Inpatient vs. Observation: Where Level-of-Care Denials Turn

Observation-status downgrades — where a payer retrospectively reclassifies an inpatient admission as outpatient observation — are one of the most common payer tactics against community hospitals and one of the most consistently under-argued denial categories. The reason they are under-argued is that most appeals make a specific analytical mistake: they argue the wrong question.

The wrong question

"The patient was sick enough to be admitted" is not the argument. The payer already knows the patient was sick enough to present to the hospital. The denial is not disputing that. The denial is disputing whether the admitting physician had a reasonable expectation that the patient would require inpatient care spanning at least two midnights — at the time of the admission decision, based on the clinical picture as it existed at that moment.

This is a prospective standard, not a retrospective one. It does not ask whether the patient ultimately stayed two midnights. It asks whether it was reasonable to expect that they would, given what the physician knew when the admission order was written. Appeals that describe the patient's hospital course — what happened during the stay — are answering a retrospective question that nobody asked.

What the Two-Midnight Rule actually requires

The physician's expectation at admission time must be documented and clinically reasonable. "Clinically reasonable" means: given the presenting complaint, vital signs, lab values, imaging, comorbidities, and clinical trajectory at the time of admission, a prudent physician would have expected the patient to require hospital-level services for a period that would span at least two midnights.

The appeal must reconstruct the clinical picture as of the admission decision — not as of discharge. It must explain what the physician saw, what was uncertain, what risks were present, and why the expectation of a two-midnight stay was reasonable given those specific facts at that specific moment. (See a full observation defense memo →) Retroactively pointing out that the patient turned out to stay only 36 hours is the payer's argument, not the provider's.

Where the documentation usually fails

The admitting physician's order often says something like "Admit to inpatient — pneumonia." That is a disposition, not a clinical justification. It does not explain why the physician expected the patient to need inpatient care spanning two midnights. Without that documented reasoning, the appeal must reconstruct the clinical picture from the chart — which is possible but harder than having the prospective justification documented at admission time.

The strongest defense against observation downgrades is prospective: a physician attestation at admission that states the clinical basis for the inpatient expectation, documented before the stay is complete. When that doesn't exist and the appeal must be constructed retrospectively, the argument shifts to reconstructing what the physician reasonably would have expected based on the documented clinical picture at admission.

Why this requires physician-level reasoning

The observation-versus-inpatient dispute is fundamentally a clinical-judgment question: was the admission decision reasonable given the information available at the time? Answering that requires a physician who can evaluate the clinical picture, assess the uncertainty, and articulate why the admitting physician's judgment was clinically sound — even if the patient's course turned out to be shorter than expected. A coder can identify that the admission was downgraded. Constructing the clinical-judgment argument for why the admission expectation was reasonable is a physician function.

The pattern that makes this preventable

Observation downgrades are unusually systematic. Certain payers apply them to specific clinical presentations — short-stay chest pain, short-stay abdominal pain, single-midnight pneumonia — as a routine post-payment strategy. That means the denial pattern is predictable, the clinical arguments are repeatable, and a structured escalation approach can address them at portfolio scale rather than case by case.