Case Study 2 — The Payment Policy That Was Finalized and Never Happened: E/M, 2018–2021

DataField.Dev

Case Study 2 — The Payment Policy That Was Finalized and Never Happened: E/M, 2018–2021

Built on the public record of federal rulemaking: two Medicare Physician Fee Schedule final rules published a year apart, and the revision of the office/outpatient evaluation and management (E/M) code set that took effect January 1, 2021. Those are Tier 1 and the citations are in this chapter's further reading. No dollar figure, payment rate, or relative value appears in this case study, because every one of them has been revised since and would be wrong by the time you read this. What is durable here is the sequence and what it did to people's expertise — which is why this is the career chapter's second case and not Chapter 15's.

Background: the skill

For roughly two decades, the single most portable technical skill in outpatient coding was scoring an office visit against the E/M Documentation Guidelines.

Two sets of them were in force — one published in 1995 and one in 1997, and a provider's note could be scored under either, whichever was more favorable. Both worked the same way. You counted. Elements of the history of present illness. Systems in the review of systems. Items of past, family, and social history. Then organ systems or body areas in the examination, or, under the 1997 rules, bullets — literal enumerated bullet points, counted against a specialty-specific template. Then medical decision making, scored on its own grid. The level of service fell out of the combination.

This was a real skill and it was hard. It rewarded precision, it took years to get fast at, and it was genuinely difficult to do consistently — two competent auditors scoring the same note under the same guidelines could and routinely did arrive at different levels, which was itself a documented and much discussed problem. An entire professional apparatus grew on top of it: audit tools, scoring sheets, templates built to generate countable elements, education programs, an auditing credential, and a large population of coders, auditors, educators, and consultants whose livelihood was the counting.

It is worth being precise about what that apparatus was for, because it explains what happened next. The counting existed to determine payment. It was not a clinical instrument. Nobody counted review-of- systems items to take better care of a patient. The elements were countable because a payment system needed something countable, and the note grew to fit the count — which is the origin of a great deal of the cloned, bloated documentation Chapter 4 §4.6 describes.

The issue, part one: a policy that was finalized

In the CY2019 Medicare Physician Fee Schedule final rule, published in November 2018, CMS finalized a substantial restructuring of how office and outpatient visits would be paid.

WHAT WAS FINALIZED IN NOVEMBER 2018, EFFECTIVE CY2021
[the public record; read the rule itself, not this summary]

  * A SINGLE BLENDED PAYMENT RATE for office/outpatient E/M
    visit levels 2 THROUGH 4, for both new and established
    patients -- one payment amount regardless of which of the
    three levels was reported. Level 1 and level 5 retained
    separate rates.

  * ADD-ON CODES to restore some differentiation for
    complexity and for certain specialties, plus a prolonged
    services add-on.

  * DOCUMENTATION FLEXIBILITIES, several of which took effect
    in 2019 rather than 2021 -- including that a practitioner
    need not RE-DOCUMENT information already entered in the
    record by staff or by the patient.

  * The effective date was DELAYED to 2021 in the final rule,
    from the 2019 date that had been proposed.

Read what that meant for the skill. If levels 2, 3, and 4 all pay the same amount, then the distinction between a level 3 and a level 4 office visit — the distinction the entire counting apparatus existed to establish, the thing auditors argued about, the thing templates were built to support — stops carrying money. It would still matter for quality reporting, for risk adjustment, for other payers, and for the record's accuracy. It would not matter for the Medicare payment on that claim.

The reaction was substantial and public. Physician organizations, specialty societies, and much of the coding profession objected, on grounds that were partly about payment redistribution across specialties and partly about the policy's effect on documentation incentives. And the profession did what professionals do with a finalized federal rule: it began preparing. Education was written. Impact models were built. Consultants sold analyses of which specialties gained and which lost. Practices modeled the revenue effect. Two years of professional attention went into a policy with a published effective date.

The issue, part two: it was reversed before it took effect

In the CY2020 Medicare Physician Fee Schedule final rule, published in November 2019 — a year later, and a year before the blended rate was to begin — CMS did not implement it.

Instead the agency aligned with a different solution that had emerged in the meantime: the AMA CPT Editorial Panel's revision of the office/outpatient E/M code set, with values developed through the AMA/Specialty Society RVS Update Committee. That revision, not the blended rate, took effect January 1, 2021.

So the single blended payment rate for levels 2 through 4 — finalized in a federal final rule, with a published effective date, prepared for by a substantial part of an industry — never took effect for a single claim.

And what actually happened on January 1, 2021

The revision that did take effect changed the skill rather than the payment structure, and it changed it at the root.

OFFICE/OUTPATIENT E/M, BEFORE AND AFTER JANUARY 1, 2021

  BEFORE   Level determined by HISTORY + EXAMINATION + MEDICAL
           DECISION MAKING, scored against the 1995 or 1997
           Documentation Guidelines. Counting.

