Dr. Sameen Farooq's account reveals healthcare's dangerous blind spot: we measure whether an authorization is approved, but do we measure whether the decision arrives while it can still help?
“To the patient, all of this looks like one delay. To the practice, each is a different failure point requiring a different solution.” — Comron Saifi, MD, orthopedic spine surgeon, KevinMD
Healthcare has a peculiar superpower: It can solve an administrative problem after the opportunity to use the solution has disappeared.
A request is denied. Someone gathers more documentation. A
physician makes a phone call. An appeal follows. A decision is reconsidered.
Eventually, the answer changes from no to yes.
The dashboard turns green.
Everyone can breathe again.
Except, sometimes, the patient.
On October 16, 2026, New York City physician Dr. Sameen
Farooq, MD, described a dying patient whose planned transfer for
end-of-life symptom management was delayed after an insurer denied coverage for
“lack of medical necessity.”
A peer-to-peer review and appeal followed. About a week
later, the denial was reversed.
But the patient's condition had deteriorated. Her blood
pressure had fallen into the low 70s, and Dr. Farooq considered transportation
unsafe. She died on the hospital's medical-surgical unit.
The patient, her daughter, the hospital, and the insurer
were not named in the published account.
We should respect those limits rather than manufacture
identities to make the story more dramatic.
The facts reported by Dr. Farooq are already sobering
enough.
The insurer eventually approved the care. The
circumstances had changed.
The paperwork caught up. The patient did not.
This account does not independently establish that the
denial caused the patient's death or that a different process would have
changed the outcome. It does, however, raise a question healthcare should not
dismiss:
Can a decision be administratively successful and
operationally useless?
The world's most expensive game of telephone
Consider what happens when a clinical decision must travel
through a fragmented administrative system.
A physician documents the patient's condition. A staff
member translates that information into a request. The request goes to a payer
or utilization-management organization. A reviewer evaluates it against
applicable criteria. If the request is denied, someone must understand the
reason, obtain any additional information, and decide whether to appeal.
Each handoff creates an opportunity for delay, confusion, or
missing information.
Now add a patient whose condition is changing while everyone
waits.
The clinical situation is moving. The administrative request
may not be.
It's like tracking a moving vehicle with a photograph taken
last Tuesday.
The photograph might be accurate. It might even be
beautifully documented. But it cannot tell you where the vehicle is now.
Healthcare does not need to eliminate every administrative
review. Coverage rules, medical necessity, patient safety, and responsible
resource allocation matter.
It does need to recognize that a request is not a static
object when the patient's condition is changing.
A process that cannot incorporate meaningful new information
risks answering yesterday's question while everyone is trying to manage today's
reality.
Three clocks. One patient. No pause button.
In an October 13, 2026, KevinMD article, orthopedic spine
surgeon Comron Saifi, MD, described prior authorization as three
distinct operational clocks: completing the clinical record, preparing a clean
submission, and waiting for the payer's decision.
He wrote:
“To the patient, all of this looks like one delay. To the
practice, each is a different failure point requiring a different solution.”
That is a useful distinction.
Clock one: Clinical documentation. How long does it
take to complete the relevant record?
Clock two: Submission readiness. How long does it
take to assemble the required information, route the request correctly, and
submit it?
Clock three: Payer review. How long does it take to
reach a decision after submission?
These clocks have different owners and different remedies.
An unfinished note may require a better workflow. A request
sent to the wrong organization may require better routing. A complete request
that remains under review may require payer follow-up or escalation.
Measuring only the total number of days tells us that the
process was slow. It does not necessarily tell us why.
And without understanding why, organizations risk applying
the wrong solution to the right problem.
You cannot fix a routing error by hiring more appeal
writers. You cannot fix a genuine clinical disagreement by making the printer
faster.
The patient experiences one delay. The organization needs
to identify every avoidable interval inside it.
The contrarian idea: A denial may be the receipt, not the
original problem
The healthcare revenue cycle tends to become highly
attentive after something goes wrong.
A claim is rejected. A document is missing. A payer requests
clarification. A denial arrives.
