Allergy Injection Scheduling and the Front Desk

Allergy injection scheduling breaks the assumption every scheduling system is built on, which is that a patient books an appointment. An immunotherapy patient does not book an appointment. They enter a series that runs for years, arrives at a cadence set by a protocol, and takes up a chair for a while after the part that took two minutes.

The injection schedule and the provider schedule are two different businesses sharing one front desk.

Provider visits behave normally. They are booked ahead, they have a duration, and a no-show costs one slot. Injection visits behave like a subscription. The same patients return on a repeating cadence for a long time, often walking in during defined hours rather than booking, and the constraint is not provider availability but the capacity of the injection area and the staff covering it.

Most practices run the second business on paper, or on a whiteboard, or in the head of the nurse who has been there longest. It works until volume grows, someone is out, or the patient calls asking when they are due, and nobody can answer without walking to a binder.

The front desk absorbs all of it. Every call about when the next injection is due, every patient who fell off the cadence and reappeared months later, every question about whether they can come at four fifteen.

The visit has a tail, and the tail is the capacity constraint

Subcutaneous immunotherapy is administered as a series, and practices commonly ask patients to remain on site for a period after each injection so staff are present if anything happens. The injection takes a moment. The stay is the part that consumes the room.

That single fact determines the shape of the schedule. Capacity is not how many injections can be given per hour, it is how many people can be seated and observed at once. A clinic that books against injection time rather than chair time will overfill the room every afternoon and discover it only when patients are standing.

So the scheduling unit worth modeling is occupancy, not throughput. How many observation seats exist, how long each is held, and therefore how many arrivals per fifteen minutes the space can absorb. Practices that get this right often stagger arrivals rather than opening a free-for-all window, and the staggering is the scheduling product.

What none of this involves is a decision about the patient. The protocol, the interval, and the observation period are all set by clinicians in advance. The front office is placing arrivals inside constraints it was handed.

The template cannot say what the practice means

This is where athenaOne configuration decides how much manual work survives.

Practices build schedules out of generic slot types, and the EHR will hand back a generic slot when you search for a specific one. A generic fifteen-minute slot might be fine for an established injection patient and completely wrong for a first visit in a series. Which specific appointment types a generic slot is actually eligible for, per provider and per department, is the mapping problem underneath most scheduling automation.

Allergy practices add their own version. Injection visits usually need their own appointment types and often their own template, separate from provider visits, because the resource being consumed is different. Where that separation has not been built, injection patients get booked into provider slots and the provider schedule silently degrades.

The work is unglamorous and it is most of the value. Enumerate the appointment types that exist. Map which of them belong to the injection line and which to the provider line. Confirm which generic slots each one may occupy, per department. Then automation has something true to book against.

Skip that and you have automated booking into a template that was already lying to you.

Falling off the cadence is a front-office event

The most common failure in an immunotherapy program is not a scheduling collision. It is a patient who simply stops coming.

When MGMA asked practice leaders where they would focus on patient access in 2026, no-shows were the largest single answer at 27%, with online scheduling at 24% and phone access at 22% close behind. In a specialty where the patient relationship is a multi-year series, a lapse is different from a missed appointment. Nobody notices immediately, because there was no booked slot to sit empty.

That makes the recall list the real asset. Who is in an active series, when were they last seen, and what does the practice’s protocol say about the interval. Patients who have drifted past their expected window are a working list, and the outbound call that brings them back is the highest-value call the front office makes that day.

What the automation does with that list is book, not advise. It reaches the patient, tells them the practice is expecting them, and offers times that fit the arrival pattern. If the patient has been away long enough that the practice’s protocol requires a clinician to weigh in before the next injection, that is a rule the practice wrote, and the call routes to clinical staff instead of booking.

Answering the question patients actually ask

The single most common inbound call in an allergy practice is some version of when am I due, and it is embarrassing how often it cannot be answered quickly.

The information exists. Last visit, appointment type, and the interval the protocol specifies. What is usually missing is any surface that assembles those into an answer, so the call becomes a hold, a walk to a binder, and a callback.

An automated front door handles this well precisely because it is a lookup rather than a judgment. Identify the patient, read their injection visit history, apply the practice’s stated interval, and offer the next available arrival window. Book it, confirm it, and send a reminder that reflects the arrival time rather than the injection time, because arriving late in a staggered window is the same as not arriving.

The caller who asks anything beyond scheduling, about the injection itself, about a reaction, about whether they should still come this week, gets a person. That handoff should be fast and unapologetic. It is also rare enough that handling it well costs almost nothing.

What a well-run injection line looks like

Four properties, all of them administrative, all of them measurable.

The injection line has its own appointment types and its own template, so its volume never distorts provider availability. Arrivals are staggered against seat occupancy rather than injection duration, so the room does not overflow at four o’clock. Every patient in an active series appears on a list with a last-seen date and an expected window, so lapses surface within days rather than months. And the front desk can answer when am I due without leaving the phone.

None of that requires new clinical capability. It requires the practice’s existing protocol to be represented somewhere a system can read, and the schedule to be built around the resource that is actually scarce.

The payoff shows up in two places. The injection room stops having bad afternoons, and the series completion rate stops depending on whether a particular staff member remembered to call somebody.

Key Takeaways

Model capacity as observation seat occupancy, not injections per hour, and stagger arrivals against it.
Give the injection line its own appointment types and template so its volume does not consume provider slots.
Map which generic slots each injection appointment type may occupy, per provider and per department, before automating any booking.
Maintain an active-series list with last-seen dates so lapses surface in days rather than months.
Let automation answer when am I due as a lookup against visit history and the practice’s stated interval.
Route anything about the injection itself to clinical staff immediately, and keep that path short.

An allergy practice’s schedule is really two schedules, and only one of them is about providers. Build the injection line around seat time and a live recall list, and the front desk stops running it from memory.

Sources