What Happens While the Hospital Is “Waiting on Insurance”?

There is another phrase you hear constantly in hospitals:

“We’re waiting on insurance.”

Patients hear it.

Families hear it.

Nurses hear it during rounds.

Case managers say it when explaining why the discharge still has not happened.

And from the outside, I think it can sound remarkably passive.

Like somebody submitted a form, everyone is waiting for an email, and nothing much is happening until the insurance company responds.

That is not really what it looks like inside a hospital.

Usually, by the time insurance authorization becomes the thing everyone is talking about, something important has already happened:

The patient is getting better.

The acute part of the hospitalization may largely be over.

The IV antibiotic may have been stopped or converted to something that can be given outside the hospital.

The medication regimen may have settled.

The physician may no longer be changing the treatment plan every few hours.

Discharge may now be the main conversation happening around the patient.

And yet the patient is still there.

The nurses are still taking care of them.

The bed is still occupied.

The case manager is still working.

The family is still waiting.

And somewhere else in the hospital, another patient may be waiting for that same bed.

All because one part of the transition has not happened yet.

I think it helps to start with what the phrase usually does not mean.

It usually does not mean:

“We have no idea whether this patient is medically ready to leave.”

Those are different questions.

If somebody still needs hospital-level treatment, they are not simply sitting around because the insurance company has not authorized rehabilitation.

They are still being treated.

The situations I am talking about are the ones where the conversation has shifted from:

What do we need to do medically to get this person better?

to:

Where can this person safely receive the next level of care, and will that care be covered?

Maybe the team is recommending a skilled nursing facility (SNF).

Maybe an inpatient rehabilitation facility (IRF), a different setting with different coverage requirements.

Maybe another post-acute service that requires insurance review.

Whatever the setting, somebody eventually has to determine whether the requested service meets the patient's benefit and coverage requirements.

And this is where the insurer's side of the story matters.

UnitedHealthcare's published skilled-nursing-facility process, for example, shows that authorization is not simply someone checking whether the patient has an insurance card. For certain plans, nurses and medical directors review clinical information before making a coverage determination. Its guidance also lists orders, therapy evaluations, progress notes, treatment length and discharge planning among the records used for ongoing clinical review after SNF admission. That ongoing-review checklist should not be mistaken for a universal pre-admission requirement. [1]

That is one insurer's process, not a universal description of every plan.

But it illustrates what is actually being decided.

The payer is asking some version of:

Does this patient qualify for this level of care under this coverage?

That is a legitimate question.

The problem is what happens while healthcare waits for the answer.

This is one of the strangest parts to watch as a nurse.

Whether a patient is medically ready for discharge or not, nursing responsibility does not disappear.

I still assess the patient.

I still give medications.

I still answer the call light.

I still respond if something changes.

I still document.

They are still a patient in a hospital bed, and they still receive hospital nursing care.

But the type of care may feel completely different from earlier in the admission.

Think about someone who initially needed IV antibiotics, frequent monitoring and repeated adjustments to the treatment plan.

Earlier in the hospitalization, there may have been constant discussion:

Is the medication working?

Do we need to change the dose?

What did the new lab show?

Do we need another test?

Are they getting better?

By the time the person is medically ready to move on, much of that may be settled.

Now I may be handing them the same oral medications they are going to take at the rehabilitation facility or eventually at home.

The acute treatment has ended.

But because the next step has not been authorized or arranged yet, the hospital keeps providing the infrastructure around them.

There is something strange about that.

We have essentially said:

You no longer need this hospital to get better.

But the rest of the healthcare system has not yet created a path for you to leave it.

What surprised me when I started researching this was that there is evidence behind what hospital staff see every day.

One study looked at stroke patients who had already been medically cleared to leave the hospital for either inpatient rehabilitation or a skilled nursing facility.

Among 289 patients, those who required insurance precertification waited an average of 1.5 days after medical clearance, compared with 0.8 days among patients who did not require precertification.

That is an observed difference of about seven-tenths of a day in this particular hospital and patient population. The study was retrospective, not a randomized test of authorization. [2]

That is not a national estimate.

It was one stroke center, and it was published years ago.

But another multicenter study found something similar from a different angle.

Researchers examined 1,543 adults with traumatic brain injuries across four Level I trauma centers. They defined a discharge delay as remaining hospitalized at least 24 hours after case-management documentation said the patient was ready to leave.

Fourteen percent experienced a delay.

Among those delayed patients, the median delay was three days.