Suddenly, everyone is investigating.
The billing team opens another portal. The physician
receives another message. Someone searches for a document that everyone
remembers seeing but nobody can find.
The administrative detective story begins.
But what if the missing information could have been
identified before submission? What if the required documentation had been
easier to locate? What if a handoff had been completed correctly the first
time?
These are important questions, although Dr. Farooq's account
does not establish that a documentation or data-quality defect caused the
denial in his patient's case.
The broader operational principle is this:
Some failures become visible downstream even though their
contributing conditions existed upstream.
A denial is an event that can be counted. The underlying
process weakness may be harder to see.
It could be inconsistent information, unclear
responsibility, an incomplete request, a routing error, or a genuine
disagreement about coverage. Those are different problems, and they should not
be lumped together.
The mistake is treating every denial as the same kind of
failure.
Sometimes an appeal is exactly what is needed. Sometimes the
answer is better documentation. Sometimes the payer's decision deserves
challenge. Sometimes the process needs redesign.
The first job is to find out which situation you are dealing
with.
Otherwise, healthcare risks becoming exceptionally good at
repairing the same problem repeatedly.
The green dashboard problem
Imagine a management dashboard showing the following:
- Authorization
initially denied.
- Appeal
submitted.
- Decision
reversed.
- Case
closed.
From an administrative perspective, that may be a successful
resolution.
But what does “case closed” actually mean?
Was the intended care delivered? Was it still clinically
appropriate? Did the patient remain eligible for the planned service? Did
another barrier appear after approval?
A reversed denial tells us that the decision changed. It
does not, by itself, tell us what happened next.
Nor does every reversal prove that the original decision was
unreasonable. An appeal may contain new evidence or clarify an initially
ambiguous request.
The point is not to presume fault. It is to distinguish
administrative outcomes from the outcomes they are intended to support.
A closed ticket is not the same thing as a completed
mission.
Imagine an airline reporting that your delayed flight was
successfully resolved because the airline eventually approved your boarding
pass—after the plane had departed.
The documentation would be immaculate. Your vacation would
still be in trouble.
Healthcare's stakes are obviously different. The underlying
lesson is the same: process completion should not be mistaken for real-world
success.
The metric healthcare should stop overlooking
Organizations commonly measure authorization turnaround
times, denial rates, approval rates, and appeal outcomes.
Those measures matter. But considered alone, they do not
capture the full operational journey.
Healthcare organizations should consider tracking:
Time to initial decision: How long did the first
response take?
Time to final resolution: How long did the request
remain unresolved after a denial or dispute?
Time to action: How long after approval did the
intended service actually occur?
Documentation-related rework: How often did requests
require corrections or additional information?
Escalation time: When clinically meaningful changes
occurred, how quickly did the relevant information reach the appropriate
decision-maker?
Feasibility after delay: Was the original plan still
appropriate and safely achievable when the decision arrived?
Reason for reversal: Did the appeal introduce new
information, correct an administrative problem, or reveal a substantive
disagreement?
These are proposed measures, not claims that all
organizations currently ignore them.
They also require careful interpretation. A poor outcome
after a delay does not automatically prove that the delay caused it. Clinical
review is necessary to assess the relationship between the administrative
process and the patient's outcome.
But organizations cannot learn systematically from events
they never connect.
The goal is not simply to shorten every clock. It is to
understand which clock is creating avoidable delay—and whether that delay
matters to the care being delivered.
Seven changes that could make the process more
intelligent
1. Make urgency visible at the beginning
A routine authorization request and one involving a rapidly
deteriorating patient may require different escalation pathways.
Use clinically appropriate urgency criteria, clear
ownership, and a reliable method for communicating material changes.
A request should not remain frozen simply because its
original submission is complete.
2. Make denial reasons useful
If documentation is missing, identify what is missing when
possible.
If the dispute concerns medical necessity, explain the
relevant criterion.
If a request went to the wrong organization, identify the
routing problem.
“Denied” is a status. It is not a complete action plan.
