One family’s extraordinary decision in Kolkata inspires a contrarian question for healthcare leaders: What if the next major breakthrough in medical billing is not doing more work faster, but eliminating the work that should never have existed?
“I'm looking for technology to enhance the science of medicine so that I can lean into the art of medicine.” — Judd Semingson, president and CEO of Community Clinic, as reported by Axios
Three Lives. One Family’s Decision. A Question Healthcare
Needs to Answer.
In October 2026, a family in West Bengal, India, faced a
loss no family wants to experience.
Tamal Chakraborty, 59, of Dharampur, Chinsurah, in
Hooghly district, suffered an intracranial hemorrhage and was taken to Apollo
Multispeciality Hospitals in Kolkata.
Doctors later confirmed brain death.
Then came a decision that would change other families’
lives.
His wife, Soma Chakraborty, and son, Sagnik
Chakraborty, consented to organ donation. According to The Times of
India, Tamal’s liver and kidneys were transplanted into three patients. His
corneas were also collected and preserved at Disha Eye Hospitals. The
recipients were not identified by name in the report.
Sagnik explained that the family had heard about organ
donation but did not know how to navigate the process. When the hospital team
approached them, his mother was the first to decide that her husband’s organs
should help others survive.
Consider the weight of that moment.
A family was navigating grief and uncertainty. A medical
team had to communicate clearly. Information had to reach the right people.
Multiple clinical teams and institutions had to coordinate their work.
And amid an unimaginable loss, three patients received the
possibility of a new beginning.
The family’s generosity deserves the attention. The
clinicians and transplant teams deserve recognition. This story is not evidence
that administrative inefficiency affected Tamal’s care, and organ donation
should never be reduced to a business-process analogy.
But the story does invite a broader question about
healthcare systems.
When the stakes are high, we understand that
communication, accurate information, and coordination matter. Why do we
tolerate so much preventable confusion in everyday healthcare administration?
In medical billing, a missing detail can trigger a
correction. A poorly defined handoff can trigger another phone call. An
incomplete record can send a claim into a cycle of rework.
The consequences are different from those of organ
transplantation, but the operational question remains worth asking: Could
better preparation and clearer processes prevent avoidable problems before they
become everyone else’s emergency?
That is where healthcare innovation needs a rethink.
The Contrarian Question: What If We Are Making the Wrong
Work Faster?
Healthcare has developed a serious appetite for speed.
Faster documentation. Faster coding. Faster claim
submission. Faster denial analysis. Faster appeals. Faster follow-ups.
Now AI promises to accelerate even more of the process.
Some of these advances are genuinely useful. Automation can
reduce repetitive tasks, identify inconsistencies, and help staff find relevant
information. AI can support documentation, classification, and analysis when it
is implemented with appropriate safeguards.
But speed is not the same as improvement.
Imagine a practice that repeatedly submits claims with
incomplete information. It buys software that identifies missing information
sooner.
That is progress.
Now imagine another practice redesigns its intake and
documentation processes so the required information is captured correctly
before the claim is submitted.
That may be an even greater improvement.
The first practice becomes better at detecting errors.
The second reduces the conditions that create them.
Both approaches can be necessary. But they solve different
problems.
One improves the response. The other improves the process.
What if healthcare’s next big innovation is not
automating more work, but preventing unnecessary work from being created in the
first place?
That question is less glamorous than an AI product
demonstration. Nobody is likely to applaud a redesigned intake field or a
clearer handoff between registration and billing.
But preventing a recurring error can be more valuable than
repeatedly processing it.
Healthcare organizations can become remarkably good at
compensating for broken processes. They build queues for unresolved tasks,
dashboards for incomplete records, alerts for delayed responses, and reports
explaining why the earlier reports were wrong.
Eventually, the organization becomes extremely efficient at
managing the consequences of its own design.
It is a little like installing a sophisticated
water-monitoring system while refusing to fix the leaking pipe.
The dashboard is beautiful. The floor is still wet.
Automating unnecessary work is still unnecessary work.
Before buying another tool, healthcare leaders should ask
whether the underlying task is necessary, whether its cause can be removed, and
whether automation will genuinely reduce complexity or simply move it somewhere
else.
