AI Recall Systems for Dental Practices: How to Stop Losing Hygiene Revenue

Recall is where a dental practice should have a structural advantage. You know who was seen, when they were seen, and what the clinician recommended. Unlike a new patient, a recall patient has already chosen you, completed intake and found the office. The only open question is whether someone reaches them at the right moment and makes it easy to book.

In many practices that "someone" is a front desk team member who is already handling the patients in the building. So recall gets done in bursts: a postcard run, then a call list that gets worked when there is time. This post covers what independent research says about recall intervals and reminders (it is less glamorous than vendor claims), how to measure your own recall, what automation can and can't do, and a revenue example built on ranges instead of promises.

First: The Recall Interval Is the Clinician's Call, Not "Every Six Months"

A recall system is only as good as the interval it carries. Six months is the traditional default, but the evidence for treating it as universal is weak. A Cochrane review updated in 2020 pooled the randomised trials of recall intervals (two studies, 1,736 participants). For adults attending check-ups in primary care, it found little to no difference between risk-based and six-month recall in tooth surfaces with caries, gum bleeding or oral-health-related quality of life over four years, and rated that evidence high-certainty. For children and adolescents the evidence was uncertain (Fee et al., Cochrane Database of Systematic Reviews, 2020).

That review is about check-up intervals in general, not hygiene visits specifically, and we aren't clinicians. The practical point is a design rule: the interval belongs to the dentist and hygienist, set patient by patient, including the shorter intervals clinicians choose for periodontal maintenance. A recall system should read each patient's interval from your records and respect it, not fire a generic "it's been six months" message at everyone.

How Many Patients Are Overdue? The Numbers That Exist

At national level, 66.7% of U.S. adults had a dental exam or cleaning in the past year, according to the CDC's early release of 2025 National Health Interview Survey data (CDC FastStats). That means roughly one adult in three did not. Some of them are your lapsed patients, though many lapsed patients have simply moved or changed practices. The ADA Health Policy Institute's research on access to dental care points to affordability as the main barrier (ADA HPI, Dental Coverage, Barriers, and Outcomes), which matters for how you word a message to someone who hasn't been in for a while.

What about your own recall rate? You will see "the average dental practice recalls 55–65% of patients" quoted widely. The page we checked that publishes that range says plainly that these are compiled ranges rather than figures from a single dataset, and that no government or peer-reviewed source publishes them as an authoritative series (Dentx, recall rate benchmarks). We don't use them. Measure yours instead:

• •Reappointment rate. Of the patients who finished a hygiene visit this month, what share left with their next visit already scheduled? It is the cheapest number to improve, because the patient is standing at your desk.
• •Recall completion. Of the patients who came due in a given month, what share completed a visit within, say, 60 days of the due date? Pick the window and keep it the same each time.
• •Overdue list size. Active patients, seen in the last 24 months, with no future appointment. Most practice management systems can produce this report; if yours can't, that's worth knowing before you buy anything.

What the Research Says About Reminders and Coming Back

One study looked directly at reminders for a preventive dental visit. Researchers assigned 389 dental outpatients to a reminder by postcard, text message or phone call, or to no reminder, and tracked who returned for tooth scaling (reminders came with short health-education messages). Patients who were reminded were about 2.6 to 2.9 times as likely to come back as those who weren't. The confidence intervals are wide (the lower bound of the larger estimate is 1.1), so the size of the effect is imprecise. And even the best-performing group, patients who were both satisfied and reminded, returned at 26% (Cheng et al., 2013, abstract on PubMed).

The sensible reading: reminders help, and they do not bring most overdue patients back on their own. Claims that texts convert "three times better than postcards" or that automation recovers a specific share of your overdue list are things we couldn't trace to a source, and we removed them from this post. Treat any vendor's recovery figure, ours included, as a hypothesis to test on your own list.

Why Manual Recall Usually Underperforms

We don't have a measured number for how much worse manual recall performs, but the mechanics are easy to see:

• •One touch. A single postcard is easy to ignore, and many patients need more than one prompt. Manual recall rarely has the time for a second and third.
• •One channel. Some patients read mail, some answer texts, some only open email. A single channel misses whole groups, and which group a patient belongs to is something you can learn from how they respond.
• •Batch timing. A monthly mailing contacts some patients too early and some after they've already drifted away. Patient-specific timing, based on each person's own due date, avoids both.
• •No easy way to say yes. A card that says "call us to schedule" asks the patient to remember to phone during office hours. A link that shows open times takes less effort, and a phone that gets answered when they do call matters just as much (more on that in our post on dental missed calls).
• •It stops when the desk is busy. The patients overdue the longest are often on a call list that only gets worked when everything else is done.

What Automated Recall Does, and What It Needs From You

An automated recall system replaces the stop-and-start pattern with a process that runs in the background. In outline:

• •Patient-specific timing. Each patient's own interval and due date drive the first message, starting a few weeks before the due date rather than after it.
• •A short sequence across channels. A text first, an email if there's no reply, a further text after a few days, then a task for a person to phone. Then it stops. Nobody should be messaged indefinitely.
• •A way to book in one step. Each message links to a booking page, offers a reply, or lets the patient call an agent that can book. Which of these is possible depends on your scheduling software.
• •A separate track for long-lapsed patients. Someone who hasn't been in for 12 months or more needs a different message from someone due next month, with no judgment and an easy way back.
• •Confirmation and reminders after booking. A recalled patient who books and then doesn't show up has cost you twice. We cover reminders in our no-show post, and the no-show cost calculator shows what that costs at your schedule size.

