Physical Medicine and Rehab Prior Authorization Cadence

Rehab prior authorization is a different animal from the one most practices are set up to handle. In almost every other specialty, an authorization is a gate you clear once before a procedure. In physical medicine and rehabilitation it is a budget, denominated in visits, spent a little at a time, expiring on a date, and running out while the patient is mid-course and the front desk is looking at something else.

The structural problem is that the approval and the consumption are separated by weeks.

An authorization arrives covering a number of visits inside a window. The patient starts coming. Each visit spends one unit of something nobody is watching, because the moment of celebration was the approval and the moment of pain is a month away. Then a patient arrives for what turns out to be visit thirteen against a twelve-visit authorization, and the practice has already delivered care it cannot bill.

The patient absorbs the second half of the failure. They are mid-course, they have built the appointments into their week, and now there is a gap while somebody submits paperwork. Continuity breaks for administrative reasons, which is the most avoidable kind.

And the tracking that would prevent this usually lives in a spreadsheet maintained by one person, updated when they have time, and consulted by nobody at the front desk on a Tuesday morning.

The regulatory floor is moving, and it is not the whole answer

The environment is improving, which changes the math without solving the problem.

CMS finalized rule CMS-0057-F requires impacted payers, including Medicare Advantage organizations, state Medicaid and CHIP programs, and Medicaid managed care plans, to send prior authorization decisions within 72 hours for expedited requests, with API and operational requirements phasing in through compliance dates beginning January 1, 2026 and January 1, 2027.

Faster decisions help. They do not change the shape of a visit-limited authorization, which is the part that hurts a rehab practice. A payer who answers quickly still answers a question somebody has to remember to ask, on a date somebody has to calculate, based on a visit count somebody has to maintain.

So the operational work is upstream of the payer entirely. It is knowing, on any given day, which patients are inside four visits of running out, and starting the re-authorization while there is still runway.

Burn-down is the number nobody owns

Ask a rehab practice how many patients are within two visits of exhausting their authorization and watch what happens. Usually somebody opens a spreadsheet.

The data to answer it exists. Authorizations live in the chart with a visit count and a date range, and appointments live on the schedule. Remaining visits is the difference between them, and the reason it is hard is not the arithmetic. It is that nothing computes it continuously and puts it where the front desk can see it.

The useful design is a daily list rather than an alert. Every patient with an active authorization, visits used, visits remaining, days until expiry, and a flag when either crosses the threshold the practice set. Thresholds should be measured in visits and days, because a patient coming three times a week and a patient coming once a week burn the same budget at very different speeds.

Once that list exists, the re-authorization request stops being a reaction to a denied claim and becomes a scheduled task with lead time. That is the entire improvement, and it is worth more than any single payer’s turnaround time.

Rescheduling quietly breaks the link

This one costs practices real money and almost nobody has it written down.

When an appointment carrying an authorization is rescheduled, the new appointment is not automatically attached to the existing authorization. One multi-site practice’s answer was a human team whose job was re-linking. There is a second edge on the same problem: if a patient reschedules an appointment whose authorization has not come back yet, that appointment can only be pushed later, never pulled earlier, because there is no approved date range to schedule inside of.

Rehab schedules reschedule constantly. Patients move appointments around work, weather, and transportation, often several times in a course of care. So the link breaks more often here than in a specialty where the authorized event happens once.

The automation worth having is unexciting. On every reschedule of an appointment tied to an authorization, re-check the link, and if it did not follow, flag it before the visit rather than after the claim. And where an authorization has not returned, enforce the forward-only rule at booking time instead of hoping the scheduler remembers it.

Booking rules that reflect what the plan will actually do

The best-run practices bake authorization behavior into how they offer appointments in the first place.

One group’s rule is simple and worth stealing. A visit that needs no authorization gets booked same-day or next-day. A visit that does need one has its earliest offered date set out far enough for the authorization team to submit and hear back. In practice they generalized it by plan type, because HMO products essentially always require authorization while PPOs mostly do not.

That rule is appointment type and plan type together, not either one alone. It is exactly the kind of logic that lives in a scheduler’s head, degrades when that scheduler is out, and can be stated in one sentence to a system that will apply it identically every time.

The measurable version of getting this wrong is well documented in adjacent specialties: at one multi-site procedural practice, roughly 6 to 8% of procedures were being scheduled outside their authorization window by human schedulers, which surfaces later as denials. The AMA maintains ongoing resources on prior authorization burden precisely because this pattern is industry-wide rather than a local failure.

The front office cannot make a payer faster. It can stop offering dates the authorization will not cover.

What stays with a person

Everything above is counting, calendar arithmetic, and paperwork status. None of it involves deciding what care a patient should receive, and the boundary should be stated rather than assumed.

The clinical content of a re-authorization request, the justification for continued care, belongs to the clinician. The automation assembles the packet, pulls the visit history and the dates, routes it for review and signature, submits it, and tracks the response. It does not write the justification and it does not decide whether continued care is warranted.

A denial goes to a person immediately, with the history attached. So does any case where the visit count in the chart disagrees with the visit count the payer believes, because that reconciliation requires somebody to make a call.

And the patient conversation about what happens if authorization does not come through is a conversation with staff, not with an automated system. That one is not a technical limitation. It is a judgment about which moments in a course of care deserve a human voice.

Key Takeaways

Treat a rehab authorization as a visit budget with an expiry, not a one-time gate, and compute remaining visits continuously.
Publish a daily burn-down list with visits used, visits remaining, and days to expiry, thresholded on both.
Re-check the authorization link on every reschedule, because the new appointment does not inherit it.
Where an authorization has not returned, enforce forward-only rescheduling at booking time.
Set earliest offered date by appointment type and plan type together, so you stop offering dates the plan will not cover.
Keep the clinical justification, denials, and count disputes with staff. Automation assembles, routes, submits, and tracks.

The practices that stop losing visits to authorization are not the ones with a better payer relationship. They are the ones who know every morning which patients are running out, and started the paperwork a week earlier.

Sources