Transportation Calls Are Half the Dialysis Front Desk

Ask anyone who has worked a dialysis front desk what the phone is actually for and the answer is rides. Not clinical questions, not billing, not new patients. The dialysis front desk spends its day on transportation, because a treatment three times a week is a transportation problem twelve times a month, and every one of those trips has a way to go wrong.

This volume is invisible in most planning because it does not look like healthcare work. The ride is late. The ride went to the old address. The patient switched to the early shift and nobody told the transport provider. The standing order lapsed and now there is no ride at all on Wednesday.

Each call is small. Together they are the majority of the inbound, and they land on the same people who are supposed to be handling everything else.

What makes it worse is the coupling. A ride problem is not a customer service problem, it is a missed treatment, and a missed treatment is a clinical event the staff will spend the rest of the week chasing.

So the work is worth automating not because the calls are annoying but because the failure they cause is expensive, and because almost none of it requires a person.

The trip is a covered benefit with its own paperwork

Transportation to and from providers is a required Medicaid benefit, known as non-emergency medical transportation. MACPAC describes it as a benefit states must provide, delivered through a range of models including brokers and transportation network companies, with rules that vary considerably by state.

That matters to the front desk in a practical way. Rides are usually arranged through a broker or plan rather than by the clinic, so the clinic is a party to the arrangement without controlling it. When something breaks, the patient calls the clinic anyway, because the clinic is the number they know.

The population is also large and highly dependent on this. According to the United States Renal Data System 2023 annual data report, more than 808,000 people in the United States are living with end-stage kidney disease, with 68% on dialysis and 32% with a kidney transplant. The people on dialysis are making the trip several times a week, indefinitely.

For a practice administrator this frames the job correctly. The clinic is not running a transportation service. It is running the communication layer that keeps a transportation service and a treatment schedule pointed at each other, and that layer is mostly confirmations, changes, and chasing.

A schedule change is a transportation change

The single highest-yield rule in this segment is that nothing about a patient’s treatment time changes in one place.

Rides for recurring treatment are usually standing orders, arranged once and repeating. That is efficient until the treatment schedule moves. A patient who shifts from the afternoon chair to the early one now has a standing ride that arrives at the wrong time, and the mismatch shows up as a missed treatment rather than as a scheduling error.

The fix is to treat the ride as a dependent of the appointment rather than as a separate arrangement someone remembers. When the recurring series changes in athenaOne, a transportation update becomes a required task with an owner and a deadline, not a note in a handoff.

The same coupling explains a failure that confuses new administrators, which is the reminder that covers only half of the day. Reminder tools generally fire on one appointment, so a patient with a linked pair gets told about one and misses the other. In a clinic where the ride and the treatment are effectively two halves of the same commitment, a confirmation that mentions only the treatment leaves the part that actually fails unconfirmed.

Confirmations here should name both. The time of the treatment, the pickup arrangement as the clinic understands it, and a way to say something is wrong that does not require calling during business hours.

The three calls that eat the day, and what replaces them

Almost all of the transportation volume is one of three patterns, and each one has an automated form that works.

The first is the confirmation. Yesterday’s outbound to confirm tomorrow’s treatment and the ride attached to it. This is pure logistics, it happens on a schedule, and it can be delivered in the patient’s language through whichever channel the record says they use, including a secure portal message for patients who prefer it. What it produces is not a courtesy, it is early warning. A patient who says on Tuesday that their ride is cancelled gives the clinic a day to fix it.

The second is the change. A patient calls to move a treatment, a shift is reassigned, or a ride provider changes. Handled well, one contact updates the series, flags the transportation dependency, and confirms both back to the patient. Handled badly, it updates the calendar and leaves the ride pointing at the old time.

The third is the failure in progress. The ride has not arrived, the patient is stranded, or the driver went to the wrong address. This one is genuinely urgent and it is the one that most deserves a human, so the front-office layer’s job is to identify it fast, capture the details, and put a named person on it rather than leaving the patient in a queue.

That split is the whole design. Two of the three patterns leave the phone entirely, which is what makes the third one answerable quickly.

What has to stay with the clinical team

Transportation work sits close to a clinical boundary and the line has to be drawn deliberately.

The automation confirms treatment times and ride arrangements, records what the patient says about their transportation, updates the recurring series when a shift changes, chases a lapsed authorization for rides, and documents every contact against the chart. All of that is logistics.

It does not assess anyone. A patient who says they are unwell, who describes symptoms, or who is calling about how they feel rather than about a ride is not handled by the front-office layer. That contact goes to clinical staff immediately, with what the patient said captured as they said it.

The same applies to consequence. Whether a missed treatment needs urgent attention is not a scheduling determination, and nothing in the outreach process should behave as though it is. The automation notices that a treatment was missed and tells the people whose job it is to decide what that means.

That boundary also protects the useful part. A clinic that can say precisely which patients confirmed, which flagged a ride problem, and which went silent has given its nurses better information than a stack of message slips, without anyone in the front office making a judgment they are not licensed to make.

The measures that show the phone got quieter

The right metrics for this work are boring, which is a good sign.

Start with confirmed treatments as a share of scheduled treatments, measured the day before rather than on the day. That number is the early warning system working or not working.

Then count ride problems reported in advance against ride problems discovered at the chair. Moving volume from the second column to the first is the entire point, and it is the number that predicts missed treatments.

Track how many transportation changes followed a schedule change within the same day. A gap there means the dependency is still being carried by memory.

Finally look at inbound call volume by reason. If confirmations and routine changes have moved off the phone, the remaining calls should be shorter, fewer, and disproportionately the urgent kind, which is exactly the queue you want a person answering.

Key Takeaways

Treat the ride as a dependent of the appointment, so a shift change in athenaOne creates a transportation task with an owner and a deadline.
Confirm the treatment and the pickup arrangement in the same message, since reminder tools fire on one appointment and leave the other unconfirmed.
Run day-before confirmations as early warning rather than courtesy, because a ride problem found on Tuesday is fixable and one found at the chair is not.
Deliver confirmations in the patient’s language on the channel their record says they use, including the portal for patients who prefer it.
Keep a human on the ride that failed in progress, and let automation identify and escalate it in seconds rather than answer it.
Send anything about how a patient feels to clinical staff immediately, and keep the front-office layer out of it.
Measure ride problems reported in advance against ride problems discovered at the chair, because that ratio predicts missed treatments.

The dialysis front desk is not busy because patients ask a lot of questions. It is busy because a treatment that repeats three times a week generates a transportation event every single time, and nothing in the schedule knows about it. Connect the two, automate the confirmations and the changes, and the phone goes back to being for the problems that need a person.

Sources