Case Study 1 — The Backlog at OMHA: When the Appeal System Itself Failed

Real and public. The Office of Medicare Hearings and Appeals backlog is one of the best-documented administrative failures in the Medicare program's history — litigated in federal court, tracked in quarterly status reports filed with a judge, and studied by the Government Accountability Office and the HHS Office of Inspector General. Specific figures below are stated at the precision the public record supports and no further; where a detail is characterized rather than quoted, it is labeled.

Background

Section 30.6 presented the five levels of Medicare appeal as a clean ladder: file within the window, receive a decision within the statutory timeframe, climb if you disagree. Level 3 — the administrative law judge hearing at the Office of Medicare Hearings and Appeals (OMHA) — carries a statutory decision standard of 90 days.

For the better part of a decade, that standard was fiction. This case study is about how a well-designed appeal system was overwhelmed, what that did to the providers inside it, and what it took — a lawsuit, a court-ordered timetable, settlements priced in cents on the dollar, and a congressional funding increase — to dig it out.

What happened

The volume arrived faster than the capacity. In the early 2010s, several forces converged on OMHA's docket at once. The Recovery Audit Contractor (RAC) program went national in 2010 (Chapter 37 §37.5), generating large volumes of postpayment denials — prominently hospital inpatient-status denials, the "should this have been observation?" question from Chapter 16 §16.3 — and providers appealed them at scale, because the amounts were large and the reversal rates providers experienced at hearing made appealing rational. Medicare Advantage growth and ordinary claim volume added more. Appeal receipts at OMHA grew several-fold in a few years while adjudication capacity grew hardly at all.

By early 2014, OMHA announced it would defer assigning most new provider appeals to judges — the appeals were docketed, acknowledged, and then waited, unassigned, in a queue. The pending count grew into the hundreds of thousands of appeals. Average processing times at level 3 stretched to roughly three years, against the 90-day statutory standard.

What the wait did to providers

The cruelest feature of the backlog was an interaction with the recoupment rules. Under the limitation-on-recoupment provision enacted in the Medicare Modernization Act (the "935 protections"), Medicare generally may not recoup an alleged overpayment while a redetermination or reconsideration is pending. But once the QIC decides against the provider, recoupment may proceed — even though the provider is still appealing. In a system where level 3 took ninety days, that was a tolerable design. In a system where level 3 took three years, it meant the money was taken years before any judge heard the case — with interest repaid if the provider eventually won, but with the cash gone in the meantime. For small providers under extrapolated overpayment demands (Chapter 37 §37.6), the wait itself was the penalty; some litigated the recoupment separately on due-process grounds, with mixed results (the Family Rehabilitation litigation in the Fifth Circuit is the best-known example).

Meanwhile the deadlines still bound the appellant. The provider's sixty days to reach the ALJ level never flexed. The system's ninety days to decide had, in practice, no consequence at all — escalation rights existed, but escalating past the backlog mostly delivered an appellant to a Council that was itself accumulating one. Every clock in the ladder kept perfect time except the ones that ran against the government.

How it was resolved

Three instruments, none of them fast.

Settlements, priced in cents on the dollar. In 2014, CMS offered hospitals a settlement of eligible pending inpatient-status appeals at 68 percent of the net allowed amount — a documented, public figure, and a remarkable one: the program's own price for not adjudicating. A later initiative, the 2018 Low Volume Appeals settlement, offered qualifying appellants 62 percent. Tens of thousands of appeals left the docket this way; the hospitals that accepted traded a likely win later for a certain payment now. (Both percentages are from CMS's published settlement terms; participation figures are public but not quoted here.)

Litigation and a court-ordered timetable. The American Hospital Association sued to compel HHS to meet its statutory deadlines. After the D.C. Circuit held in 2016 that mandamus could be appropriate, the district court in late 2018 ordered a schedule: reduce the backlog 19 percent by the end of fiscal 2019, 49 percent by 2020, 75 percent by 2021, and eliminate it by the end of fiscal 2022, with quarterly status reports to the court.

And money. Congress substantially increased OMHA's funding; OMHA hired judges and staff, expanded settlement-conference and statistical-sampling programs, and worked the docket down. HHS met the court's schedule, and by the end of fiscal 2022 the backlog, as defined in the litigation, was effectively eliminated. Processing times fell accordingly — a functioning level 3 again, for the first time in roughly a decade.

What it shows

First, the transferable sentence:

**An appeal right is a promise about process, not about time — and a right that takes three

years to exercise is, for many appellants, not a right.**

Second, capacity is part of system design. Every individual rule in §30.6 worked as written throughout the backlog years: windows, forms, levels, standards. The system failed anyway, because its designers had sized it for one volume of disagreement and its other components — the RAC program above all — manufactured another. Chapter 29 taught that a denial is data about the payer's logic; a backlog is data about the system's assumptions.

Third, asymmetric deadlines are a design choice with consequences. The appellant's deadlines were enforced by forfeiture; the adjudicator's were enforced by nothing until a federal judge supplied the consequence. When you read §30.10's insistence on tracking the payer's clock as well as your own, this is the argument at national scale.

Fourth, the settlement percentages are the honest price of delay. Sixty-eight cents on the dollar is what certainty was worth against a three-year wait — accepted by sophisticated institutions with strong cases. Delay does not merely postpone value; it converts winnable claims into discounted ones. That is recurring theme 6 — every day a claim sits, it is worth less — written into a federal settlement offer.

And fifth, the fix required an outside forcing function. Not better intentions — a court order with dates and a judge reading quarterly reports. Providers inside the system had spent years reporting the problem accurately to no effect, which should sound familiar: it is Chapter 29 §29.9's "when the answer is no," played at the scale of the federal government, and resolved the same way — a dated record of the problem, kept until somebody with authority had to act on it.

The lesson

For a working practice, the backlog years teach three durable habits.

Win at levels 1 and 2, where the clocks still work. The redetermination and reconsideration decided on time throughout the crisis. §30.7's front-loading discipline — complete argument at level 1, complete record at level 2 — is not just evidentiary hygiene; it is how you avoid needing the level that history has shown can fail.

Never build a revenue plan on money that is in an appeal queue. An appealed amount is not receivable on any schedule you control — Chapter 31 treats pending appeals accordingly, and the backlog is the proof.

And when an overpayment demand is in play, know the 935 recoupment timeline cold — because the interaction between recoupment and appeal pace, not the merits, may be the most consequential fact in the file. (Chapter 31 §31.9 and Chapter 37 §37.6 carry this forward.)

Discussion questions

1. The RAC program (Chapter 37) paid contractors contingently to find overpayments; the appeal system absorbed the results. Should the volume a review program generates be that program's design constraint? Who owned the seam between them?

2. Providers' reversal experience at the ALJ level was part of why appealing was rational at scale. If a review program's findings are frequently reversed on full review, what does that say about the findings — and what does it say about the incentives of the reviewer?

3. The 68 percent settlement was accepted by many hospitals with strong cases. Construct the decision from the hospital's side. What would have to be true for 68 cents now to beat a likely full win three years out?

4. The appellant's deadlines were enforced by forfeiture; the government's by nothing, for years. Design a consequence for missed adjudication deadlines that does not simply pay every appellant — what are the options, and what does each distort?

5. §30.10 tells a practice to track the payer's clock and escalate in writing when it lapses. At what point does the OMHA story suggest that individual escalation stops working — and what did work?

6. Compare this case with Chapter 29's Case Study 1. In both, people inside the system reported the truth for years without effect. What finally made each report actionable, and what is the common ingredient?