  AFTER    Level determined by MEDICAL DECISION MAKING **or** by
           TOTAL TIME on the date of the encounter. History and
           examination are performed as clinically appropriate
           and DO NOT determine the level.

           99201 deleted. A new prolonged-services structure.
           MDM restated around problems, data, and risk.

The counting stopped determining the level. Not gradually, not by attrition — on a published date, for the highest-volume family of codes in outpatient medicine.

Be precise about the scope, because overstating it is the common error: the 1995 and 1997 guidelines did not vanish. Other E/M families continued to use them until subsequent revisions extended the decision-making-or-time framework further, and the guidelines remain historically relevant to any record predating the change and to any payer that had not adopted the revision. But for the office visit — the bread and butter, the thing most outpatient coders spend most of their day on — the scoring skill that had defined the specialty for twenty years was retired by an editorial panel and a final rule.

What it shows

Two different things went wrong here, and conflating them is how people draw the wrong career lesson.

The first is a reversal. A finalized federal payment policy, with a published effective date, was reversed before it took effect. That is not a scandal and it is not evidence of bad faith — it is evidence that rulemaking is a process with more steps after "final" than the word suggests, and that a policy can be finalized, litigated in public, overtaken by a better solution, and abandoned. The people who lost were the ones who had treated "finalized" as "settled."

The second is an obsolescence. A large body of genuine, hard-won expertise stopped being the thing that determined payment. Nobody did anything wrong to those professionals. The rule they had mastered was replaced with a better one — most people in the field agree the 2021 framework is a real improvement, because it scores the visit on the physician's actual thinking rather than on how many systems got documented. The improvement and the obsolescence are the same event.

And notice who was fine. Not the people who had memorized the bullet counts. The people who were fine were the ones whose expertise was the classification and the reasoning — who could read a note, identify the problems addressed, assess the data reviewed and the risk assumed, and defend a level against a reviewer. That skill did not merely survive the change; the change made it more central, because medical decision making went from one-third of the scoring to the whole of it.

The same asymmetry appears in the documentation flexibilities. The practitioner who no longer has to re-document what staff already entered was freed by a change in a rule. The coder who understood why the re-documentation requirement had existed — attestation, authorship, the legal health record (Chapter 4 §4.4 and §4.8) — understood immediately what the flexibility did and did not change about who stands behind the note. The one who had only learned the requirement learned nothing from its removal.

The lesson

"Finalized" is not "settled," and the half of your expertise that is a scoring convention has an expiration date you will not be told in advance.

Specialize in a body of knowledge, never in a configuration — and a payment rule is a configuration, however federal it is, however many years you have spent inside it, and however hard it was to learn.

There is a second lesson here about how to read a change like this while it is happening, and it is the more practical one:

The people who were least disrupted were reading the rules. Not the summaries. The CY2019 rule said what it finalized and when. The CY2020 rule said what it was doing instead. The AMA published the revised code set and its guidelines in advance of the effective date. All of it was available, in writing, on a schedule, months before it mattered, to anybody who had built the habit §40.10 prescribes. The professionals who were surprised in January 2021 were surprised by a document that had been published, free, the previous autumn.

Discussion questions

§40.9 says a specialty is "partly a bet on a payment system." Apply that sentence to a coder who had spent fifteen years becoming excellent at E/M auditing under the 1995 and 1997 guidelines. What part of that expertise was a bet, and what part was not? Be specific — name at least three skills on each side.
The blended-rate policy was finalized in November 2018 and reversed in November 2019, having never applied to a claim. A practice manager asks whether she should have prepared for it. Answer her, and then say what "preparing" should have meant given that it might not happen. (Compare Case Study 1's compliance-date question — the two cases point in opposite directions and reconciling them is the exercise.)
This case study says "the improvement and the obsolescence are the same event." Give one other example from anywhere in this book where a change that was clearly correct destroyed value somebody had built. (Chapter 22's Case Study 2 is one candidate, and it is uncomfortable for a different reason.)
The 2021 revision made medical decision making the whole of the scoring rather than one-third of it. Trace the effect on Account 10-4471: Chapter 15 §15.13 leveled it as a 99214 on MDM, and the note says time was not used. Would that level have been reached the same way under the pre-2021 framework? Say what additional information you would need to answer, and where in Chapter 4 §4.10's note you would look for it.
The people least disrupted were the ones reading the primary rules rather than summaries. That is easy to say and hard to do while employed full time. Design the smallest version of that habit you would actually sustain — name the specific documents, the frequency, and the time budget — and then say honestly which month of the year you would first skip it.
The hard one. A colleague argues that the right conclusion from this case is to stop investing in deep expertise at all, since payment rules change: learn broadly, stay shallow, stay flexible. Answer him. Use §40.9's portable-versus-expiring distinction, and be concrete about what a shallow generalist cannot do that a deep specialist can — including on the day the rule changes.