The most frequently documented cause was insurance authorization, accounting for 52% of delays, followed closely by the lack of an accepting bed at 41%. [3]

That second number matters just as much as the first.

Because it confirms something that becomes obvious once you work around discharge:

Insurance is not the only thing patients wait for.

You can fix authorization and still have a bed problem.

You can have a bed and still have a transportation problem.

You can have authorization, a bed and a transport plan, and still have some other operational issue prevent the patient from leaving.

Healthcare discharge is a chain.

Insurance can be one very important link without being the entire chain.

This is where the problem expands beyond the person waiting.

Hospital beds are not abstract things.

A bed upstairs usually represents the ability to admit somebody else downstairs.

And in my nursing career, I cannot remember the emergency department being some permanently empty place where every admitted patient immediately gets a room upstairs because capacity is always available.

Especially during busy periods, it is not.

People wait.

Sometimes they wait a long time.

If every hospital bed is occupied because every person upstairs still genuinely needs acute hospital care, that is an overloaded hospital caring for a lot of sick people.

That problem is difficult, but understandable.

What feels different is knowing that one of those beds belongs to someone whose acute medical reason for being there has effectively ended.

The person is not staying because the hospital still has a major medical intervention left to accomplish.

They are staying because the next part of the healthcare system has not been able to receive them.

And there is now national evidence suggesting that this is large enough to affect hospital capacity.

A 2025 JAMA Internal Medicine study examined 89.3 million hospital admissions involving people in Medicare Advantage and Traditional Medicare.

From 2017 through the third quarter of 2023, average length of stay increased from 6.0 to 7.1 days for Medicare Advantage admissions, compared with an increase from 5.8 to 6.3 days in Traditional Medicare.

The difference was particularly pronounced among patients ultimately discharged to skilled nursing facilities.

Using the study’s year-specific Table 2, the adjusted relative increase translates to about 1.9 million additional bed-days in 2022—roughly 286,000 average-length admissions. That is a capacity equivalent, not a count of additional patients hospitals could necessarily have admitted. [4]

That does not prove prior authorization caused those additional days.

The researchers could not identify when patients became medically ready or isolate insurer behavior as the cause. Network limitations, post-acute capacity, patient differences and pandemic effects are alternative explanations. [4]

But it tells us that the difference is not merely an annoyance experienced by a few case managers.

Something is happening at a scale large enough to consume meaningful hospital capacity.

Even that makes “waiting on insurance” sound simpler than it really is.

Because another misconception is:

Insurance approves the request, and the patient leaves.

Sometimes the pieces line up quickly.

Sometimes they do not.

Suppose a patient is going to a skilled nursing facility.

Insurance authorizing skilled care does not magically create a bed.

The facility still has to accept the patient.

The facility has to be able to meet the patient's needs.

A bed has to exist.

Transportation may need to be arranged.

The timing has to work.

Different organizations have different operating hours and processes.

A late-day authorization may leave very little time to coordinate transportation and admission.

Weekend and holiday processes may differ from weekday processes.

In fact, UnitedHealthcare's own SNF guidance explicitly contains separate notification rules for weekday versus weekend and holiday admissions—an important reminder that weekend admissions do occur, but that the administrative workflow itself changes depending on timing. [1]

That is why I would no longer describe this as:

Insurance finally says yes and the patient goes.

It is more like:

One organization finally says yes, allowing the other organizations to keep moving.

Healthcare fragmentation can sound like an academic phrase.

This is what I think it looks like in real life.

One organization says:

The patient no longer needs us.

Another says:

We need to determine whether we cover the next level of care.

Another says:

We can accept the patient if we have the right bed and can meet the patient's needs.

Another has to answer:

How are we physically getting the patient there?

Each decision can make sense independently.

Put them together, though, and the patient waits.

And while the patient waits, the hospital keeps supplying the infrastructure around them.

Nursing care continues.

Medications continue.

Meals continue.

Housekeeping continues.

The room stays occupied.

The case manager continues coordinating.

The family continues asking for updates.

And the person in the emergency department continues waiting for somewhere to go.

The phrase “waiting on insurance” compresses all of that into four words.

I also think saying we are “waiting” understates what hospital staff are actually doing.

Somebody may be tracking the authorization.

Someone may be checking whether more documentation is required.

Someone may be communicating with the receiving facility.

Someone may be updating the family.

Someone may be arranging transportation.

Someone may be developing a backup plan if the first facility falls through.

And that is happening alongside every other patient that case manager is responsible for.

This is something I am especially interested in now as I move from bedside nursing into case management myself.

Nurses experience healthcare delays mostly from the patient-facing side.

