Internal Medicine Care Coordination That Fills the Schedule

Internal medicine care coordination is rarely a knowledge problem. Your staff know which patients are overdue for an annual wellness visit, which ones have an open follow-up order sitting in athena, and which ones walked out of the hospital last week without a follow-up on the books. What they do not have is four uninterrupted hours a day to call all of them, which is why the recall list keeps growing while the phones keep ringing.

The math is unforgiving in a chronic-heavy panel. Every refill request, every result callback, and every specialist status check lands on the same phone line as the outreach work, and inbound always wins because inbound is loud. Outreach is the thing that gets done when there is time, and there is never time.

Meanwhile the schedule is not actually full. The single most requested analysis in our customer calls is a location-by-location view of open appointment slots against the patients who need to be scheduled, and the picture is almost never balanced. Some locations show more than a hundred open slots and a dozen pending outreach items. Others show three open slots and forty patients waiting to be called.

Both halves of that are a revenue problem, and neither one gets solved by hiring another scheduler to sit on the phone.

The recall list is already in athena. Nobody is working it

Internal medicine practices generate outreach work automatically. Open orders and follow-up ticklers accumulate in athena as visits close, wellness visit recalls come due on their own clock, and post-discharge follow-ups arrive with a short window attached. Medicare covers an annual wellness visit every year for beneficiaries who have had Part B for more than twelve months, which is a predictable, bookable volume for any adult panel.

What makes it stall is that the work is outbound and repetitive. A person has to call, reach a patient who is at work, leave a message, note the attempt, and try again in three days. It is the first thing dropped on a busy morning and the last thing picked up.

AI that reads the athena order and tickler queues can run that outreach continuously instead of in bursts. It calls the patient, offers slots that are actually eligible for that visit type and provider, books the appointment, and writes the result back onto the task so staff can see what happened without asking. That continuous outbound stream is what turns recall work into real schedule optimization rather than a list somebody feels guilty about.

The primary provider field is wrong, and it routes the call

Here is a complication that catches most automation projects late. The chart’s primary provider field is stale almost everywhere, because patients change doctors and nobody goes back to update it. You can only trust that field if the practice has been disciplined about maintaining it, and most have not.

That matters for recall outreach specifically, because the whole point is to get the patient back to the right clinician. Booking a wellness visit with a physician the patient stopped seeing two years ago produces a cancellation, or worse, a visit that annoys everyone involved.

The fallback that works is behavioral rather than declarative: look at who has actually seen this patient recently, and offer that provider first. When the recent history is ambiguous, the call captures the patient’s own preference and the appointment goes to a human to place. The automation does not guess about continuity, it either has evidence or it hands off.

Two open items on one patient, and both of them call

A common failure is a patient having an open follow-up task and an open order at the same time. Working the task does not close the order, so the order stays alive and the patient gets called a second time about something they already scheduled. Compounding it, staff move items between queues by hand, and an item that belongs in one bucket can get moved into the queue the automation reads from.

Whether completing a booking satisfies and closes the underlying task is a per-practice policy, not a universal rule. Scheduling may or may not close the tickler, and every clinic configures it differently; closing an order that has an authorization linked to it has its own quirks.

So the rule gets written per practice, up front: which queue is authoritative, what closes a task, what closes an order, and what happens when both exist on the same patient. It is unglamorous configuration work and it is the difference between outreach that patients find helpful and outreach that makes the practice look disorganized.

Enrollment logistics, not clinical management

Chronic care programs carry an administrative tail that has nothing to do with medicine. Patients have to be told the program exists, consent has to be captured, and the recurring appointments have to land on a calendar. Medicare’s chronic care management benefit has explicit enrollment and consent requirements before any of the ongoing work can be billed.

That enrollment logistics layer is automatable. The outbound call explains the program in the practice’s own scripted language, captures the consent the practice requires, books the first appointment, and routes the documentation to the right place in the chart.

Everything past that point belongs to the clinical team. Which patients belong in the program, what happens during the visits, and what the care team does with the results are decisions for clinicians, and our system never touches them. It moves the paperwork and fills the calendar so the clinical staff spend their time on the part only they can do. The scripting and consent capture follow the same rules as any other primary care patient communication workflow.

What the schedule looks like when outreach runs continuously

The measurable change is not call volume, it is the gap between open capacity and pending outreach closing. When the same engine can see both sides in athena, it can work the pending list against the slots that are actually open at that location, for that provider, in that appointment type.

AHRQ’s work on ambulatory safety points at the same operational truth from a different direction: the follow-up steps between visits are where things fall through, and the fix is a reliable process rather than more reminders to staff.

A practice can wire the open-slots-versus-pending view to its marketing spend, turning campaigns on and off as capacity opens. That is the version of this that pays for itself twice. Both sides of that view come from working directly inside athena, not from a nightly export.

Key Takeaways

Work the athena order and tickler queues as a continuous outbound stream, not a project someone picks up when the phones go quiet.
Do not route recall calls off the chart’s primary provider field. Use who has actually seen the patient recently, and hand off when the history is ambiguous.
Write down, per practice, what closes a task and what closes an order. Patients getting called twice is a configuration failure, not a bad list.
Split chronic care work into enrollment logistics, which automates cleanly, and clinical management, which does not and should not.
Track open slots against pending outreach by location. That single view usually explains more revenue than any call volume report.
Capture consent and preferences on the outreach call itself so staff are never re-collecting information the patient already gave.

Internal medicine practices do not need a longer recall list. They need the one they already have to be worked every day, against the slots that are actually open, with the results written back where staff can see them. That is administrative work, it happens inside athena, and it is the kind of steady volume that compounds quietly across a chronic panel.

Sources