The Hidden Price of “Just One More Follow-Up”
Ask a medical billing specialist what happens when required
information is missing.
The answer is rarely, “Nothing much.”
A staff member identifies the gap. Someone contacts the
patient or clinician. The record is updated. Another person checks whether the
update reached the correct system. The claim is corrected and resubmitted.
Someone else tracks the response.
If the correction is incomplete, the cycle starts again.
Each task might take only a few minutes. But those minutes
accumulate, and the true cost extends beyond the clock.
There is context switching. A medical assistant stops
preparing for the next patient to investigate an administrative issue. A
clinician is asked to clarify documentation after the encounter. A billing
specialist must reconstruct events from weeks earlier. A practice manager tries
to determine which step is holding up payment.
The work crosses departments, systems, and people.
Every handoff creates an opportunity for information to be
lost, misunderstood, duplicated, or delayed.
For a small physician-owned practice, the impact can be
especially painful. The person managing billing may also be handling patient
calls. The front-desk coordinator may be checking eligibility, arranging
referrals, and answering the phone. The physician may finish seeing patients
only to start reviewing documentation queries.
The practice was established to deliver healthcare—not to
become a full-time help desk for disconnected systems.
Yet that is what administrative friction can create.
The more interesting question is not whether every follow-up
is necessary.
It is how many follow-ups exist only because an earlier
step failed to provide the information or clarity needed to complete the job?
This is where practices can begin to distinguish necessary
administrative work from preventable rework.
The Numbers Behind the Administrative Problem
The administrative burden in healthcare is not simply a
collection of frustrating anecdotes.
The American Medical Association’s 2025 Prior Authorization
Physician Survey, reported in May 2026, described a persistent burden on
physician practices. The survey included 1,000 practicing physicians.
Among the reported findings:
- 95%
said prior authorization delays access to necessary care.
- 79%
reported that patients abandon treatment because of prior authorization
challenges.
- 26%
said prior authorization had contributed to a serious adverse event for a
patient in their care.
- Physicians
reported completing an average of 40 prior authorization requests per
week.
- Physicians
and staff spent an average of 13 hours per week on prior
authorization work.
- 94%
said prior authorization contributes to physician burnout.
These are survey findings based on physicians’ reported
experiences. They do not establish that every authorization causes harm or that
every practice experiences the same burden. They do, however, illustrate why
administrative processes deserve serious attention.
Source: American
Medical Association — Prior authorization reform survey.
The financial opportunity is substantial, too.
The 2024 CAQH Index estimated that healthcare’s
medical and dental administrative transactions involved approximately $90
billion in annual spending, with an additional $20 billion in potential
annual savings from moving remaining manual transactions to fully
electronic workflows.
These figures concern the administrative transactions
measured by CAQH. They do not represent the entire cost of healthcare
administration or medical billing alone.
Source: CAQH
— 2024 Index report.
The easy conclusion is that administrative inefficiency is
expensive.
The more important conclusion is that digitizing a
process does not automatically make it a good process.
A digital form can still ask the wrong question. An
automated workflow can still move incomplete information from one system to
another. An AI model can still generate a polished summary of a poorly
documented encounter.
Technology can remove friction. It can also disguise
friction beneath a more attractive interface.
A paper form that creates rework is frustrating.
A digital form that creates the same rework is not a
breakthrough. It is the same problem wearing a better outfit.
Garbage In, Beautifully Summarized, Is Still Garbage
Much of the AI conversation in healthcare focuses on what
happens after information has been created.
Can AI read a clinical note? Suggest a code? Summarize an
encounter? Identify a possible denial? Draft an appeal?
These capabilities may be valuable. But a deeper question
comes first.
Was the necessary information captured correctly,
completely, and consistently at the source?
If documentation is incomplete, AI may help identify the
gap. If information is contradictory, it may flag the inconsistency.
But a practice should not expect an AI system to invent
missing facts, silently resolve conflicting records, or substitute plausible
language for evidence.
AI can generate a convincing explanation. Convincing is not
the same as correct.
It can make a process faster. Faster is not the same as
complete.
It can identify patterns in past denials. That does not mean
every future denial can be prevented.
And it can help staff process more information without
necessarily reducing the amount of administrative work.
There is a fundamental difference between using AI to
process information and designing a process that reliably captures the
right information in the first place.
