We asked a practice two questions on the same Tuesday morning, in two different rooms.
In the first room, we asked the person running intake how many referrals were sitting unscheduled. She knew, roughly, and had the exact figure inside a few minutes. People had been waiting. Some of them a while.
In the second room, we asked the person running the schedule how many slots had gone unfilled the week before. She knew that too, and pulled it just as fast.
Neither of them had ever seen the other number. Not once. Both had worked there for years.
Why do patients wait weeks when we have open appointment slots?
Because the referral queue and the schedule belong to different people and get compared by nobody. Intake knows who is waiting. Operations knows what is open. Neither report contains the other’s information, so the overlap never surfaces.
The answer everybody gives first
Ask a practice why new patients wait three weeks and you get the same sentence. We are full.
Sometimes that is exactly right. Often it is an impression formed by looking at today’s schedule, which is genuinely packed, and assuming next month looks the same. Today is always full. Today was booked weeks ago.
What almost nobody does is check whether the weeks in between were full, because that requires putting two reports next to each other that live in two different jobs.
Three situations that look identical at the front desk
Long waits with full slots is a capacity problem. There is no room. The answer is more providers, more hours, or another location, and all three are expensive and hard to reverse.
Long waits with open slots is a process problem. The room existed and nobody moved anyone into it. The answer costs nothing and sits entirely inside your control.
Short waits with open slots is a demand problem, and if referrals are arriving but not converting, the loss is happening before scheduling ever sees them. That belongs to marketing and referral relationships, not to operations at all.
Same symptom in every case. A patient on the phone being told three weeks. Three completely different responses, and a practice that cannot tell them apart will eventually hire its way out of a problem hiring does not touch.
Why nothing surfaces it
Nothing is broken, which is what makes it hard to find.
Every system is working. Every person is doing the job they were given, competently. The failure lives in the space between two roles, and no report is built to look there, because reports are built around roles.
There is a reporting asymmetry underneath it too. Open slots are visible if you go looking for them. Waiting patients are visible if you go looking for them. Waiting patients and open slots in the same week is a comparison nobody has been assigned to make, so it does not exist anywhere as a number.
What the schedule is actually telling you
Most practices track one capacity figure: how much of the schedule got booked. That number looks forward and it counts intent.
There is a second figure underneath it, which is how many of those slots were actually seen. It is always lower, because a booked slot that no-shows was never used. McKinsey found health system leaders are routinely surprised by the realised number, because the forward-looking one does not account for late cancellations and no-shows and therefore overstates both scheduling effectiveness and clinic utilisation.
A practice booking 92% of its slots and seeing 78% has 14% of its clinical capacity claimed and then lost. That is a different problem from not having enough slots, and it responds to different work.
HFMA already publishes a key for the first half of this, percentage of patient schedule occupied, and it applies to physician organizations as well as hospitals. Most practices have never run it per provider, which is the only version that can be acted on, because groups are rarely evenly busy and an average that blends a full schedule with an empty one describes nobody.
What is an unfilled slot actually worth?
Run this on your own numbers rather than anyone else’s. It takes a calculator and about two minutes.
Start with slots available in a month, per provider. Subtract the slots that were actually seen, which is the realised figure, not the booked one. Multiply what is left by your average revenue per visit for that provider and visit type.
That figure is your monthly empty slot cost. Annualise it if you want the number that gets attention, and do it per provider, because the total tells you the size and the split tells you where.
A worked example, on assumptions you should replace
Assume a provider with 160 available slots a month. Assume 92% get booked and 78% get seen, which is the gap described above. Assume revenue per visit of 200 dollars, and change that to your own figure immediately, because it is the number that moves the answer most.
That is 125 slots seen against 160 available. Thirty-five unused. At 200 dollars a visit, 7,000 dollars a month for one provider, 84,000 across a year.
Every number in that paragraph is an assumption and none of them are a finding. The arithmetic is the point. Put your own figures in and the answer is yours.
Why the number is worse than it looks
An appointment slot is perishable. Thursday at two o’clock exists once, and at three o’clock its value is zero regardless of what anybody intended.
That is different from almost every other revenue problem in a practice. A stalled claim is late money and most of it eventually arrives. A patient balance can be worked for months. An empty slot does not age. It ends.
Which means the two halves of this problem carry very different price tags. Slots empty because the schedule genuinely had no demand is a marketing question. Slots empty while patients sat waiting to be scheduled is money that was available, in the building, on that day, and expired anyway.
Then set that against what the fix costs. Comparing two reports weekly and moving waiting patients into open gaps requires no new staff, no software, and no capital. If your annualised figure is meaningful and the fix is a standing half hour on somebody’s calendar, that is the whole business case and it does not need a spreadsheet.
How every other appointment business handles it
An airline does not look at the passenger list and the seat map in separate meetings. An empty seat on a departed flight is unrecoverable revenue, so the two numbers live on one screen and somebody is measured on closing the gap between them.
A hotel does the same with rooms and the booking queue. A repair shop does it with bays and the job list. In every one of them the perishable asset and the waiting demand are compared continuously, by a named person, because the asset expires at the end of the day whether or not anybody noticed.
A clinical hour is exactly that kind of asset. Practices are one of the few appointment businesses that routinely hold the two halves in separate hands.
Found, fixed, and held
The find is one page: referrals waiting and slots unfilled, side by side, weekly, per provider, for the same period.
The fix is somebody owning that page with authority to move patients into gaps. Not a new report. A named person and a standing time.
What holds it is the comparison running on a schedule rather than when somebody thinks of it. Left to memory, this gets checked during a bad month and forgotten during a good one, which is precisely backwards, because a good month is when the slack accumulates.
What this means for you
Run the comparison once for a single week you have already lived through. Referrals that were waiting on Monday, slots that went unfilled by Friday.
If one of those numbers is zero, you have a clean answer and a clear direction. If both are non-zero, you were turning patients away while paying for time nobody used, and it happened without a single person doing anything wrong.
Most practices have never seen those two figures next to each other. The first week is usually uncomfortable, because the overlap is rarely nothing.
Grab 30 minutes with us. Prep nothing. You will see which of the three situations you are actually in.
Questions people ask
How do I know if my practice has a capacity problem or a process problem?
Compare wait times against unfilled slots for the same week. Long waits with full slots is capacity. Long waits with open slots is process, and process costs nothing to fix. The two look identical from the front desk.
Should we hire more providers to reduce patient wait times?
Only after checking whether slots are going unfilled while patients wait. Hiring adds slots and does nothing about slots that already exist and are not being filled. Measure both figures for a month before committing to the cost.
How do I calculate the cost of unfilled appointment slots?
Take slots available in a month per provider, subtract the slots actually seen rather than the slots booked, and multiply what is left by your average revenue per visit. Do it per provider, because the total shows the size and the split shows where.
What is the difference between slot fill rate and slot utilisation?
Fill rate counts booked slots and looks forward. Utilisation counts slots actually seen and looks backward. The gap between them is no-show and late-cancellation cost, and the forward-looking figure overstates performance by leaving it out.