I have written a lot in the past about the two-midnight rule exception for physician judgment of the need for inpatient admission with an expectation of a stay of under two midnights. As a reminder, when the two-midnight rule was first adopted, the only approved exceptions to the two-midnight expectation were unplanned mechanical ventilation and inpatient-only surgery. (Patients who have a rapid recovery, die, transfer, or leave against medical advice, etc., are not “exceptions;” they had an expectation of two midnights on admission, but their stay ended sooner than expected.)
But in response to intense lobbying, the Centers for Medicare & Medicaid Services (CMS) added an exception for physician judgment on Jan. 1, 2016. In short, if a physician determines that a patient who has an expectation of staying under two midnights warrants inpatient admission, then that admission will be considered appropriate for Part A payment. This exception actually makes no sense whatsoever, and CMS admitted that, stating that it disagreed with the need for the exception, but approved it nonetheless.
As the agency has stated, there are no services that can be provided to an inpatient that cannot be provided to an outpatient, so there is no patient (other than a patient undergoing inpatient-only surgery) who “requires” inpatient admission.
Why would a physician deem it necessary to admit a patient with an expectation of under two midnights? It’s really semantics. For years we have viewed inpatients as being sicker than observation patients. We have also gotten much more efficient at providing medical care and been better able, through the wonders of modern technology, to discharge patients much sooner than ever before. While a patient who suffered a heart attack used to spend days and weeks in the hospital, we now have “code cardiac” teams that can intervene in under an hour, stopping the heart attack in its tracks, allowing the patient to go home the next day. But the thought of treating an acute myocardial infarction, a life-threatening condition that kills thousands every year (many of them “dropping dead”) as an outpatient seems incomprehensible.
Likewise, from the hospital side, a patient with an acute myocardial infarction requires intensive, expensive resources that exceed those provided to a patient who has an elective cardiac catheterization and stent – yet if that acute myocardial infarction patient is treated as an outpatient, the hospital will receive an Ambulatory Payment Classification (APC) payment, which is several thousand dollars less than the corresponding Diagnosis-Related Group (DRG) payment.
When this exception was proposed, many providers contacted CMS for clarification, asking for case examples to help us understand how to apply it compliantly. CMS refused, instead referring inquiries to the Quality Improvement Organizations (QIOs), which would be doing the actual audits.
Initially, the QIOs would not supply examples, and in one instance they even referred questioners back to CMS. But that changed this month, when Dr. Laura G. Shawhughes, MD, a physician advisor to utilization review, care management, and CDI (clinical documentation improvement), posted on a user group a copy of the Livanta “Short Stay Review Program Overview and Update.” In this presentation, Livanta provided, in writing, several cases that did not have an expectation of a two-midnight stay but were approved for Part A inpatient payment. Their case summaries are as follows:
- A 75-year-old with CAD who presented with increasing chest pain. Hemodynamically stable, positive enzymes, no EKG changes. Admitted with NSTEMI, had cardiac catheterization with PCI on day of presentation. Patient was kept overnight, remained stable, and was discharged following a one-midnight stay.
- A patient who missed dialysis presented with CHF, K of 6.9, and ST segment elevations. The patient was admitted for monitoring and urgent dialysis, and discharged with normal K and better fluid balance after an overnight stay.
- Patient with DKA (pH 7.25, glucose 750) and altered mental status. The patient was admitted and given an insulin drip. She improved over a one-midnight stay and was discharged.
- An 84-year-old male presents post-fall three hours prior to arrival. The patient is on ASA and Plavix. The patient is alert and oriented without focal neurologic deficit. The CT of the brain shows acute subdural hematoma without midline shift. The patient was admitted for frequent neurochecks and repeat head CT in morning, to be discharged home if stable.
Livanta also listed several examples in the same section of approved one-day admissions, but these cases represent patients whose stays could rightly be expected to last over two midnights when the initial decision was made. Those cases that would fit into the unexpected rapid recovery category were:
- Acute abdominal pain requiring nasogastric decompression, q1-2 hr. VS, and frequent physician checks
- Diverticulitis with walled-off abscess and normal abdominal findings, no fever, no leukocytosis
- Acute electrolyte disturbances associated with symptoms:
- Hyperkalemia with EKG changes
- Hypercalcemia with seizures
- Hypocalcemia with tetany
- Hyponatremia with obtundation
Does a presentation from Livanta, a QIO, represent “official CMS guidance?” Of course not. Fortunately, Dr. Edward Hu, the president of the American College of Physician Advisors and system executive director of physician advisory services at the UNC Healthcare System, asked CMS about these examples on an Open Door Forum call. And while CMS could not provide an answer on the call, they went the extra mile and provided Dr. Hu a response via email. In a surprisingly straightforward answer, the CMS representative stated,” yes… it would be reasonable to look to Livanta’s education for examples, as some providers are (based on geographic location) may be interacting with both KEPRO and Livanta. And the QIOs are implementing the same Part A payment policy.”
In summary, these case examples demonstrate that it is appropriate to consider the use of this exception for patients whose life is in immediate jeopardy without nearly immediate intervention, and for whom treatment could result in a “cure” in fewer than two midnights. The examples provided by Livanta all fit that description; I also believe that patients with complete heart block requiring an emergent pacemaker and patients with severe anaphylactic shock would fit, and perhaps others. Noting that these are only my personal recommendations based on the above information, each hospital’s utilization review team should work with its compliance team and determine how to utilize this information, and to speculate how the Recovery Audit Contractors (RACs) will address this if and when they start auditing again.
With nearly a million denied cases awaiting review at the administrative law judge level, one could fault a hospital for choosing to not use this exception. And of course, no one would endorse expanding this exception to patients “at high risk” of a disease. But whatever is decided, we now have enough information for each hospital to make an informed decision.