The second challenge is less glamorous. It rarely produces a
spectacular product demonstration.
Nobody is likely to post a viral video celebrating a
well-designed field that prevents a missing insurance identifier. Nobody will
applaud a clean handoff between a medical assistant and a billing specialist.
Yet those improvements can prevent hours of rework.
The right question is not merely, “What can AI do with the
information we have?”
It is also, “Why did we not have the right information when
we needed it?”
This is not an argument against AI. It is an argument for
putting AI in the right place.
The best technology may not be the technology that does
the most work. It may be the technology that makes unnecessary work disappear.
Medical Billing Is Not Just a Billing Problem
Medical billing is often treated as a downstream function.
The encounter happens. Documentation is completed. Codes are
assigned. A claim is submitted. Payment is pursued.
If something goes wrong, the billing team investigates.
This sequence makes billing look like the final stage of the
process.
But many billing problems begin much earlier.
Consider a hypothetical example.
A patient visits a specialist. The appointment is completed,
but insurance information was not verified correctly. A referral document has
not reached the appropriate location. A required detail in the clinical record
is unclear. A staff member assumes someone else has checked the authorization.
The claim goes out.
Then it comes back.
The billing specialist may be excellent at identifying the
denial. The software may correctly classify it. The practice may have a
well-organized follow-up queue.
But the problem may have begun during scheduling,
registration, eligibility verification, authorization, documentation, or a
handoff between these functions.
That distinction changes the questions a practice should
ask.
Instead of asking only, “How do we resolve this denial?”
ask:
- Where
did the missing information first become relevant?
- Who
was responsible for capturing or verifying it?
- Was
the required information available at that point?
- Did
the process make the next action clear?
- Could
the problem have been detected before claim submission?
- If
this issue has occurred repeatedly, what has changed to prevent it from
recurring?
These questions move the conversation from individual blame
to process design.
They also help avoid a common trap: blaming the last person
who touched the claim for a problem created several steps earlier.
A denial is an outcome. It is not necessarily the origin
of the problem.
When practices treat the outcome as the origin, they can
become trapped in a cycle of corrections without meaningful prevention.
Activity Is Not the Same as Completion
Healthcare administration often measures activity because
activity is easy to count.
How many claims were submitted? How many calls were made?
How many tasks were closed? How many denials were appealed?
Those numbers can be useful, but they can also create a
misleading sense of progress.
A submitted claim is not a payment.
An authorization request is not an approval.
A task marked complete is not necessarily a resolved
problem.
An appeal sent is not an appeal won.
A message delivered is not proof that the recipient
understood it or acted on it.
This is not an argument against operational metrics. It is
an argument for measuring outcomes alongside activity.
A practice might track claims submitted each day. It should
also understand how many claims are accepted on the first pass, how many
require correction, how long payment takes, and which recurring issues generate
the most rework.
Likewise, a practice might count completed denial
follow-ups. But the more revealing question is whether the underlying causes of
those denials are becoming less common.
A team can close hundreds of tasks while the same problem
continues to return.
That is not necessarily a staff performance issue. It may be
a signal that the system rewards closure more clearly than prevention.
Imagine a leaking pipe.
One employee spends the day mopping the floor. Another
identifies the leak and fixes the pipe.
Both are busy. Only one has reduced the need for future
mopping.
Healthcare needs fewer celebrations of the mop.
It needs better plumbing.
Seven Ways Physician-Owned Practices Can Prevent Rework
This does not require replacing every system or redesigning
an entire practice overnight. Start with one recurring problem and investigate
where it begins.
Step 1: Trace the problem backward
Choose one frequent administrative issue: a common denial,
missing documentation, an eligibility discrepancy, or a delayed authorization.
Trace it backward from the point where it becomes visible.
What happened immediately before it? What information was
missing? When was that information first needed? Who could reasonably have
captured or verified it?
Do not begin by asking who made a mistake.
Begin by asking why the process allowed the mistake to pass
unnoticed.
Step 2: Measure rework, not just output
Record how often the problem recurs and how much staff time
is spent correcting it.
A practice does not need an elaborate analytics platform to
begin. A carefully maintained spreadsheet may be enough.
Track a small number of measures:
- First-pass
claim acceptance.
