Her cardiology referral went in on a Tuesday.
She told me about it weeks later at a family thing, the way people tell me things once they find out what I do for a living. Someone was supposed to call her. No one did. She waited, then called the number on the printout herself, sat on hold, and hung up at eleven minutes because her break was over.
She got in eventually. Different system. Four months later.
No one in that story delivered bad care. The referral was correct. The cardiologist was excellent. She just never got there.
That isn't just a service failure. That's a revenue event.
Somewhere in that organization a finance team closed the month and never saw her. There is no account for patients who couldn't reach you. No line called Demand That Didn't Convert. The money left and the books balanced anyway.
I have been telling versions of this story for years, and for years the answer to "what did that cost us" was a shrug and a survey score. That answer stopped being acceptable this year.
What does patient access actually cost a health system?
Innovaccer put a number on it this year. The Economics of Patient Access in 2026 drew on structured interviews and surveys with 110 hospital CFOs, COOs, and chief growth officers representing $84 billion in combined net patient revenue.
A typical 400-bed health system loses $6.2 million a year to avoidable referral leakage. That is 270 to 315 basis points of operating margin.
Now hold that against what a hospital actually earns. Kaufman Hall put the median hospital operating margin at 1.3% through December 2025, across more than 1,300 hospitals.
The low end of that leakage range is more than double the entire median margin.
Written as a percentage it sounds abstract. Written as dollars it's the difference between a system that invests next year and a system that cuts.
Where the money actually goes
The research traces most of the loss to five failure modes. Not one of them is clinical.
Calls abandoned before anyone picks up
Same-day availability that doesn't exist
Scheduling workflows split across systems that don't talk to each other
Insurance and prior authorization friction
Referral loops that close too slowly, or never close at all
Every one of those is a decision somebody made about staffing, software, or who owns the handoff. None of them require a physician to do anything differently. Which means none of them will ever surface in a quality review, and all of them show up in the cash.
The last one is the quietest and the most expensive. A referral loop that never closes means nobody in your organization ever finds out the patient didn't arrive. The order was placed. The task was completed. From the inside, the process looks like it worked.
An abandoned call is not an abandoned call
Most access teams report abandonment against a service target. Healthcare call centers average about 7%, high performers run 2 to 3%, and multi-site hospital systems average closer to 13%.
Here is the trouble with that number. It's a service level. There is no revenue inside it. You can hit your target, report green, and still have no idea what left the building.
So stop reporting an abandoned call as an abandoned call. Price it as a missed opportunity.
You already know why people call you. Every handled call in that queue got tagged with a reason: scheduling, billing, results, prescriptions, records. That distribution is sitting in your system right now.
Take the same distribution and apply it to the calls you abandoned. There is no reason to believe the people who hung up wanted something different from the people who got through. They just waited in the same queue and lost.
Then apply your average call-to-visit conversion rate to the scheduling share of those abandoned calls, and multiply by net revenue per completed visit. That's your missed scheduling revenue, from data you already own.
Run the same move on the billing share. Apply your payment rate on handled billing calls to the abandoned ones and you have the collections you didn't make, on top of the visits you didn't book.
Two things make this more urgent than it looks. More than 60% of patients in queue hang up after one minute and over 90% are gone by five. And 85% of the people who never get through will not call back. They don't complain. They don't appear in a survey. They reschedule their lives around a different organization.
Your abandonment rate is a service metric. Your abandoned calls are a revenue forecast.
One is a number you defend in an operations meeting. The other is a number you take to the people who set the budget.
The spread is the story
That 35-point gap costs bottom performers $2.8 million a year, and it is widening at roughly 8% annually. Systems at the same revenue tier, in the same kind of market, routinely sit thirty points apart.
Geography doesn't explain that. Payer mix doesn't either. The executives in that sample tied the spread to deliberate choices: what got funded, how the process was designed, and whether anyone was actually accountable for the handoff.
Speed does a surprising amount of the work. Systems that respond to a scheduling inquiry within five minutes convert two out of three patients, while systems that take 24 hours convert fewer than one in ten. Each lost interaction runs about $294 over the following twelve months.
$294 sounds like nothing. Multiply it by the calls your access team didn't get to last Tuesday.
And it compounds. Systems in the bottom quartile on referral leakage give up roughly $110 million in organizational value over five years against the top quartile. Not revenue in a single year. Value, permanently, because market share moved and didn't move back.
The research is blunt about the window: organizations that wait eighteen months may not close the gap no matter what they spend afterward. Patients who found a faster door don't come back to test whether yours got better.
The access lever almost nobody puts in the model
Access is not only whether the phone gets answered. It's whether the person on the other end can be understood. It's whether the patient walking into your building can find the right room without stopping three strangers to ask.
Language access gets classified as a compliance obligation, so it gets budgeted like one. A line item to satisfy, not an investment to size. That classification is costing organizations real money, and the evidence has been sitting there for years.