We know the person is still there.

We know the family is frustrated.

We know another admission is coming.

Now I want to understand the machinery producing that wait.

How much time is spent actually making a complex clinical decision?

And how much time is spent moving information between organizations?

Those are very different forms of work.

It would be easy to turn this into an article arguing that insurance authorization is simply pointless bureaucracy.

I do not think that is fair.

An insurance company is paying for care.

It has a legitimate interest in determining whether the requested service is covered and whether that level of care is medically appropriate.

UnitedHealthcare's utilization-management documentation describes reviews involving benefit eligibility, medical necessity, appropriateness, level of care and clinical criteria, with some potentially adverse determinations escalated for physician or medical-director review. [5]

We probably would not want a healthcare system where one organization could order unlimited expensive services and another organization was automatically required to pay without review.

So utilization management itself is not an irrational concept.

The real question is:

How much friction is necessary to perform that review?

Because there is a cost to unnecessary care.

But there is also a cost to making appropriate care wait.

Federal policy is beginning to recognize that.

As of January 1, 2026, Medicare Advantage organizations generally must make decisions on non-drug prior-authorization requests within 72 hours for expedited requests and seven calendar days for standard requests, while still acting faster when the patient's medical condition requires it. Similar timeframes apply to specified Medicaid and CHIP payers under the rule, with payer-specific details and lawful extensions. [6]

Seven days is an ordinary standard deadline, not a statement that seven days is clinically or operationally appropriate for every request. Under applicable program rules, extensions of up to 14 calendar days can be permitted in specified circumstances. [6]

CMS also requires affected payers to publicly report prior-authorization metrics such as approval and denial percentages and average and median decision time. The first 2025 metrics were due by March 31, 2026. Those aggregate reports still do not tell us how long medically ready hospital patients wait for a particular post-acute placement. [6]

That data should make future versions of this article better.

Because the important question is not merely:

What is the legal deadline?

It is:

How long are medically ready patients actually waiting?

This is where my original instinct turned out to be too simple.

My first thought was:

If an insurance company could approve a patient four days sooner, surely the insurer would save four days of hospital costs.

That sounds logical.

Healthcare payment does not always work that way.

Traditional Medicare—the fee-for-service program, also called Original Medicare—generally pays hospitals covered by the acute inpatient payment system prospectively. Under the Inpatient Prospective Payment System, the hospital generally receives a predetermined payment for the case or discharge, largely based on the patient's diagnosis-related group rather than receiving a new full payment for every additional day the patient occupies the bed. [7]

That changes the economics.

If the hospital receives roughly the same case payment whether the patient stays five days or six, the sixth day still consumes nursing time, medications, meals, housekeeping and physical capacity.

But it may not produce another full day of reimbursement.

In that situation, the hospital can absorb added variable costs and the cost of tying up capacity. That is not the same as saying it spends another average hospital-day cost in cash: some staffing and building costs remain even if one patient leaves. A newly available bed also only creates an opportunity for another admission if staffing and demand permit it.

Transfer-payment adjustments and unusually costly cases can change the payment calculation, so even prospective payment is not an absolute “same check regardless of stay” rule. [7]

Medicare Advantage gets even more complicated.

CMS’s 2021 guidance on negotiated-rate reporting illustrates that hospitals and Medicare Advantage insurers can negotiate payment in different ways. Some arrangements may resemble case-based Medicare payments. Others can involve per-diem rates, percentages of Medicare rates, or more complicated combinations. Under a per-diem arrangement, another covered, payable hospital day can increase payment; under another structure, it may not. Coverage and contract terms still matter. [8]

So there is no intellectually honest way to say:

“Insurance always pays for every unnecessary extra day.”

Sometimes it may.

Sometimes the hospital may bear much more of the marginal financial burden.

And suddenly the incentive problem becomes much more interesting.

That question is probably more important than anything I originally planned to write in this article.

Imagine a hospital has strong financial pressure to discharge a medically ready patient because another day consumes resources without generating much additional reimbursement.

At the same time, another organization controls one of the approvals required for the patient to move.

The insurer has good reasons to review the request.

But depending on its contract with the hospital, the insurer may not bear the same financial consequence from an extra day that the hospital does.

That does not mean the insurance company wants the patient to remain hospitalized.

It means the incentives may not be perfectly aligned.

The hospital may experience the delay as:

I need this decision because I need this bed.

The payer may experience the same case as:

I need enough information to make the correct coverage determination.

Both perspectives can be rational.

The patient is physically sitting in the middle.

