Sliding-Fee Conversations an FQHC Front Desk Can Have Confidently

Sliding-fee conversations are the ones a health center front desk dreads, because the patient is asking a pricing question and the honest answer depends on three facts nobody has in front of them.

The patient wants to know what the visit will cost before they agree to come in. The person answering the phone knows the health center has a discount schedule. What they do not know, in the four seconds they have, is which tier this household lands in, whether the paperwork behind that tier is still current, and what the patient already owes.

The question that decides whether the patient books

Cost is not a billing conversation at a health center. It is an access conversation, and it happens at the top of the call rather than at checkout.

A patient who cannot get a straight answer about what they will pay does one of two things. They book and then do not show, or they do not book at all. Either outcome is a slot that produced nothing, and neither one shows up in a report labeled as a pricing failure.

What makes it hard is that the answer is genuinely conditional. The discount depends on household size and income, the documentation supporting it expires, and the patient may be carrying a balance from a prior visit that changes what staff should say next. Those three facts live in different places, and the front desk is expected to combine them live, on the phone, while someone waits.

Staff cope by hedging. They tell the patient the health center has a sliding scale and someone will go over it at the visit. That is a true statement and a lost booking, because the patient called precisely to avoid the surprise.

What the discount schedule is actually required to do

The rule underneath all of this is more specific than most front desk scripts assume, and the specificity is useful.

Federal regulation requires a health center to prepare a schedule of fees designed to cover its reasonable costs, and a corresponding schedule of discounts adjusted on the basis of the patient’s ability to pay. That schedule has to provide a full discount to individuals and families with annual incomes at or below the federal poverty guidelines, and no discount to those with annual incomes greater than twice those guidelines. Nominal fees may still be collected from patients at or below the full-discount level.

Read operationally, that is a decision tree with hard edges. Below one threshold the answer is a nominal fee. Above another the answer is the full fee. Between them the answer is a tier. The boundaries are not a matter of judgment, which is exactly why the conversation can be scripted and automated up to the point where judgment starts.

The part that does require a person is everything upstream of the tier: verifying income, sizing the household, and accepting documentation. That is eligibility work, and it belongs to the staff who do it. The automation’s job is to know which tier is already on file and to say so plainly.

The complication: the tier is on file, the paperwork lapsed

This is the failure that produces angry calls, and it looks like nothing on the schedule.

Sliding-fee determinations are made per household and re-verified on a cycle. The determination sits in the chart. The documentation behind it has a date. When the date passes, the patient is still in the system at their old tier, and a front desk quoting from what they see will quote a discount the patient is no longer approved for.

The patient hears a number on the phone, arrives, and is told something different at the window. From their side that is a health center that changed its price. From the practice’s side it is a recertification that came due between two visits.

There is a related pattern worth naming, because health centers run into it the same way practices run into payer enrollment grids. The full eligibility picture usually lives partly outside the EHR, in a spreadsheet or a scanned folder that scheduling has no view into. Asking for the complete grid rarely works. The workable artifact is the narrower one: the current tier, its effective dates, and a flag for whether re-verification is due.

An automation handling this reads the sliding-fee plan attached to the patient through GET /slidingfeeplans before it says anything about cost, and it checks the effective dates rather than the tier alone. When the determination is current, it states the tier. When it is expired or missing, it does not guess and it does not quote the old number. It says the discount needs to be re-verified, offers the appointment, and routes the re-verification to the eligibility staff with the patient’s contact details attached.

That handoff is the whole design. The automation is allowed to report a determination. It is not allowed to make one.

Balances collide with the discount conversation

The second complication arrives inside the first. A patient asking what a visit costs frequently also owes something from a previous one.

Those are two different questions and staff often answer them as one, which is how a routine scheduling call turns into a collections call the patient did not expect. The right sequence is to answer the access question first, book the appointment, and treat the balance as a separate conversation with its own script.

The automation can read what is outstanding through GET /patients/{patientid}/claims/patientoutstandingdetailed, which matters because the useful version of this is itemized rather than a single total. A patient is far more likely to engage with a balance they recognize than with a lump sum they cannot place.

Where it stops is negotiation. Setting up a payment plan, adjusting a balance, applying a hardship exception, or deciding that a balance should not block scheduling are all decisions with money attached, and they go to a person. The automation surfaces the itemized balance, offers a payment link when the practice wants one offered, and routes anything that sounds like a request for relief to billing staff.

One rule prevents most of the damage here: an outstanding balance should never silently block a booking at a health center. If the practice has a policy about that, it is a policy a human applies, not a condition an automation enforces on a phone call.

Where the phone hours actually go

The staffing case for automating this is easier to make once you look at what the phones are already consuming.

In a March 10, 2026, MGMA Stat poll of practice leaders, the most time-intensive phone tasks were eligibility and prior authorization at 45%, scheduling at 31%, intake at 9%, prescription refills at 6%, and an other category at 9%. The poll had 294 applicable responses.

Eligibility is the largest single line, and at a health center it is larger still, because the sliding-fee determination sits on top of the payer question rather than replacing it. A patient can be uninsured and sliding-fee eligible, insured with a balance, or insured with coverage that changed since the last visit. Each of those needs a different first sentence.

Most of that time is not judgment. It is lookup, restatement, and routing. Reading a tier, reading a balance, and saying both accurately is repetitive work with a correct answer, which is the category that scales badly with headcount and well with automation. The staff time it frees goes back to the eligibility work that actually requires a person, which is the part health centers are usually behind on.

What to measure once it is running

The obvious metric is the wrong one. Calls handled says nothing about whether patients got a usable answer.

The number that tracks the actual problem is the share of cost questions answered with a current, documented tier rather than deferred to the visit. That is measurable, it moves quickly, and it maps directly to bookings that hold.

Two more are worth watching. The first is re-verification lead time: how long between the automation flagging an expired determination and eligibility staff closing it. If that number grows, the flag is being generated and ignored, and patients will keep hearing one price on the phone and another at the window.

The second is the rate of quoted-tier disputes at check-in. It should fall to near zero, and if it does not, the automation is reading a field that is not the one the front desk trusts. That is a configuration problem worth finding early rather than a reason to stop reading the field.

Key Takeaways

Treat the cost question as an access question. A patient who cannot get a straight answer about price is a booking you already lost, and it will not appear in any report as a pricing failure.
Read the sliding-fee determination and its effective dates before quoting anything. The tier alone is not enough, because an expired determination still displays.
Never let an automation make a sliding-fee determination. It reports what is on file and routes re-verification to eligibility staff with the patient’s details attached.
Answer the access question first and book the appointment before raising a balance. Collapsing the two turns a scheduling call into a collections call the patient did not agree to.
Surface balances itemized rather than as a single total. Patients engage with charges they recognize and disengage from lump sums they cannot place.
Send every request for a payment plan, adjustment, or hardship exception to a person. Anything with money attached is a human decision.
Do not let an outstanding balance silently block a booking. If the health center has a policy there, a person applies it.
Measure the share of cost questions answered with a current documented tier, plus re-verification lead time. Calls handled tells you nothing about whether the patient got an answer.

Health centers are usually described as complex because of the payer mix, but the harder complexity is that eligibility, discounting, and collections are three separate systems that a patient experiences as one question asked at the front desk. The answer exists. It is just spread across a determination record, a set of effective dates, and a claims balance that nobody can assemble in the four seconds a phone call allows. An AI team working inside athenaOne can read the sliding-fee plan and its dates before it speaks, state the tier when the documentation is current, decline to guess when it is not, present an itemized balance without turning the call into a negotiation, and hand every judgment call to the staff who are supposed to make it.

Sources