Provider Offboarding: What to Automate in the Final 60 Days

Provider offboarding is a rare event at a single practice and a standing operational load at an MSO. If you support forty sites, somebody is inside a notice period essentially all the time, which changes what good looks like. A single practice can afford to improvise a departure. An organization that runs one a week cannot, and the ones that keep improvising pay for it in patients who quietly do not come back.

The difference between MSOs that handle this well and MSOs that do not is almost never effort. It is whether the work is a playbook or a scramble. In a scramble, the regional manager finds out about the booked panel in week four, the site tries to call through it in the last ten days, and the practice absorbs whatever it did not get to. In a playbook, the panel is pulled the day notice is logged, the rebooking runs on a schedule, and the regional team is watching a burn-down instead of taking a phone call about it later.

The runway is sixty days and most of it gets wasted at the front

Notice periods cluster around sixty days, and the first three weeks of that are usually spent on everything except the schedule. Credentialing gets started, recruiting gets briefed, the clinical coverage conversation happens. The booked appointments sit untouched because no single role owns them.

That delay is the whole problem, because rebooking capacity is a rate, not a lump. Three hundred appointments over six weeks is manageable. Three hundred over ten days is not, and no amount of urgency in the last stretch recovers the weeks already spent. An MSO can fix this once, centrally, by making the panel pull an automatic consequence of logging a departure rather than a task someone remembers.

The standing playbook is short. Day one, pull the booked panel across every department and site the provider touches. Day two, the clinicians mark which visits have to keep their window. Day three, rebooking starts on the rest and runs daily until the list is clear.

At MSO scale the panel is never on one schedule

A provider at a single-site practice has one schedule. A provider inside an MSO very often does not.

They cover a satellite two days a month, they hold a block at a second location under a different department, they carry a telehealth template that lives somewhere else again. Pulling the booked set for one provider therefore means reading across departments rather than pulling one calendar, and this is the single most common way a departure plan comes up short. The site that got missed finds out on the day, when patients arrive for a provider who no longer works there.

The fix is structural rather than diligent. Resolve the provider record first, enumerate every department they are active in, and pull booked appointments across all of them as one set. Then reconcile that count with what the site managers believe they have. When those two numbers disagree, and they frequently do, the query is right and the recollection is wrong.

Standardize the offer, localize the slots

The reason departures go badly across a portfolio is that every site invents its own script under time pressure, and the quality of the patient experience ends up tracking whoever happened to be at the desk.

What should be standard across the organization: what patients are told about why the appointment is moving, the order the panel gets worked in, how many contact attempts each patient gets before a letter goes out, and what happens when someone asks for their records instead of a new date. That last one deserves a real answer prepared in advance, because it is the moment the patient is deciding whether to stay.

What has to stay local: the actual slots. Which colleague absorbs which visit type, what the receiving provider’s template allows, and which location is a reasonable ask for that patient. Centralizing the offer while localizing the inventory is the whole trick, and it is what lets one automated calling workflow serve forty sites without sounding like a call center that has never heard of the practice.

The load lands on phones that are already the constraint

Rebooking a panel is outbound work stacked on top of inbound demand that has no slack. When MGMA asked practice leaders to name their top patient access priority for 2026, the answers split across no-shows at 27%, online scheduling at 24%, phone access at 22% and wait times at 21%. Phone access sits near the top of the list in normal operations, before anyone adds a departure to it.

At portfolio scale this compounds in a way single sites never see. Two concurrent departures in one region is not twice the work, it is twice the work landing on the same regional float staff who were the contingency plan for the first one.

The part that automates cleanly is the repetitive middle: place the call, explain that the appointment needs to move, offer the approved slots for that visit bucket, write the reschedule back to the record, retry the ones that did not answer, and escalate to a person the moment the patient asks something outside the script. The decision about which visits could move at all was made in week one, by the clinicians. What is left is contact volume, which is exactly the kind of work that should not be rationed by how many people happen to be free.

What the regional dashboard should show

Departures are a portfolio metric, not a site anecdote, and the reporting should match.

For each active departure: panel size, share confirmed with a new date, days remaining, and the implied rate needed to finish. A site trending short shows up in week two, which is when help still helps.

Across departures: retained share, meaning the proportion of each panel that rebooked with the organization rather than cancelling. This is the number that tells you whether the playbook is working, and it is the one nobody tracks because a cleared schedule looks the same either way.

And one leading indicator worth watching separately: new-patient wait time at the receiving sites. If retained share is high but new-patient access got worse, the panel was absorbed by crowding out new demand. That trade can be the right call for a quarter. It should never be one you discover after the fact.

Key Takeaways

Make the panel pull an automatic consequence of logging a departure. The first three weeks of a sixty-day notice period are where the runway gets lost, and no urgency at the end recovers them.
Enumerate every department the provider is active in before pulling appointments. The missed satellite site is the most common failure, and patients discover it by showing up.
Reconcile the queried panel count against what site managers believe they have. When those disagree, trust the query.
Standardize the script, the working order, the contact-attempt count, and the records-request answer across the portfolio. Keep slot inventory local to the site.
Two concurrent departures in a region is not twice the work, because both fall back on the same float staff. Plan capacity regionally, not per site.
Report panel size, confirmed share, days remaining and required rate per departure, plus retained share across departures and new-patient wait time at receiving sites.

For an MSO the argument for automating provider offboarding is not that any single departure is hard. It is that departures never stop, the work is nearly identical every time, and the cost of doing it badly is invisible in the moment and permanent afterwards. A patient who could not get a new date does not file a complaint. They just book somewhere else, and the practice attributes the gap to attrition six months later without ever connecting it to the last day of a physician who left in March.

Sources