That is healthcare finance.

Not simply:

Who is the bad guy?

But:

Who makes the decision, who absorbs the cost, and do those incentives line up?

This is where I think there may actually be room for improvement.

Right now, a tremendous amount of coordination can fall on the hospital.

The case manager needs to know what the payer requires.

They need to identify appropriate facilities.

They submit information.

They follow the status.

They respond if something is missing.

Then they have to combine the payer's decision with facility availability, transportation and everything else involved in making discharge happen.

But the insurance company already possesses information that the hospital is trying to obtain.

It knows its own network.

It knows its own rules.

It knows what documentation is required.

It knows where the request is inside its own process.

So I found myself wondering:

Why should another healthcare worker have to repeatedly extract that information from the payer?

Could the payer identify documentation requirements immediately?

Could authorization status be visible in real time?

Could the system automatically tell the hospital exactly what is missing?

Could straightforward requests move automatically while human clinical reviewers concentrate on the complicated cases?

When I said that aloud, it felt a little idealistic.

Then I looked at what CMS is implementing.

CMS’s rule requires electronic Prior Authorization APIs—interfaces for systems to exchange requests and responses—with compliance dates generally beginning January 1, 2027. The covered groups include Medicare Advantage, specified Medicaid and CHIP payers, and qualified plans on federally facilitated Exchanges; exact dates vary by payer type. The rule’s authorization requirements exclude drugs. [9]

These systems are supposed to allow providers to identify which services require authorization, determine the documentation requirements, submit the request and receive a response electronically.

The payer's response must be able to indicate approval, provide a specific reason for denial, or request additional information. [9]

In other words, while thinking through this article, my instinct was:

Why can't the hospital and insurance company simply exchange this information automatically instead of making humans chase it?

And federal policy is moving in almost exactly that direction.

CMS explicitly describes the goal as improving electronic exchange and making the authorization process more efficient from end to end. [10]

That does not eliminate the need for clinical judgment.

Some requests will still need nurses, physicians and medical directors to actually review the case.

CMS itself acknowledges that automation will not make every authorization a real-time decision because complicated cases still require human clinical review. [6]

That seems reasonable.

The goal should not be:

Remove humans from medical decisions.

It should be:

Stop making humans perform information-moving work that computers can do.

This is also not simply a theoretical technology project.

When CMS finalized its interoperability and prior-authorization reforms, it estimated that the package of policies would generate approximately $15 billion in savings over ten years by reducing administrative burden and improving information exchange among patients, providers and payers. [10]

That number needs an important caveat.

It is not an estimate saying hospitals will save $15 billion in unnecessary bed-days.

It is a broad federal estimate of savings from multiple interoperability and prior-authorization reforms.

Still, it tells us something.

Administrative friction in healthcare is large enough that the federal government believes making information move more efficiently can create savings measured in billions of dollars.

That gets back to the question I was asking from the bedside:

How much expensive healthcare labor are we using simply to make organizations talk to one another?

After watching discharge from the bedside, I think this is the larger point.

When a family hears:

“We're waiting on insurance,”

they are hearing about one unresolved step.

But the actual process includes a chain of dependencies, often worked on in parallel:

The patient becomes medically ready.

The next level of care is recommended.

Clinical information is assembled.

Coverage is reviewed.

A decision is made.

A facility accepts.

A bed becomes available.

Transportation is arranged.

The timing works.

The patient finally leaves.

That is not a mandatory sequence: referrals and facility acceptance may precede authorization, and planning should start before medical readiness.

Any one of those links can stop the process.

The traumatic-brain-injury study captures this remarkably well: insurance authorization was the most common documented reason for delay, but an unavailable accepting bed was almost as common. [3]

So fixing prior authorization will not fix every delayed discharge.

But that is not an excuse to leave preventable administrative delay alone.

If one part of the chain can become faster, clearer and more automated without sacrificing necessary clinical review, then it probably should.

Because while those organizations coordinate, the delay accumulates somewhere physical.

It accumulates in a hospital room.

The nurses keep caring for the patient.

The case manager keeps coordinating.

The family keeps asking when they are leaving.

And somewhere else in the hospital, somebody may be waiting for that room.

So maybe the phrase needs a little more translation.

“Waiting on insurance” does not mean nothing is happening.

It often means the medical part of the hospitalization is approaching its end while the financial, administrative and operational parts of the healthcare system are still trying to catch up.

And the most important question may not be who deserves the blame.

It may be:

Who controls each part of the wait, who pays for it, and how much of that waiting still needs to exist?