- Denials
by reason and payer.
- Missing-information
corrections.
- Days
in accounts receivable.
- Time
spent on repeat follow-ups.
- Percentage
of recurring issues with a documented corrective action.
Define each measure consistently. Interpret changes
alongside payer mix, claim complexity, and other relevant factors.
The goal is not to create another dashboard staff must feed.
It is to discover which problems are worth preventing.
Step 3: Capture information when it first becomes
available
Information is often easiest to obtain at the moment it is
created.
Insurance details may be available during registration.
Referral information may be available before the appointment. Documentation
questions may be clarified while the encounter is still fresh.
Waiting until a claim is rejected can turn a simple
clarification into an investigation.
Capture essential information at the earliest appropriate
point, without burdening clinicians with unnecessary data entry or collecting
information the practice does not need.
The aim is not to ask more questions.
It is to ask the right questions at the right time.
Step 4: Make handoffs explicit
A handoff should not depend on a vague assumption that
someone else will take care of it.
Define what information must accompany the task, who owns
the next step, and what counts as completion.
If a task is blocked, the reason should be visible. If
another person must act, the responsibility should be clear.
This does not require turning every interaction into a
bureaucratic checklist. The goal is to remove ambiguity, not create a new
species of paperwork.
Step 5: Separate preventable errors from unavoidable
complexity
Not every denial is caused by poor data. Payer policies,
coverage exclusions, changing requirements, clinical documentation standards,
and legitimate disputes all matter.
A practice should not assume that every denial can be
eliminated by improving its own process.
Classify problems by cause. Some may be preventable through
better registration or documentation. Others may require payer clarification,
clinical review, an appeal, or a policy change.
This distinction helps teams focus on what they can control
without blaming themselves for every external obstacle.
Step 6: Automate only after examining the process
Automation can reduce repetitive effort, improve
consistency, and help staff identify problems earlier.
Examples may include eligibility checks, claim-status
inquiries, structured data validation, reminders, and routing work to the
appropriate person.
But before automating a process, ask whether the process
itself is necessary.
If the same information is entered into three systems,
automating all three entries may preserve the duplication. If a task exists
because two systems do not exchange information reliably, adding another
task-management tool may not solve the underlying issue.
Map the workflow first.
Remove unnecessary steps where possible.
Then automate the steps that still need to happen.
Step 7: Check whether the problem actually declined
After changing a process, measure what happened.
Did the same denial category become less frequent? Did
correction time decrease? Did first-pass acceptance improve? Did staff spend
less time chasing missing information?
If the metric did not improve, investigate why.
A new workflow is not a success simply because it has
launched. A software implementation is not a success simply because employees
have been trained to use it.
The test is whether the change improves the work and the
outcome.
The ultimate measure of a better process is not how
efficiently the team manages the problem. It is whether the problem becomes
less likely to return.
A Practical 30-Day Experiment
Small practices do not need to wait for a major technology
project to begin improving administrative work.
A focused 30-day experiment can reveal useful opportunities.
Week 1: Find the recurring problem.
Select one frequent source of rework. Review a sample of
recent cases and document where each problem first appeared. Use de-identified
or appropriately authorized information for the analysis.
Week 2: Identify the earliest practical intervention.
Look for a missing field, unclear responsibility, delayed
verification, inconsistent process, or failed handoff. Ask the staff who
perform the work what actually happens rather than relying only on the written
procedure.
Week 3: Test one change.
Make a limited adjustment. This might be a clearer intake
requirement, an earlier verification step, a more explicit handoff, or a rule
that flags incomplete information before submission.
Avoid introducing multiple changes simultaneously if doing
so would make the results difficult to interpret.
Week 4: Compare the results.
Review the same measures used in the baseline. Did the
problem occur less frequently? Did the change create extra work elsewhere? Did
it introduce new risks or delays?
Keep the change if the evidence supports it. Modify it if
the results are mixed. Stop if it makes the process worse.
This is not glamorous. It is disciplined operational
learning.
And it may be a better starting point than buying software
in the hope that a vendor will somehow discover how the practice actually
works.
Five Questions to Ask Before Buying Another AI Tool
AI can be useful, but physician-owned clinics should
evaluate it against the actual problem they need to solve.