3. Give every pending case an owner
A pending request should have a responsible person, a
defined next step, and a follow-up date.
“Pending” is not a plan. It is a condition.
The difference is the same as the difference between saying
“the car is broken” and saying “the mechanic has the part and will install it
tomorrow.”
One describes the problem. The other describes the next
action.
4. Connect approval to delivery
Authorization is not the final objective. It is one step
toward a service being delivered.
Track what happens after approval and identify additional
barriers that prevent the intended action.
Otherwise, an organization can celebrate a decision without
knowing whether it accomplished anything.
5. Learn from reversals
When a denial is overturned, ask why.
Was relevant evidence overlooked? Was the request
incomplete? Did new information change the clinical picture? Was the policy
interpreted differently on appeal?
The objective is not automatic blame. It is pattern
recognition.
6. Separate preventable administrative failures from
genuine disagreements
Not every denial is avoidable. Not every appeal should
succeed.
A mature process distinguishes incomplete documentation,
routing failures, coverage disputes, and clinical disagreements rather than
reporting them all as a single number.
Different problems require different interventions.
7. Measure time in the context of clinical reality
A two-day delay does not mean the same thing in every
situation.
Organizations should assess whether the delay affected the
feasibility or timing of the intended care, while avoiding unsupported
assumptions about causation.
The objective is a more meaningful measure of
performance—not simply a faster stopwatch.
AI can write the appeal. Can it help prevent the problem?
Healthcare AI is often marketed as a way to automate what
happens after a denial.
It can help draft appeals, summarize records, classify
denial reasons, and organize supporting documentation.
Those functions may reduce administrative workload.
But there is a larger opportunity.
What if technology could help identify missing information
before submission? What if it could flag a mismatch between a request and known
requirements? What if it could identify a stalled case, surface the next
action, or alert an authorized human reviewer when a defined escalation
threshold is reached?
That is a different question from whether AI can write a
better appeal letter.
The most valuable administrative AI may be the kind that
makes the appeal unnecessary—not by suppressing legitimate disputes, but by
preventing avoidable process failures.
That distinction matters.
AI must not fabricate clinical facts, manufacture
documentation, guarantee authorization, or replace clinical judgment. It needs
appropriate validation, privacy safeguards, auditability, and human oversight.
And even a well-designed system cannot eliminate every delay
or guarantee a better clinical outcome.
The goal is more modest—and more useful: identify
preventable problems earlier, make the relevant information easier to find, and
help the right person act at the right time.
The OnnX thesis: Healthcare should fix the information
before the rework
This is the operational problem that motivates OnnX,
an AI-powered medical billing SaaS initiative focused on small and medium-sized
clinics.
The central thesis is that some revenue-cycle failures begin
before a claim is submitted or a denial appears.
Missing information, inconsistent records, disconnected
workflows, and late identification of requirements can create avoidable rework.
The opportunity is to identify those problems earlier and prompt appropriate
human intervention.
OnnX is being developed around that upstream-prevention
approach. It was not involved in Dr. Farooq's patient's care, and there is no
evidence that OnnX—or any particular technology—would have changed the outcome
described in the account.
The story illustrates a broader design question for
healthcare technology:
Are we building tools that help organizations repair
problems faster, or tools that also help them identify preventable problems
before they spread through the workflow?
Healthcare needs both.
A genuine coverage dispute may require an appeal. A missing
document may require a better intake process. A routing error may require
workflow redesign. A changing clinical situation may require timely escalation.
The challenge is to identify the actual problem instead of
treating every failure as a generic billing issue.
For small and medium-sized practices, the stakes are
immediate. Staff time is limited, and administrative rework competes with
patient coordination and essential operations.
Nobody opened a medical practice because they dreamed of
spending their afternoon searching three portals for a document uploaded
yesterday.
The joke is familiar. The wasted time is real.
OnnX's goal is to help clinics identify preventable
billing problems earlier—not merely process the aftermath faster.
That is a product thesis to validate, not a promise to
eliminate denials or solve every healthcare administrative failure.
