Emergency Medicine Follow-Up After the Visit Ends

Emergency medicine follow-up is the largest unowned workflow in American outpatient care. The patient is discharged, handed paperwork that says to follow up with their doctor in a few days, and walks out into a gap that no one on either side of it is responsible for closing.

The handoff is a sentence on a piece of paper. That is the entire mechanism.

The emergency group’s involvement formally ends at discharge. The primary care office does not know the visit happened unless the patient tells them or a record arrives and someone reads it. The patient, who is tired and was recently unwell, is expected to make the connection themselves, from home, during business hours, on a phone tree.

A meaningful share of them do not. They feel better, or the number was busy, or they could not get an appointment inside the window they were told about, so they wait. Some of those patients come back to the emergency department, which is the most expensive possible version of a follow-up appointment.

Nothing about closing this gap requires a clinician. It requires somebody to call the patient, confirm they have an appointment, and help them get one if they do not. It is a booking problem wearing a care coordination costume.

The volume that goes home

The scale of this is easy to underestimate because the visible part of emergency medicine is the part that gets admitted.

CDC reports 155.4 million emergency department visits in a year, with 11.5% of visits resulting in hospital admission. The overwhelming majority go home. Every one of those discharges carries some instruction about what happens next, and for most of them the next step is an outpatient appointment that does not yet exist.

That is the population. It is not a niche, and it is not the sick end of the distribution. It is ordinary people who were seen, treated, and sent home with a plan that depends on their own follow-through.

For a group that owns any part of the downstream relationship, whether that is affiliated clinics, a health system partnership, or its own follow-up program, this is the largest addressable list in the organization and the least worked.

What the follow-up call is, and firmly is not

This is the article’s most important paragraph, because the temptation to overreach here is severe and the consequences are real.

The call confirms logistics. Did you get your paperwork. Do you have a follow-up appointment scheduled. Would you like help making one now. Do you know where you are going and when. Were you able to fill what you were given. Do you have questions for the office you are following up with, and if so, here is how to reach them.

The call does not ask how the patient is feeling in any way that invites an answer the system would have to act on. It does not check whether symptoms improved. It does not decide whether the patient needs to be seen sooner. It does not evaluate anything.

That line is not squeamishness, it is the design. A follow-up program built on logistics can run at full volume across every discharge, because none of it requires a license. A program that starts sorting patients by how they sound has quietly become something else, with a different risk profile and a different regulatory footing, and it will be staffed and scaled accordingly, which means it will not reach everyone.

So the choice is between a narrow program that reaches every discharged patient and a broad one that reaches a few. Narrow wins, and it wins by a lot.

Booking is what makes the call worth making

A follow-up call that ends with advice to call your doctor has moved nothing. The patient already knew that.

The version that works ends with an appointment. That means the outreach has to be able to see the encounter that generated the follow-up need, check whether the patient already has an upcoming appointment, and where they do not, offer real availability and book it.

In athenaOne those are ordinary reads. Recent encounters establish who is on the list and when they were seen. A patient’s appointments establish whether the loop is already closed, which matters enormously because calling patients who already booked wastes contact and irritates them. Open slots make the offer concrete.

Where the follow-up belongs to a practice outside the group’s system, the achievable version is warmer than a reminder and less than a booking. Confirm which office the patient is following up with, give them the number, offer to connect them, and record what happened. Even that is a substantial improvement over a line on a discharge sheet, because somebody now knows whether the handoff landed.

Timing and attempts, which is where these programs actually fail

Most post-discharge outreach fails on mechanics rather than concept.

Call too early and the patient is asleep or has not been home long enough to have looked at their paperwork. Call too late and the window they were told about has closed. The workable pattern is a first contact inside the first day or two, then a defined attempt schedule across different times of day, because a patient unreachable at ten in the morning is often reachable at six in the evening.

Channel matters as much as timing. A phone call from an unknown number after an emergency visit goes unanswered constantly. A text that identifies the sender and offers a callback or a booking link converts better and costs less, and the two together beat either alone.

And the attempt schedule needs a defined end. After the agreed number of attempts across the agreed channels, the case closes with its history recorded rather than looping forever. AHRQ maintains extensive resources on care coordination as a discipline, and the recurring theme is that handoffs fail in the space between organizations, which is exactly the space this program operates in.

What good looks like on a report

Four numbers, and they should be reviewed by someone with authority to change the program.

Share of discharges contacted inside the target window. Share reached, which is different and usually lower than anyone expects. Share who already had an appointment, which tells you how much of the program is confirmation rather than rescue. And share who did not have one and left the interaction with one booked, which is the number the program exists to produce.

The fourth number is the only one that represents work created out of nothing. Track it per site and per shift and the operational picture gets clear fast.

One caution is worth stating about interpretation. A high already-booked share is good news about the population and bad news about the targeting, since it means contact is being spent on patients who did not need it. Filtering on existing appointments before the call is cheap and it concentrates the effort where the gap actually is.

Key Takeaways

Most emergency department visits end in discharge, so the follow-up population is the majority of volume rather than an edge case.
Keep the call to logistics: paperwork, appointment, location, timing, and questions for the receiving office. No condition questions.
End the call with a booked appointment where you can, because advice to call your doctor moves nothing.
Filter out patients who already have an upcoming appointment before calling, to concentrate contact where the gap is.
Use a defined multi-attempt schedule across times of day and channels, with an explicit stopping point and recorded history.
Report contacted, reached, already booked, and newly booked, and treat newly booked as the program’s actual output.

The gap between an emergency visit and the appointment that should follow it is a phone call nobody was assigned to make. Assign it, keep it to logistics, and measure the appointments it creates.

Sources