1. Which specific problem will this solve?
“Improve efficiency” is too vague. Identify the task,
failure point, or measurable outcome.
2. Does the problem originate in the workflow or in the
information?
If information is incomplete at the source, a tool that only
processes it downstream may not address the root cause.
3. What happens when the system is uncertain?
Healthcare administration involves consequential
information. A useful system should make uncertainty and exceptions visible
rather than quietly treating a guess as a fact.
4. How will the practice measure success?
Agree on baseline measures before implementation. Include
error rates, staff time, rework, and meaningful financial outcomes where
appropriate.
5. What work will the tool add?
New technology brings implementation, training, oversight,
integration, security, and maintenance requirements. Those costs belong in the
evaluation.
The right question is not whether a product uses AI.
It is whether the product improves a process that
matters—and whether the improvement outweighs the new complexity.
Efficiency Must Never Become an Excuse to Cut Corners
Preventing rework is important, but efficiency is not a
reason to compromise accuracy, privacy, or compliance.
Medical billing and practice operations involve protected
health information, documentation requirements, payer contracts, and applicable
federal and state laws.
Practices should consider several safeguards:
- HIPAA
and privacy: Access to protected health information should be
appropriate to the task. Vendors handling protected health information may
require appropriate agreements and safeguards.
- Documentation
integrity: Automation must not invent clinical facts, silently alter
records, or encourage documentation unsupported by the encounter.
- Billing
compliance: Claims must accurately reflect the services provided and
applicable coding and coverage requirements.
- Human
oversight: Staff should be able to review uncertain results, correct
errors, and escalate exceptions.
- Transparency:
Practices should understand how automated decisions are made, what
information is used, and where the system has limitations.
- Security
and accountability: Technology should support clear responsibility for
decisions, corrections, and incident response.
The details depend on the tool, workflow, payer, and
circumstances. Practices should seek qualified compliance and legal advice when
needed.
The objective is not to make healthcare administration
faster at any cost.
It is to make the work more accurate, less wasteful, and
more accountable.
The Human Cost of Administrative Noise
It is easy to describe administrative burden in hours,
dollars, denial rates, and days in accounts receivable.
Those measures matter. But they do not fully capture the
human experience.
A patient may never see the spreadsheet used to track a
missing authorization. A family may not know how many calls were needed to
resolve an administrative issue. A physician may not mention the time spent
correcting documentation after clinic hours.
Yet these tasks consume attention.
And attention is finite.
When clinicians and staff spend more time navigating
avoidable administrative problems, less time remains for work that requires
human judgment, compassion, and expertise.
That does not mean every administrative task is wasteful.
Verification, documentation, authorization, and accurate billing can serve
legitimate purposes. Nor does it mean billing inefficiencies caused or affected
Tamal Chakraborty’s organ donation story. There is no basis for making that
claim.
The connection is broader: healthcare depends on people and
systems coordinating around important information.
Tamal’s family was able to make a meaningful decision after
communication with the hospital team helped them understand the possibility of
donation. That is a reminder of the human importance of clarity and
coordination.
In routine practice operations, the same principles deserve
attention, applied in a very different context.
Clear information helps people act. Explicit
responsibilities reduce ambiguity. Well-designed handoffs make it easier to
complete the next step.
Preventing avoidable administrative work can give people
more time to focus on the work that truly requires them.
What OnnX Believes: Fix the Source, Not Just the Symptom
As a physician-entrepreneur and founder of OnnX, an
AI-powered medical billing SaaS focused on small and medium-sized
physician-owned clinics, I believe the industry needs to ask a different set of
questions.
We should not only ask how to code claims faster, pursue
denials more efficiently, or automate repetitive revenue-cycle tasks.
We should also ask:
- Why
did the information arrive incomplete?
- Why
did the process depend on someone remembering an undocumented step?
- Why
did the same error require multiple people to correct it?
- Why
was the problem not visible until the claim reached the payer?
- What
can we eliminate before we decide what to automate?
This is not an argument that technology cannot help with
downstream work. It can.
It is an argument that better downstream tools cannot
compensate indefinitely for poor upstream information and process design.
The premise behind OnnX is that healthcare billing is not
only a tooling problem. It is also a data-quality and process-design problem.