What it needs from you: a clean recall list, intervals set by your clinicians, a rule for who must never be contacted (patients who moved, transferred, opted out or have an account issue you'd rather handle personally), and the right to text them. It also needs somewhere to read recall status from and write bookings to, which brings us to integrations below.

The Constraint That's Easy to Overlook: Hygiene Chair Capacity

Recall automation fills hygiene schedules, which only helps if you have open hygiene time. In the ADA Health Policy Institute's Q1 2026 poll of dentists (796 responses), 60.3% said they had an adequate number of dental hygienists, so about two in five did not (ADA HPI, State of the U.S. Dental Economy, Q1 2026). If your hygienists are already booked for weeks, a recall system won't add revenue; it will add a waiting list, and the better answer may be to hire or extend hygiene hours first. In the same poll, 32% of dentists said they were not busy enough and could have treated more patients, which is the situation where recall helps most.

The Revenue Math, With Ranges Instead of Promises

Here is a worked example for a practice with 1,500 active patients. Every input is an assumption; replace each with your own numbers from your practice management reports.

Two things to check against your own practice. First, capacity: 180 extra visits a year is about 3.6 more hygiene visits a week over 50 weeks, which has to fit in the schedule. Second, price: the AI Practice System on our dental page is $5,000 to set up plus $1,500 a month ($23,000 in the first year) and covers call answering, reminders and recall together. Recall alone at the low end of this range ($9,900) wouldn't cover that, which is why we think about missed calls, no-shows and recall as three levers added together, not one. If recall is your only problem, a recall feature in your existing practice management software or a basic messaging tool may be the better buy.

We have deliberately not put a number on the restorative treatment that follows a hygiene visit. Whether a visit leads to treatment depends on diagnosed need, and we found no verified conversion figure. An earlier version of this post claimed an additional $50,000–$150,000 and a total of $100,000–$200,000 a year; we removed both because we couldn't support them.

Reactivating Patients Who Have Lapsed

Patients who haven't been in for a year or more are a different job, and a more delicate one:

• •Clean the list first. Some of these patients moved, transferred to another practice or asked not to be contacted. Messaging them wastes effort and can annoy them.
• •Lead with welcome, not guilt. Acknowledge the gap in one line and make booking easy. If cost may be the reason, mention options only if your practice really offers them, such as a payment plan or an in-house membership.
• •Don't imply clinical urgency you haven't documented. A message shouldn't suggest a health problem that the chart doesn't support.
• •Cap the attempts. Three touches over a few weeks, then stop and let the patient come to you.

What a Message Sequence Can Look Like

These are drafts to adapt, not tested copy. Test them on your own patients and watch the reply and opt-out rates. They deliberately carry no clinical detail.

Consent, Privacy and Integrations: What We Do and Don't Claim

Privacy. A message telling someone it is time for their next dental visit is tied to the fact that they are your patient, so assume patient information is involved. Under HIPAA's Privacy Rule, a practice may let a vendor handle protected health information on its behalf only if it has satisfactory written assurances, documented in a written contract or agreement (45 CFR 164.502(e)). We don't claim HIPAA compliance. Our standard setup isn't designed to handle protected health information, so recall messages should carry only what is needed to book, and patients shouldn't share clinical details with the agent. If your practice needs a HIPAA-covered deployment, ask for the business associate agreement in writing before anything goes live, from us or from any vendor.

Consent. Automated texts are regulated. Have your own counsel confirm that you hold the right consent for the messages you plan to send and that opt-outs, such as replying STOP, are honored promptly. This post is general information, not legal advice.

Integrations. We connect to your practice management software where it has an API we can use, and many platforms offer limited or no API access. Where a direct connection isn't possible, we can work from a regular export of the recall list plus a booking link, or hand the booking request to your front desk to confirm. We don't promise a specific integration before confirming it for your system, which we do in the free audit.

A 60-Day Recall Test You Can Run First

1. Pick the cohort. Take the patients who come due in the next 30 days and split them into two groups at random, for example by alternating down the list.
2. Run the new sequence on one group only. The other group gets your normal process, so you have something to compare against.
3. Count at day 60. For each group, how many patients have a scheduled visit, and how many completed one.
4. Be honest about noise. With around 100 patients in each group, a difference of a handful of visits can be chance. If the gap is small, extend the test or enlarge the cohort before deciding.

The Bottom Line on Dental Recall Automation

Recall should be your most reliable source of schedule demand, and a system that sends the right message at the right time can help. The honest version of the claim is smaller than the one most vendors make: reminders help, they don't bring most overdue patients back on their own, the interval should come from your clinicians, and the payoff depends on having hygiene capacity to fill. Measure your own recall, run the 60-day test, and let the result, not a benchmark, decide whether to spend more.

If you'd like a second pair of eyes on your numbers, the AI receptionist and recall setup for dental practices starts with a free audit of your calls, reminders and recall list. And if patients call back after a recall message, the call coverage cost comparator shows what it costs to make sure someone answers, while the missed call revenue calculator estimates what it costs when no one does.