learned to tolerate.
And if this article made you think about someone on your team who is quietly carrying a broken process every day, share it with them.
Maybe they aren’t the problem.
Maybe they are the person holding the system together.
Frequently Asked Questions
Is Patricia “Pat” Pendleton a real person?
Yes. Patricia “Pat” Pendleton is the grandmother in the three-generation nursing family at Antelope Valley Medical Center in Lancaster, California. She became a nurse at age 40 after her husband’s cancer diagnosis and spent 25 years at the hospital.
Who are the four women?
The family consists of Patricia “Pat” Pendleton, Terryl Call, Tayler Rodriguez and Amanda Starler. Terryl is Pat’s daughter. Tayler and Amanda are Terryl’s daughters and Pat’s granddaughters. Both Tayler and Amanda became registered nurses and work in the emergency department at Antelope Valley Medical Center.
Did Pat force the next generation into nursing?
No. The family has described the path as something that developed naturally from observing the previous generation rather than through an explicit mandate. That distinction is important: example can be more powerful than instruction.
What does a nursing family have to do with medical billing?
More than it initially appears.
The story illustrates how behavior, values and systems are learned through observation. Healthcare organizations pass down not only professional ideals but also workflows, habits and workarounds.
Is medical billing really a data-quality problem?
Often, but not exclusively. Some problems are payer-driven, contractual or genuinely unpredictable. The argument is that practices should distinguish preventable information and workflow failures from genuine payer-side exceptions rather than treating every denial as the same problem.
Will AI replace medical billers?
That should not be the objective. The more useful question is which repetitive, predictable tasks can be automated so skilled people can focus on exceptions, judgment, relationships and complex problem-solving.
Why is upstream prevention important?
Because the farther downstream a problem is discovered, the more expensive it often becomes to correct.
A wrong insurance record at registration can eventually become a denied claim, a delayed payment, staff rework and potentially a frustrated patient.
What should practices measure?
At minimum:
Clean claim rate.
First-pass acceptance.
Preventable denial rate.
Eligibility errors.
Authorization failures.
Denial resolution time.
A/R aging.
Staff rework hours.
Revenue leakage.
Exception recurrence.
What is the biggest misconception about healthcare efficiency?
That efficiency means processing more work faster.
Sometimes the best efficiency gain is eliminating the work altogether.
Myth Busters
Myth: “More staff means better operations.”
Reality: More staff can increase capacity, but it doesn’t necessarily fix the process generating the work.
Myth: “More AI means more innovation.”
Reality: AI applied to a bad workflow can simply automate the bad workflow.
Myth: “If the information exists in the EHR, the problem is solved.”
Reality: Information must also be current, connected, contextualized and actionable.
Myth: “Denial management is the same as revenue-cycle improvement.”
Reality: Denial management is often downstream recovery. Revenue-cycle improvement should also ask why the denial happened.
Myth: “The person who fixes the problem owns the problem.”
Reality: The person who discovers the problem may simply be the final link in a chain that started much earlier.
The Bigger Question
Patricia Pendleton’s family legacy is inspiring because it happened organically.
Nobody needed to convince three generations that caring for people mattered.
They saw it.
They lived around it.
They absorbed it.
And eventually, they chose it.
Healthcare organizations are doing the same thing every day.
They are teaching the next generation what healthcare is.
Not through speeches.
Through systems.
Through workflows.
Through behavior.
Through technology.
Through what they tolerate.
Through what they celebrate.
Through what they refuse to question.
And perhaps that is the most provocative lesson of all:
The next generation will inherit more than our medical knowledge.
They will inherit our operating system.
So let’s make sure it’s worth inheriting.
What are you passing down?
About the Author
Dr. Daniel Cham is a physician and medical consultant with expertise in medical technology consulting, healthcare management, and medical billing. He focuses on delivering practical insights that help professionals navigate complex challenges at the intersection of healthcare and medical practice.
Connect with Dr. Cham on LinkedIn to learn more and continue the conversation.
Important Note
This article provides general educational and professional commentary on healthcare operations, technology, administrative workflows and medical billing. It is not legal, medical, financial, coding, compliance or reimbursement advice. Specific circumstances should be evaluated with the appropriate qualified professionals.
Continue the Conversation
The conversation continues beyond this article.
Explore Dr. Cham’s broader work, including his healthcare and technology discussions across LinkedIn, Spotify, YouTube, X and Facebook.
Listen to the podcast on Spotify
Knowledge drives progress. Start your journey here.
Free Resource
A related free resource is available through the Featured section on LinkedIn.
No signup required.
Read it. Challenge it. Share it with your team.
And decide what your practice should stop tolerating.
References
Antelope Valley Medical Center — Three Generations, One Calling
The hospital’s account of Patricia Pendleton, Terryl Call, Tayler Rodriguez and Amanda Starler provides the primary source for the family’s nursing history and multigenerational connection.
The Atlantic — “Medicine Needs to Get Serious About AI”
Ezekiel J. Emanuel and Vinod Khosla’s September 8, 2026 essay provides the current AI quote and examines the emerging debate over physician-controlled versus increasingly autonomous AI in medicine.
Patients Know Best — September 2026 Profile
The September 2026 profile examines fragmented medical records and the effort to connect information across healthcare organizations, reinforcing the article’s broader argument about information fragmentation.
#Healthcare #HealthcareInnovation #Physicians #MedicalBilling #RevenueCycleManagement #HealthcareAI #AIinHealthcare #HealthTech #MedicalPractice #PracticeManagement #HealthcareLeadership #HealthcareOperations #PhysicianLeadership #IndependentPractice #RCM #MedicalTechnology #HealthcareTechnology #DenialsManagement #PriorAuthorization #HealthcareTransformation
Leave a Reply