# How to Find Overdue and Lapsed Patients in Your EHR

Recall software does not find your overdue patients, it contacts the list you give it. Overdue is not a field in most systems but a calculation across a last visit date, an interval and a future appointment, each living somewhere different.

**Category:** Articles
**Written by:** Zain Qayyum
**Last updated:** Oct 27, 2026
**Source:** https://signals.medsender.com/articles/find-overdue-lapsed-patients-in-your-ehr

---

## TL;DR

- Recall software does not find your overdue patients. It contacts the list you give it, so the list is the part that decides whether any of it works.

- "Overdue" is not a field in most systems. It is a calculation across a last-visit date, a recall interval and a future appointment, and each of those three lives somewhere different.

- The standard report lies in one predictable direction. It counts patients who have no future appointment, which is not the same as patients who need one.

- Three filters keep the list honest: patients who have left the practice, patients already booked under another provider, and patients whose record says they should not be contacted.

- A list goes stale within a week and an unowned one stops being rebuilt entirely. Give it a named owner with allocated hours before you automate anything that reads from it.

- If the list is too long to work, the problem is the interval, not the outreach.

## What "Overdue" Means in Your EHR, and Why the Report Lies

Most practices start by running whatever the system calls a recall report, and most find it unusable within a week. The reason is worth understanding, because it also explains what to build instead.

A recall report usually answers one question: which patients have no future appointment scheduled. That is a cheap query and it is not the question you asked. A patient with no future appointment may be overdue, may have transferred, may have died, may be booked next week with a partner at another site, or may never have been on a recall cycle in the first place. All of them land in the same report.

The question you actually want answered has three parts. When did this patient last complete a visit, what interval applies to them, and is there anything on the calendar that already satisfies it. Those three facts are rarely in one place. The last visit date sits in the encounter history, the interval is either in a recall field, implied by the diagnosis, or in nobody's system at all, and the future appointment lives in the schedule.

That is why the report lies. It has one of the three facts.

## Building the List From What the Chart Already Holds

Start from the encounter history rather than the recall field, because the encounter history is always populated and the recall field frequently is not.

Pull every patient with a completed visit in a defined window, typically the last three years. Going back further adds names faster than it adds appointments, and a patient who has not been seen in four years is a reactivation problem rather than a recall problem. They are worth a different message and usually a different list.

Then attach the interval. Three sources, in descending order of reliability. An explicit recall interval set on the patient is best, where your system supports one and your clinicians have used it. A diagnosis-driven default is next, which is how most specialty practices actually work even when nobody has written it down. A single practice-wide default is the fallback, and it is a blunt instrument that will put some patients on the list months early.

Finally, subtract anyone with a future appointment that satisfies the interval. This is the step most practices skip, and it is the one that generates the complaint that patients are being contacted about an appointment they have already booked.

The output is a list with a date attached to each name, which is the thing that makes it workable. A list without dates cannot be prioritized and will be worked from the top down until somebody gets bored.

## The Three Filters That Keep the List Honest

Three exclusions do most of the work, and each one exists because of a specific failure that is unpleasant when it happens.

The first is patients who are no longer yours. Transfers out, deceased patients, and patients who have formally left. Most systems hold a status flag for this and most recall reports ignore it. The cost of missing this filter is not a wasted message, it is a call to a family that should not have received one.

The second is patients already handled elsewhere in the practice. A multi-provider or multi-site group will have patients booked with a partner, at another location, or under a different service line. If your list is built per provider, these patients appear on one provider's list while sitting on another's schedule. Build the exclusion against the practice calendar rather than the provider calendar.

The third is contact preference. A patient who has asked not to be contacted by text, or not to be contacted at all, has to be excluded at the list stage rather than at the send stage. Relying on the outreach tool to honor a preference means the preference has to have reached the outreach tool, and in most practices that sync is partial.

## Who Owns the List Once It Exists

A list is a snapshot and it starts decaying immediately. Patients book, cancel and transfer, and a list built on Monday is wrong by Friday.

That makes ownership a real question rather than an administrative one. Somebody needs to own the definition, which means the intervals and the filters, and somebody needs to own the working of it. In most practices these are different people, and in practices where they are the same person the list tends to survive about as long as that person's capacity does.

The practices that make this stick route it through a central function rather than leaving it with each site or each provider. That pattern shows up consistently: where one admin function owns the list and the sites work from it, the list stays current. Where each site maintains its own, the definitions drift and within a quarter no two sites mean the same thing by overdue.

If you are planning to automate outreach from this list, set the refresh cadence before you connect anything. A nightly rebuild is standard. A list that is regenerated monthly will contact patients who booked three weeks ago.

## Deciding Where to Start When the List Is Too Long

The first honest run of this usually produces a number nobody wants to look at. Several thousand names is common in a practice that has never done it.

Resist the instinct to work it from the top. A list that large is not a work queue, it is a signal that your interval is wrong or that it has never been applied. Three questions narrow it into something a front office can actually act on.

Which cohort is both overdue and clinically time-sensitive. That is where to start, and in a specialty practice it is usually a much smaller group than the total.

Which cohort has the highest likelihood of rebooking. Recently lapsed patients rebook at a far higher rate than long-lapsed ones, so a list sorted by how recently someone became overdue will convert better than one sorted by how overdue they are.

What can your front office absorb in a week. Backfilling the schedule only helps if there is room for the patients who respond, and a campaign that books three hundred patients into a schedule with capacity for forty creates a different problem.

Start with the intersection of those three and leave the rest. The remainder is not lost. It is a reactivation list, it behaves differently, and it should be worked with different expectations and a different message.