The earlier information is captured correctly, the clearer
the handoffs, and the more visible the exceptions, the less unnecessary rework
the practice may need to manage later.
That is the direction worth exploring: fewer preventable
errors, less repetitive follow-up, more reliable information, and a clearer
view of where a problem actually begins.
The objective should not be to make the billing team heroic
at rescuing broken claims.
It should be to create fewer claims that need rescuing.
The Future of Healthcare Innovation May Be Less About
Doing More
Healthcare innovation often celebrates what a new system can
do.
Can it generate a note? Summarize a chart? Identify a denial
pattern? Automate another step?
These are reasonable questions.
But perhaps the next generation of meaningful innovation
will also be judged by what it makes unnecessary.
How many duplicate entries disappear?
How many avoidable corrections never occur?
How many handoffs become simpler?
How many hours of repetitive follow-up are returned to
staff?
How often does the right information reach the right person
at the right time?
And how much administrative complexity can be removed
without compromising safety, accuracy, or accountability?
This is a more demanding standard than simply measuring
software adoption.
It asks technology companies, healthcare organizations, and
practice owners to demonstrate that the work itself has improved.
In the end, healthcare does not need every administrative
task to become an AI-powered task.
Some tasks need better tools.
Some need clearer ownership.
Some need better information.
And some need to disappear.
Final Thought: Stop Celebrating the Mop
Tamal Chakraborty’s story is a reminder that extraordinary
human outcomes can emerge when families and healthcare teams navigate difficult
moments with compassion, communication, and coordination.
His family’s decision helped three people receive
transplanted organs. Their generosity should remain at the center of the story.
For those of us working on healthcare systems, the story
also invites a wider reflection: the quality of a result often depends on what
happens before the decisive moment.
In medical billing, that means looking beyond the denial,
beyond the correction, and beyond the next follow-up.
It means asking whether we are fixing the problem—or merely
becoming more efficient at managing its consequences.
Perhaps healthcare’s next big innovation will not be the
system that produces the most work in the least time.
Perhaps it will be the system that knows which work should
never have existed.
And perhaps the most impressive dashboard will not show how
many tasks we completed.
It will show how many preventable tasks we no longer need.
A provocative question for healthcare leaders and
physician-owners: If AI can help us automate almost any repetitive task,
shouldn't we first ask whether that task deserves to exist?
Share your perspective in the comments: What is one
recurring administrative task in your practice that should be eliminated—not
merely automated?
If this perspective resonates, repost it to encourage a more
practical conversation about healthcare innovation, medical billing, and the
work we can prevent.
About the Author
Dr. Daniel Cham is a physician-entrepreneur and
founder of OnnX, an AI-powered medical billing SaaS focused on small and
medium-sized physician-owned clinics. His work explores medical billing,
revenue-cycle management, healthcare technology, data quality, and practical
ways to reduce administrative friction so practices can focus more of their
attention on patient care.
Connect and follow
·
Listen to the
podcast on Spotify
·
Subscribe
and watch on YouTube
·
Follow
Dr. Cham on Instagram
For a free resource, visit the Featured section of my
LinkedIn profile. No signup required.
References
- Yengkhom,
Sumati. “Year’s 13th organ donation in city saves three lives.” The
Times of India, October 12, 2026. Read
the report
- Axios.
“Axios Live: How health care providers can reach patients in an era of
mistrust.” October 12, 2026. Read
the report
- American
Medical Association. “AMA survey: Prior authorization reform pledge
falls short with physicians.” May 2026. Read
the report and survey findings
- CAQH.
2024 CAQH Index. Administrative transaction costs and potential
savings from automation. Read
the report
Disclaimer: This article is for general
educational and commentary purposes. It does not provide medical, legal,
billing, coding, or compliance advice. The organ donation story is used as a
human-interest starting point for a broader discussion of coordination and process
design; no claim is made that billing or administrative inefficiencies affected
the events described.
#HealthcareInnovation #MedicalBilling
#RevenueCycleManagement #HealthcareAI #PhysicianEntrepreneur
#MedicalPracticeManagement #PhysicianBurnout #HealthcareOperations #DataQuality
#ClaimDenials #AdministrativeBurden #DigitalHealth #OnnX
No comments:
Post a Comment