Skip to main content

Practice Operations

How to Reduce Patient No-Shows in athenahealth

Reminders alone have stopped moving no-show rates. Here is where no-shows hide inside athenaOne, and the work that recovers the slot before the day is lost.

7 min read

If you want to reduce patient no-shows in an athenaOne practice, you almost certainly already send reminders. So does nearly everyone else, and the no-show rate has not cooperated. An August 2026 MGMA Stat poll found 32% of medical groups reporting higher no-show rates year to date than in 2025, with 58% about the same and only 10% lower.

The practices that got worse were using the same tools as the ones that improved: text and phone reminders, confirmation calls, fees. Another reminder is not the missing piece. What usually is missing is the work that happens after the reminder goes out: reading the answer, acting on the appointments that never confirmed, and putting a freed slot back in front of a patient who wants it. In athenaOne all of that has a place to live. It just needs someone, or something, working it every day.

Start with where your no-shows actually concentrate

A blended practice-wide no-show rate hides the problem. MGMA’s own advice in the 2026 poll write-up is to stop chasing a national figure and find where missed visits cluster: new-patient visits, procedures, behavioral health, one location, or one provider template.

In athenaOne that cut is already available. Appointment status, appointment type, department and provider sit on every booked visit, so a report of no-shows by appointment type and department for the last 90 days usually shows two or three hot spots carrying most of the loss. New-patient visits with long lead times are a common one. MGMA’s write-up names long scheduling lead times among the reasons another reminder may not help.

That matters because the fix is different in each hot spot. A new-patient slot booked five weeks out needs a confirmation loop. A procedure block needs the prep instructions confirmed. A single provider template with a high rate may just be offering slots patients cannot use.

Reminders work, but they are only half the loop

The evidence on reminders is consistent. A meta-analysis of 21 controlled studies found patients who received electronic notifications were 25% less likely to no-show than those who received none, 15% against 21%, and multiple notifications beat a single one.

Reminders also work by producing an answer. Patients who are reminded cancel and reschedule more, and a cancellation two days out is a good outcome. It is a slot you can still fill.

In athenaOne, reminders are managed through the appointment reminders API, and each appointment carries a confirmation status the practice configures. The reminder is the question. The confirmation status is the answer. Most practices measure the first and never work the second.

Work the unconfirmed list, not the whole schedule

The appointments still sitting unconfirmed 48 hours out are where tomorrow’s no-shows are. They are a short list, and they are the list worth a second contact.

Here is how that work runs when Pretty Good AI handles it inside athenaOne. Each day it reads confirmation status for the upcoming schedule, using GET /appointments/{appointmentid}/confirmationstatus against the statuses your practice defined in GET /reference/appointmentconfirmationstatus. For anything still unconfirmed, it reaches the patient and asks a direct question: keeping it, moving it, or cancelling. A confirmed answer gets written back with the PUT on the same confirmation status. A reschedule gets booked into a real open slot of the same appointment type. A cancellation gets recorded with the right reason from GET /appointmentcancelreasons, so your no-show report stays honest.

Staff stop calling down a printed list. They see what changed and handle only the patients who asked for a person.

The linked-appointment trap

One complication shows up in nearly every multi-specialty group, and it quietly creates no-shows that look like patient behavior.

Native reminders fire on the chronologically first appointment of the day. A patient booked for a 9:30 imaging slot and a 10:00 provider visit gets reminded about one of them. They arrive for the scan, leave, and the provider visit becomes a no-show. Nothing in the template enforces the pairing; it exists only as convention.

The fix is to treat the pair as one visit when confirming. When the outreach reaches the patient, it names both appointments, confirms both, and if the patient moves one, it looks for a slot where both legs still fit with the right spacing. If no pairing works, it does not book half of it. It hands the request to the scheduling team with both appointments noted, because booking one leg and losing the other is how the second appointment silently disappears.

Refill the slot the same day it opens

A cancellation only helps if the slot gets used. athenaOne has a native waitlist, readable and updatable through GET and PUT on /appointments/waitlist, and it is usually underworked because offering a slot means phoning down a list while the slot ages.

When a confirmed cancellation lands, the open slot can be offered straight away to waitlisted patients who match the appointment type, provider and department. The first patient who accepts is booked, and the waitlist entry is updated so nobody gets a second offer for a slot that is already gone.

There are rules a backfill has to respect. Many practices keep two lead-time rules: new patients cannot book inside three or four business days, so forms get completed, while established patients can book next day. A freed slot tomorrow morning goes to an established patient, not a new one. Where a rule like that is not written down anywhere, the workflow sends the slot to staff rather than guessing.

What to hand to a human

Automation should cover the routine answers and pass on anything that needs judgment or a relationship. Good handoff points:

  • A patient who says they are not coming because of cost or coverage. That goes to billing or the front desk, not into another reminder.
  • A patient with repeated no-shows where the practice has a dismissal or fee policy. Applying the policy is a staff decision.
  • Anything the patient says about how they are feeling. The outreach books, moves or cancels. A question about symptoms is routed to the clinical team’s queue untouched.
  • A reschedule that needs an appointment type or provider the rules do not cover.

Each of those arrives as a task in athenaOne with the call summary attached, so nobody has to call the patient back just to learn why they were calling.

How to measure whether it worked

Track four numbers, monthly, by appointment type and department:

  • No-show rate, defined the same way every month.
  • Share of appointments confirmed 48 hours out.
  • Cancellations received with at least a day of notice.
  • Freed slots refilled before the appointment time.

If the confirmed share rises and the late-notice cancellations become early ones, the no-show rate follows. If the refill number stays flat, the waitlist is the bottleneck, not the reminders. For the scheduling-side detail, see our pieces on confirmation status as a no-show predictor and how waitlist backfill data is structured.

Key Takeaways

  • Reminders are table stakes. MGMA’s 2026 poll shows practices using them still split three ways on whether no-shows improved.
  • Cut no-shows by appointment type, department and provider in athenaOne before choosing a fix. The loss usually concentrates in a few places.
  • The unconfirmed list 48 hours out is the working list. Reach those patients again and write the answer back to confirmation status.
  • Confirm linked appointments as a pair, or the second one quietly becomes a no-show.
  • A cancellation is only a win if the waitlist fills the slot the same day, within your lead-time rules.

Most of what reduces no-shows in an athenaOne practice is not a new tool. It is closing the loop the reminder opens, every day, on the appointments most likely to go empty. That is the work Pretty Good AI does inside athenaOne, and it hands everything else to your staff with the context already attached.

Sources

Ready to See It in Action?

See how Pretty Good AI works unconfirmed appointments and fills open slots inside athenaOne

Schedule a Demo →

Written by Kevin Henrikson