A natural experiment at an academic medical center put a dual-handset interpreter telephone at every bedside on a medicine floor. Thirty-day readmissions for patients with limited English proficiency fell from 17.8% to 13.4%. Over the same window, readmissions for English-proficient patients on that floor went the other direction, from 16.7% to 19.7%. The study counted 119 readmissions averted and, after paying for the interpreter service itself, estimated monthly savings of about $161,000.
The detail that matters most comes after the intervention. When the phones became less accessible again, the improvement disappeared.
Convenience was the intervention. Not the availability of interpretation. The proximity of it.
Which is why onsite AI translation is worth more attention than it gets. Not the language line somebody has to request and wait for. A device in the patient's hand or mounted on the wall that helps a person move through your building with confidence, at the moment they're standing in the wrong hallway.
On cost alone I'd expect that to price out well against a per-minute language line. The return nobody models is the second one: a patient who didn't feel lost in your lobby is a patient who comes back. Language access reads as an equity program. It behaves like a retention program.
That second half is my hypothesis, and I'll name it as one. The 2026 review of language access evidence in the Journal of Patient Experience makes the case that professional interpretation and language-concordant care shape patient experience directly. What almost nobody has done is put a retention number on it. It's also the kind of hypothesis you can test on one clinic in a quarter, which is more than can be said for most things in a strategic plan.
Trust isn't a survey score. It's a conversion rate.
Patient experience has spent twenty years being measured after the fact, by people who already made it through the door. The survey goes to the patient who got seen. It never reaches the woman who hung up at eleven minutes.
Your satisfaction scores are a survey of the people you didn't lose.
This is why the operational layer matters more than the encounter. The clinical care is table stakes. Getting in the door, the handoffs between steps, the follow-through afterward: that's where trust is built or spent, and it's the part almost nobody owns end to end.
There is a second signal underneath this one. A referral is a handoff, and every handoff is a place a patient can feel forgotten. Accessibility gets them to the door. Continuity is what carries them from your primary care physician to your specialist without the patient having to hold the thread themselves. When the loop doesn't close, you haven't only lost a booking. You've told that patient, without meaning to, that the parts of your organization don't know each other.
Referral leakage, defined plainly: demand that your organization generated or received, and then failed to convert into a completed visit. It's the patient your own physician sent to your own specialist, who ended up somewhere else. You paid to create that demand. Someone else billed it.
The old model treated access as an administrative cost to minimize. The evidence now treats access as a conversion funnel to optimize, where every abandoned call and unclosed referral loop is a measurable financial event.
What this means for you
If you sit in finance, this is a margin protection story before it's an experience story. Two to three hundred basis points is not a program you fund out of goodwill. It's a recovery you book.
If you run operations or patient access, this is the argument you've been missing in budget season. You've known the phones are underwater. What you haven't had is the conversion math that turns that into a number a finance committee has to answer.
Try this before your next budget conversation
Pick one high-volume service line. Not the system. One line.
Pull the call-reason distribution for every call your team handled on that line last month. Scheduling, billing, results, prescriptions, records.
Apply that same distribution to the calls you abandoned. Same queue, same demand, different luck.
Apply your call-to-visit conversion rate to the scheduling share of those abandoned calls.
Multiply by net revenue per completed visit. Then do the billing share against your payment rate on handled billing calls.
Annualize it. That is your missed revenue opportunity on one service line, built entirely from numbers you already have.
Two cautions from running this with teams. Pull from the source systems rather than the dashboard, because the dashboard was built to report the steps somebody already decided to watch. And include after-hours, where abandonment is effectively total and the volume is larger than most leaders expect.
Run it on one line and you'll have a real figure by Friday. Run it across all of them and you'll find your version of the $6.2 million. In my experience it's almost never the line the team would have guessed.
Then bring it to the person who controls the budget and say the sentence out loud: this is what we didn't collect on one service line last year, and here is the step where it left.
If you want the faster version
The Revenue Leaks Calculator walks the same math and shows you where to look. Free, a few minutes, no call required. It tells you where the leak is. What it means and what to do about it is a longer conversation. And if you don't have or know where to start to pull your numbers, let's talk. Book a Call.
Your income statement has no line for the patients who couldn't reach you. Somebody else's revenue does.
Which service line are you running first?
Let’s get to work,
Ebony
P.S. $6.2 million a year on a 400-bed system. 270 to 315 basis points, against a median margin of 1.3%. That money isn't lost to bad medicine. It's lost on hold. I walked the full financial case in this week's episode, The $6.2 Million Hospitals Lose to Broken Trust.
Ebony Langston is the founder of The Patient Experience Strategist™ and a fractional Chief Experience Officer for healthcare organizations rebuilding patient trust as a margin strategy. With more than 20 years of operations experience supporting Fortune 100 healthcare organizations, she now works with leaders across the field, from health system C-suites and patient experience teams to independent and small-group practice owners. She writes weekly for the executives and clinicians turning patient experience from a cost line into a revenue engine.
Connect on LinkedIn | Subscribe to The Patient Experience Strategist