The uncomfortable truth: Good people can operate a bad
process
It is tempting to turn a story like this into a simple
morality play: the physician is the hero, the insurer is the villain, and the
solution is to eliminate every administrative restriction.
But healthcare is more complicated.
Coverage decisions require rules. Clinical information can
be incomplete. Reviewers must interpret evidence. Providers must document
medical necessity. Resources are finite.
Those realities do not excuse avoidable delays or
inappropriate denials. They explain why reform must address both decision
quality and process design.
A system can contain competent people performing
individually reasonable tasks and still produce an unacceptable overall result.
Responsibility becomes fragmented. Information arrives too
late. Metrics reward narrow process completion. Nobody examines the full
journey from clinical need to actual care.
The failure may not sit with one individual.
It may sit between individuals, systems, and
responsibilities.
That is where better workflows, better measurement, and
carefully designed technology may have the greatest opportunity to help.
The question healthcare cannot afford to ignore
Dr. Sameen Farooq's account does not establish that a
different authorization process would have changed his patient's outcome. We
should not claim that it would.
It does raise a serious question about the relationship
between administrative decisions and the changing circumstances in which care
is delivered.
A denial was reversed. The patient still died in the
hospital, and the planned transfer was considered unsafe.
That distinction matters.
Healthcare should measure the quality of decisions. It
should also examine whether the processes behind those decisions are timely
enough to support meaningful action.
Because approval is not the finish line.
Approval is supposed to enable something.
If the intended action is no longer safely possible by
the time approval arrives, a green dashboard is a very poor substitute for
success.
And if we can identify avoidable failures upstream, why wait
for another denial to tell us where to look?
Join the discussion
Should healthcare organizations measure whether
authorization delays made intended care more difficult or no longer safely
feasible—not just whether requests were eventually approved?
And for healthcare AI builders: Are we creating systems that
prevent avoidable administrative failures, or are we teaching machines to clean
up after them?
I welcome thoughtful perspectives from physicians, practice
managers, medical billers, payer leaders, and healthcare technology builders.
- Visit
the personal website
- Listen
to the podcast on Spotify
- Subscribe
and watch on YouTube
About the author
Dr. Daniel Cham is a physician-entrepreneur and founder of OnnX,
an AI-powered medical billing SaaS initiative focused on helping small and
medium-sized clinics identify billing problems earlier and reduce avoidable
administrative rework.
Editorial note
This article discusses a physician's published account
and broader administrative-process questions. It does not independently
establish the cause of the patient's death, assign legal liability, or claim
that any particular technology would have changed the outcome.
References
Here is a clean reference list you can append to the
article. The first two sources document the physician stories; the remaining
sources support the broader discussion of prior authorization, administrative
delays, and potential reforms.
1. Farooq
S. “An insurance denial kept my patient from end-of-life care.” KevinMD.
October 16, 2026.
2. Saifi
C. “Prior authorization delays run on 3 clocks, not just 1.” KevinMD.
October 13, 2026.
3. U.S.
Department of Health and Human Services, Office of Inspector General.
“Medicare Advantage Organizations Overturned Nearly All Appealed Prior
Authorization Denials for Skilled Nursing Facility Admission, Raising Concerns
About Initial Denials.” 2026.
The report found that 95% of appealed denials for
skilled-nursing-facility admission in the reviewed Medicare Advantage cases
were overturned. This finding applies to that specific service and review, not
all prior authorization denials.
4. American
Medical Association. “AMA survey: Prior authorization reform pledge falls
short with physicians.” May 13, 2026.
The survey reports physicians' experiences with delays,
patient harm, administrative workload, and insurer reform commitments.
5. Centers
for Medicare & Medicaid Services. “Improving Prior Authorization
Processes.”
This explains applicable decision timeframes and
reporting requirements under the CMS prior authorization rule.
#Healthcare #MedicalBilling #PriorAuthorization
#RevenueCycleManagement #HealthcareAI #PatientSafety #HealthTech
#MedicalNecessity #HealthcareOperations #OnnX
No comments:
Post a Comment