How to Improve Patient Recall at Your Optometry Practice

Every optometry practice has a recall list. Almost none have a week where someone actually worked it.

Optometry patient recall tends to live in the same place as the filing nobody got to. It is real, it is valuable, and it is permanently third in line behind the patients in the waiting room and the phone that will not stop ringing. So the list grows. Patients who were once loyal drift from six months overdue to eighteen, and the list stops feeling like an opportunity and starts feeling like a reminder of what the practice does not have time for.

Most owners ask how to get through it faster. The more useful questions are which parts of the list are worth working at all, and what to do differently the moment a patient says yes.

Where these numbers come from

A recare calling analysis covering 12,662 past-due patients, called across 17 offices over three months, matched against real outcomes rather than stated intent.

Of those 12,662 patients, 287 booked an appointment. What makes the analysis worth reading is not the headline conversion rate. It is that every call was checked against two real outcomes: did the patient book, and did the patient walk in.

Five patterns came out of it. Four of them change how a recall list should be worked.

Who should you call first?

Recently overdue patients. Conversion falls steadily the longer someone has been past due.

Patients between six and nine months overdue converted at about 4%. Past 36 months, conversion was zero. 908 patients called, none booked.

Sit with that second number for a moment, because it is not really a statistic. It is several weeks of somebody’s afternoons. Nine hundred conversations that opened politely, went nowhere, and ended with a note in the chart. Every one of those calls was time that could have gone to a patient who was eight months out and would have answered differently.

The practical version is simple. Sort the list by how recently the patient went past due, and work the front of it. What tends to happen in most practices is the opposite. The list gets picked up wherever it was left off, which means the oldest and least responsive names sit at the top of the pile.

Does being years overdue mean a patient will no-show?

No. Once a patient books, how overdue they were has almost no bearing on whether they show up.

Show rates ranged from 65% to 89% with no pattern in either direction. Worth being honest about the thin cells here. The 88.9% and the 65.5% each rest on fewer than 30 patients, which is too few to read as a finding on its own. The story is the flatness, not any single row.

This corrects an assumption a lot of practices carry, which is that a patient who has been gone two years is a flight risk not worth an exam slot. The data says otherwise. How overdue a patient is should decide who gets called first. It should not decide who gets treated as unreliable once they say yes.

That is a more human read of the list, too. A patient who has been away eighteen months usually did not stop caring about their eyes. They moved, changed jobs, lost coverage, had a hard year. When someone calls and makes the next step easy, they come in.

The recall rule most practices can act on this week

Book the appointment inside the next 14 days. Lead time is the single strongest predictor of whether a recall patient actually shows up.

Patients booked within a week of the call showed up 84% of the time. Patients booked more than 60 days out showed up 63% of the time. The middle bands do not fall in a clean line, so this is best treated as a rule about the two ends rather than a precise curve.

Here is why this matters more than anything else in the analysis. How overdue a patient is belongs to the patient. Lead time belongs to the schedule, which means it belongs to the practice. It is the one variable in the entire data set that an office can change on a Monday morning.

There is a human reason behind the number as well. The yes on a recall call is genuine, but it is built on a moment. Someone caught the patient at the right time, reminded them it had been a while, and made the next step feel small. Two months later that moment is gone. The appointment is now an obligation on a calendar, competing with everything that showed up in between.

Practices that get this right hold a handful of exam slots inside the next two weeks specifically for recall bookings, rather than offering whatever is left at the back of the schedule. Protecting near-term availability feels counterintuitive. It is also the difference between a recall program that fills chairs and one that fills a calendar with people who will not arrive.

How many times should you call the same patient?

Three. Conversion holds steady across the first three attempts and then collapses.

The first three attempts converted at 5% to 6% each. The fourth converted at well under 1%. The four-attempt group is far larger than the first three, so the comparison is not perfectly clean, but the size of the drop is hard to argue with.

There is an operational read here and a human one, and they point the same way. Operationally, a fourth cycle spends real hours for almost nothing. Humanly, a patient who has not responded to three calls has communicated something. Continuing to dial does not persuade them. It wears out the relationship and the person making the calls.

Three attempts, then move the name to a different channel or a later quarter. Nobody has ever booked out of irritation.

What about patients with no insurance on file?

They converted at roughly half the average rate.

That is not surprising, and it is not a reason to skip them. It is a reason to give them something different to say yes to. The analysis points to in-house membership plans as one option practices use to lift conversion among patients without coverage.

Plan design belongs with whoever advises the practice on that side of the business, not with the person making recall calls. But it is worth naming what is happening on the other end of the phone. A patient without coverage who wants the exam is usually doing quiet math while the caller is talking. Giving them a number they can plan around changes the answer more than a better script will.

Turning recall into a system instead of a task

Score the list, tier it, and work it in order every week. That is what turns recall from something a practice intends to do into something it does.

A working version of that scoring model:

Tiers follow from the score. Tier A at 70 and above gets called first. Tier B from 40 to 69 gets called after Tier A. Tier C below 40 is low-yield and gets worked only when the higher tiers are clear.

Two things about this model deserve stating plainly rather than being buried. First, the recency scale was rebuilt on the back of this data. Patients between six and twelve months overdue now score the same, because they converted at nearly identical rates and the previous version was penalizing the nine to twelve month group for no reason anyone could point to. Fifteen to 24 months was flattened for the same reason.

Second, the heaviest weight in the model is the one piece that has not been validated. Historical treatment value was never pulled, so there was no way to test whether it predicts anything. It is being tested in a pilot. Any practice building its own version of this should treat that category as a working assumption, not a proven weight.

Who actually makes the calls

All of the above is process. Process is not the constraint. 

Every piece of this assumes someone has three or four uninterrupted hours a day for outbound calling. Working a sorted list. Protecting near-term slots. Tracking attempts. Stopping at three. A front desk that cannot reliably answer inbound calls is not going to generate that volume on top of it. Recall does not fail in most practices because nobody knows how to do it. It fails because the person who would do it is standing at the counter with a patient in front of them and two lines ringing. 

That is a capacity problem, which is why a virtual assistant for patient recall works as well as it does. Teem places skilled remote team members inside independent optometry practices to handle exactly this work: recall and recare calling, phone coverage, scheduling and rescheduling, and insurance verification. Not instead of the in-office team. Alongside it, so the people at the counter can stay with the patients in the building. 

Recall calling is its own skill, and it is worth saying what good looks like. Warmth that does not sound scripted. Reading in the first ten seconds whether someone is busy or genuinely interested. Offering two specific times instead of asking an open question. And hearing when a patient has said no without saying it. 

Every placement comes with structure around that person. Teem University is the training platform the practice uses to assign role-specific courses and keep building on them as the placement develops. The Performance Tracker shows the practice what is actually happening day to day, including calls made and appointments booked. And every client works with a dedicated Success Manager. You can see how the process runs end to end before committing to anything. 

Teem is also adding AI support behind its remote team members, with the practice’s approval. We call this approach Human-Led, AI-Amplified. In practice it means AI helps the person move faster through routine steps and gives them coaching drawn from their own real work. It never speaks with a patient, and a person makes every decision that touches one. There is more on what the model looks like inside a practice if you want the longer version. 

If recall is always the thing that waits

The pattern in most practices is not that recall gets ignored. It is that recall gets worked when there is time, and there is never time. 

If that sounds like your practice, a discovery call is a straightforward place to start. Fifteen minutes to talk through what your recall list actually looks like, how far behind it is, and who could work it consistently. If you would rather run the numbers first, the cost savings calculator will get you most of the way there.