Every group practice office manager has had the meeting where a partner insists their patients love them, their chair is full, and retention is not their problem. The production report backs them up. Then a recall interval report by provider lands on the table, and the story falls apart. That same provider has 340 patients past their assigned recall date, forty of them more than a year overdue, and an interval adherence rate ten points below the practice average. The chair is full because new patients keep arriving, not because the old ones keep coming back. This is the report that separates the two.
Most practices never look at recall this way. They watch a single practice-wide reactivation number, feel fine when it holds steady, and never notice that the average is being propped up by two strong providers while a third quietly bleeds patients out the back door. Segmenting recall by provider is uncomfortable precisely because it names names. It is also the fastest way to find the money you are already losing.
What a Recall Interval Report by Provider Actually Measures
A recall interval report by provider is not a visit count. Visit counts reward whoever has the busiest schedule, which tells you nothing about whether patients are returning on the cadence their care requires. The report you want measures two things per clinician, and both are ratios.
The first is interval adherence: of the patients assigned to this provider whose recall came due in a given window, what share rebooked inside their assigned interval? If a hygienist is on a six-month recall and 100 of her patients came due in March, interval adherence asks how many booked their next visit by roughly the end of that window rather than drifting into overdue territory.
The second is reactivation rate: of the patients who did go overdue, what share eventually came back? A provider can have mediocre interval adherence but a strong reactivation rate if the practice works its overdue list hard. The dangerous combination is low on both, because it means patients slip past due and then never get pulled back.
Layer in a third dimension, months overdue, and the report becomes diagnostic instead of merely descriptive. A provider leaking patients at one to three months overdue has a rebooking-at-checkout problem. A provider whose patients pile up at nine to twelve months overdue has a recall-outreach problem. Same low adherence number, completely different fix.
flowchart TD
A[Recall comes due for a provider] --> B{Rebooked in interval}
B -->|Yes| C[Counts toward interval adherence]
B -->|No| D[Patient goes overdue]
D --> E{Recall outreach worked}
E -->|Yes| F[Reactivation rate rises]
E -->|No| G[Patient drifts to attrition]
G --> H[Silent revenue loss on provider column]Why the Practice-Wide Average Hides Your Worst Leak
Blended numbers lie by design. Suppose your practice reports a 78% reactivation rate and the partners nod, because 78 sounds respectable. Break it apart and you might find Dr. A at 88%, Dr. B at 84%, and Dr. C at 61%. The 61 is invisible in the average because the two strong columns carry it. Yet Dr. C, with a panel of 1,400 active patients, is losing hundreds of visits a year that the other two would have retained.
Put dollars on it. If Dr. C's panel should generate two recall visits a year and 39% of due patients fail to reactivate, that is roughly 1,092 lost visits annually against a full-adherence baseline. At a conservative $180 blended value per recall visit, the gap between Dr. C and a provider running at 84% is well over $150,000 a year in production that never gets booked. None of that shows up as a line item. It shows up as a chair that feels busy because new patients keep filling the holes.
The per-provider cut also protects you from blaming the wrong person. Front desk turnover, a provider who stopped rebooking at checkout, a recall list that was never worked because one coordinator quit, all of these produce the same symptom in the blended number: a slow, unexplained softening. Only the segmented report tells you whether the leak is one provider, one visit type, or one broken step in the workflow.
Reading the Report: Four Patterns That Point Straight to the Cause
Once you have the grid, the shape of the data tells you what to fix. Four patterns recur across group practices.
Full chair, low adherence. The provider's schedule looks healthy, but their overdue count is huge and adherence is low. New-patient volume is masking attrition. The fix is not marketing for more new patients; it is plugging the recall leak so the panel stops churning underneath.
Adherence cliff at checkout. Patients rebook well at zero to one months but fall off a cliff by three months. This is a rebooking-at-checkout failure. The patient left without a next appointment on the books and no reliable system pulled them back. This is the single most common and most fixable pattern.
Channel mismatch. Split the report by channel and one provider's patients only ever reactivate after a phone call, while another's respond to text. If your outreach is text-only, the phone-responsive provider's column looks like a retention problem when it is really a channel-coverage problem.
Overdue pile-up at 9 to 12 months. Patients are not being worked once they age past the first reminder. Someone sent one text at the due date, got no reply, and the patient fell into a queue nobody touches. Every month those patients sit unworked, the odds of reactivation drop.
A worked example makes the diagnostic power concrete. Say the report shows Dr. C with 210 patients at zero to three months overdue, 80 at four to eight, and 50 at nine to twelve. That front-loaded shape says the checkout rebooking step is failing first, and the older buckets are simply yesterday's checkout failures aging forward. Fix the rebooking-at-checkout gap and the entire distribution shifts left over the next two quarters. Now compare a provider with a flat distribution, 90 patients evenly spread across every overdue bucket. That shape says the checkout is fine but nobody works the list after the first reminder, which is an outreach-capacity problem, not a checkout problem. Two providers, two low adherence numbers, two entirely different Monday actions, and only the interval-bucketed report tells them apart.
flowchart LR
A[Read provider column] --> B{Where do patients drop}
B -->|At checkout 0 to 3 mo| C[Fix rebooking step]
B -->|One channel only| D[Add voice and email outreach]
B -->|9 to 12 mo pile up| E[Work the aged overdue list]
C --> F[Adherence recovers]
D --> F
E --> FTurning the Report Into a Working Recall Engine Instead of a Blame Sheet
The report only creates value if it changes what happens on Monday. The mistake most practices make is treating a weak provider column as a performance review. It is not a person problem; it is a workflow problem. Dr. C's patients are not disloyal. They simply never got worked after the first reminder went unanswered, and no coordinator had the hours to chase a 340-patient overdue list by phone.
That is where automated recall does the labor the report exposes. Instead of one coordinator manually sorting a queue and giving up after a text or two, an automated system takes every overdue patient on every provider's list and works them across text, voice, and email until they either rebook or explicitly decline. A patient who ignores a text gets a voice reminder; one who misses the voice call gets an email with a booking link. Because the outreach is multichannel, the channel-mismatch pattern dissolves, the phone-responsive provider's column recovers without you having to hire a caller. CallSphere's self-filling scheduling and automatic recall handle exactly this, pulling due and overdue patients back onto the calendar without adding a front-desk seat; the specifics of how the recall and reminder workflows chain together live on the /features page.
Just as important, the system feeds the numbers back into the report. As reactivations land, each provider's adherence and reactivation columns move in real time, so the Monday meeting stops being a lecture and becomes a scoreboard. When Dr. C's column climbs from 61 to 80 over a quarter, nobody had to be reprimanded. The overdue list simply got worked the way a fully staffed front desk would have worked it if it ever had the hours, which it never does.
Making the Report a Standing Habit, Not a One-Time Autopsy
Run this report once and you find one leak. Run it every month and you catch leaks while they are small. A provider whose adherence slips three points in a single month is a far cheaper fix than one whose panel has been quietly aging out for two years. Set a standing cadence: pull the recall interval report by provider on the first of the month, sort by overdue count and adherence together, and put the bottom column on the agenda.
The economics make the habit worth keeping. Recovering even a fraction of one provider's lapsed panel typically pays for the automation many times over, which is why practices tend to look at the retained-visit math against the cost on the /pricing page and find the decision straightforward. Patient recall best practices in 2026 are less about sending more reminders and more about measuring the right thing per provider and then letting an automated engine work the list the measurement exposes. The report tells you where you are losing patients. The workflow decides whether you keep losing them. Start with the column at the bottom, work it until it moves, and then go find the next one.