Case Study 1 — The Two-Midnight Rule: When the Same Care Is Paid Two Different Ways

A real, public CMS rulemaking with a documented history. Tier 1 for the regulatory framework; qualitative rather than numeric where a figure would need verification.

Background

Chapter 3 §3.2 drew a line between Part A and Part B and said the line is not building-based: an inpatient admission is Part A, and an outpatient encounter at the same hospital is Part B.

That sounds clean. It is one of the most contested distinctions in American healthcare finance, because for a large category of patients — the ones who arrive at a hospital sick enough to stay but not obviously sick enough to admit — the classification is a judgment call, and the two answers produce radically different consequences for the hospital, for the physician, and above all for the patient.

A patient in observation status is an outpatient. The hospital bills Part B under the Outpatient Prospective Payment System (Chapter 34). The patient owes Part B cost sharing, which is coinsurance without an out-of-pocket maximum. Their self-administered drugs may not be covered at all. And — this is the consequence that generates the most anger — observation days do not count toward the three-day qualifying inpatient stay that Medicare has historically required before it will cover a subsequent skilled nursing facility admission.

A patient admitted as an inpatient for the identical clinical care is Part A. The hospital is paid a diagnosis-related group amount (Chapter 33). The patient owes the Part A deductible. And the days count toward the skilled nursing facility qualification.

Same bed. Same nurses. Same treatment. Same physician. Two entirely different bills, and one of them can cost a patient tens of thousands of dollars in subsequent nursing facility care that the other would have covered.

The issue

For years the classification rested on physician judgment guided by loose criteria and commercial screening tools, and it produced two opposite pressures.

Hospitals had a financial incentive to admit — inpatient payment generally exceeded outpatient payment for comparable care — and audit contractors, particularly the Recovery Audit Contractors (Chapter 37 §37.5), aggressively reviewed short inpatient stays and denied them as medically unnecessary at the inpatient level of care. The denials were retrospective, they were made by reviewers reading the record months later, and a hospital whose one-day inpatient stay was denied frequently could no longer rebill the encounter as an outpatient claim, because the timely filing window had closed. The hospital was paid nothing at all for care that had certainly been furnished.

So hospitals responded by classifying more patients as observation, which shifted cost to beneficiaries and, in the aggregate, produced a well-documented growth in observation stays and a corresponding rise in patients who discovered — at the point of discharge to a nursing facility — that they had never been admitted.

CMS addressed it in the fiscal year 2014 Inpatient Prospective Payment System final rule with what became known as the two-midnight rule: a benchmark under which a stay was generally appropriate for Part A inpatient payment if the admitting physician expected the beneficiary to require medically necessary hospital care spanning at least two midnights, and documented that expectation.

It was intended to replace an unpredictable judgment with a bright line.

What happened

The rule was immediately contested from every direction.

Hospitals objected that the benchmark substituted a clock for clinical judgment and did not fit patients who were genuinely sick but efficiently treated. The rule was accompanied by a payment reduction to offset an expected increase in inpatient admissions, and that reduction was challenged in litigation, with the courts ultimately requiring CMS to better justify it. Enforcement of the rule was subject to a probe-and-educate period rather than immediate audit, and the audit responsibility for short stays was reassigned between contractor types more than once.

CMS subsequently revised the framework to restore an explicit case-by-case exception: a stay of under two midnights could still be appropriate for Part A payment where the documentation supported the physician's judgment that inpatient care was required. That returned some discretion — and, with it, some unpredictability.

And the patient-facing problem persisted independently. Because observation status was invisible to patients, Congress enacted the Notice of Observation Treatment and Implication for Care Eligibility Act, requiring hospitals to give a written and oral notice — the Medicare Outpatient Observation Notice, Form CMS-10611 — to beneficiaries receiving observation services beyond a specified number of hours, explaining that they are outpatients and what that means for their cost sharing and their skilled nursing facility eligibility.

The right to appeal the status classification itself has been the subject of extended litigation, with courts addressing whether beneficiaries reclassified from inpatient to observation are entitled to an administrative appeal.

What it shows

First, it demonstrates that the Part A / Part B line is a decision, not an observation. Chapter 3 presented the four parts as a clean taxonomy. This case study is the disclaimer: for a substantial category of encounters, which part applies is determined by a documented physician expectation, and that determination is reviewable, contested, and worth a great deal of money to several parties at once.

Second, it shows what a payment rule does to clinical documentation. Once payment turned on a documented expectation of a two-midnight stay, the documentation of that expectation became the object of enormous attention — from utilization review staff, from clinical documentation integrity programs, and from auditors. This is the single clearest illustration in Part I of the book's first theme. The care did not change. What the record had to say about the care changed completely, and an entire professional function grew up around making sure it said it. Chapter 38 is about that function.

Third, it is the sharpest available example of a rule whose burden falls on someone who has no part in the decision. The hospital and the physician argue about status. The auditor reviews it. The beneficiary — who cannot see the classification, was not consulted, and experiences no clinical difference — bears the consequence at the nursing facility three weeks later. That asymmetry is what produced the notice statute, and the notice statute is a fix for the visibility of the problem rather than for the problem.

Fourth, and most useful for a working professional: it shows why the appeal timeline and the rebilling window matter enormously. The hospitals' original complaint was not primarily that inpatient stays were denied. It was that they were denied after the window to bill the encounter correctly as an outpatient claim had closed — so a defensible outpatient claim became zero. CMS later provided rebilling pathways with their own conditions and timeframes. Chapter 1 §1.3's two clocks, Chapter 30's appeal deadlines, and Chapter 31's timely-filing discipline are all downstream of this kind of failure.

The lesson

A status determination is a coding-adjacent decision with the financial weight of a large procedure, and it is made by clinicians under time pressure who are not thinking about claim forms.

Three concrete carry-forwards:

Know which side of the Part A / Part B line an encounter is on before you do anything else with it. It determines the form, the payment system, the patient's cost sharing, and the appeal path. Chapter 26 and Chapter 34 both assume you know.

When a rule turns on a documented expectation, the documentation of that expectation becomes the claim. Auditors do not review the decision; they review the record of the decision. A physician who made an entirely correct judgment and did not write down the reasoning has, for payment purposes, not made it.

And watch the rebilling window whenever a claim is at risk of reclassification. A denial you can rebill correctly is an inconvenience. A denial you can no longer rebill is a total loss, and the difference between them is a date.

The two-midnight framework, the exceptions, the audit contractor assignments, the notice requirements, and the skilled nursing facility qualifying-stay policy have all been revised, litigated, or waived in specific circumstances. Verify the current state of each with CMS before relying on any description of it, including this one.

Discussion questions

1. §3.2 said the Part A / Part B distinction is not building-based. After this case study, restate the actual basis of the distinction in one sentence, and say what makes it hard to apply.

2. The two-midnight rule replaced a clinical judgment with a benchmark and was then modified to restore a case-by-case exception. Argue both sides: what does a bright line buy, and what does it cost?

3. The beneficiary bears the largest consequence and has the least visibility. The statutory response was a required notice. Is a notice an adequate remedy? What would a better one look like, and who would pay for it?

4. The chapter said Medicare's rules become everyone's rules. Commercial payers also make inpatient-versus-observation determinations. Name two ways a commercial payer's version of this problem is worse for a hospital than Medicare's, using §3.1's argument about publication.

5. A hospital's utilization review nurse asks you, a coder, whether a particular stay "should be inpatient." What is the correct answer, and what is the correct thing to do next? (There is a right answer and it involves a boundary from Chapter 1 §1.6.)