The CMS proposal to eliminate the Inpatient-Only List, does not mean that all procedures are to be performed as outpatient. On Jan. 18, 2018, RACmonitor published my article titled “CMS Says OK to Admit Total Knee Replacements as Inpatient.” The title created a bit of an uproar, as many thought I was conveying the message that all total knee arthroplasties could be admitted as inpatient.
Of course, the article itself explained the true message, that patients could be admitted as inpatient if appropriate. It appears that history may repeat itself, as the Centers for Medicare & Medicaid Services (CMS) has finalized its proposal to eliminate the Inpatient-Only List, but once again while stressing that elimination of the list does not mean that all procedures are to be performed as outpatient. And once again, my message is that one-day inpatient admissions are allowed.
This impending demise of the Inpatient-Only List, announced in the Medicare 2021 Outpatient Prospective Payment System (OPPS) Final Rule, has created quite the controversy. It is clear that as surgeries are removed from the list, the Two-Midnight Rule will determine the status, but that’s the end of clarity. Ask three people what that means, and you will probably get four interpretations.
The Two-Midnight Rule can be broken down into three main categories. After seven years, most providers are comfortable with the concept of a two-midnight expectation. Starting with the initiation of care, if the patient is expected to require at least two midnights of necessary hospital care, exclusive of delays for convenience, then inpatient admission is warranted. This would clearly be the case with many surgeries performed in the hospital, such as coronary artery bypass surgery, colon resection, organ transplant, thoracotomy, and so on. CMS has also stated that admission is appropriate for patients whose recovery will require post-acute care in a skilled setting where a preceding three-day inpatient admission is required. The patient having hip arthroplasty who lives alone in a multi-level home with no support system can be admitted as inpatient to allow them to continue their rehabilitation in a Skilled Nursing Facility (SNF).
But this two-midnight expectation feature of the rule is less clear for many surgeries in which the in-hospital length of stay can vary between one and three days, such as joint arthroplasty and many orthopedic and spine procedures. Of course, CMS chose those surgeries to remove from the Inpatient-Only List as of Jan. 1, so this is no longer a theoretical discussion. But in-hospital expected length of stay is often a bit of a moving target. If a hospital has two surgeons who perform joint replacement, and one normally sends patients home the day after surgery, but another always keeps their patients for two days, does the second surgeon always admit theirs as inpatient, while the first cannot? In other words, does the two-midnight expectation depend solely on the physician’s personal practice pattern, or does the prevailing community and national standards of care have an influence?
I think it is fair to exclude the surgeon in Chicago who is known as the “orthopedist to the stars,” who is able to discharge 80 percent of his patients, including those over 65 years of age, on the same day of surgery – but that is because he has them spend the next day or two at the Peninsula or Ritz-Carlton in Chicago near his office, and he arranges for them to get in-hotel visits by his assistants. On the other hand, with the COVID-19 pandemic, we have seen more patients discharged the same day as surgery to avoid hospitalization, so that may influence the length-of-stay expectation sooner than we thought. If more doctors are able to discharge these patients the same day now, why would that practice stop simply because the public health emergency has passed?
Likewise, there is little controversy associated with the notion that if a patient has a surgery as an outpatient and a second midnight of hospital care is necessary, not including any time spent due to convenience or preventable delays, then inpatient admission is warranted. While the question of what constitutes medically necessary hospital care after a surgery is not completely settled (such as how much pain, what level of immobility, or how much nausea warrants continued hospital care), the bigger issue is ensuring that there is documentation of that need, so the second midnight and inpatient admission can be supported.
The majority of the confusion, though, began in the 2016 OPPS Final Rule, when CMS added the third component, the case-by-case exception, to the Two-Midnight Rule. Prior to this, the only exception to the rule was unplanned mechanical ventilation, meaning a patient who required mechanical ventilation but also had a one-midnight expectation could be admitted as inpatient. This new case-by-case exception allowed a physician to make a determination that for a particular patient, based on the beneficiary’s history and comorbidities, severity of signs and symptoms, current medical needs, and risk of an adverse event, inpatient admission was warranted despite an expected hospital stay of only one midnight.
It is critically important to note at this point that these exceptions are categorized by CMS under the “rare and unusual” exception category, but in the 2016 OPPS rule, CMS clearly stated that “we would like to clarify that our proposed modification to the current exceptions process does not define inpatient hospital admissions with expected lengths of stay less than two midnights as rare and unusual. Rather, it modifies our current ‘rare and unusual’ exceptions policy to allow Medicare Part A payment on a case-by-case basis for inpatient admissions that do not satisfy the two-midnight benchmark.” (I would like to thank Dr. Ralph Wuebker for asking CMS this crucial question back in 2015.)
This means that patients with a one-midnight expectation, but who are determined to warrant inpatient admission, can be admitted as inpatients as frequently as warranted. For some surgeries and surgeons, that may be infrequent, if they perform relatively low-risk surgeries on low-risk patients, but for others who are willing to perform surgery on higher-risk patients, the majority of their surgeries may qualify for inpatient admission. This is important in that while the Program for Evaluating Payment Patterns Electronic Report (PEPPER) may report the percentage of one-day inpatient surgical admissions and allow you to compare your rate to others, there is no benchmark rate, and your rate may be perfectly appropriate, even if you are in the top percentile.
But does a physician calling a patient high risk mean they are high risk? When the two-midnight rule was first introduced, many set up a medical record “smart phrase” that read, “I expect two midnights,” and instructed physicians to use it to justify their inpatient admission decision. But it quickly became clear that those four words carried no special powers, and the patient’s illness and needs must be such that a two-midnight expectation actually made clinical sense. Documenting “chest pain; I expect two midnights” did not de facto mean that inpatient admission was correct. The same applies to the patient with a one-midnight expectation who is being admitted based on risk. CMS even codified this in 42 CFR § 412.3 (d)(3), stating, “the factors that lead to the decision to admit the patient as an inpatient must be supported by the medical record in order to be granted consideration.”
Evidence to support the use of the case-by-case exception can also be found in MLN Matters SE19002, issued Jan. 24, 2019, titled, “Total Knee Arthroplasty (TKA) Removal from the Medicare Inpatient-Only (IPO) List and Application of the Two-Midnight Rule.” In this document, which is clearly specific to total knee arthroplasty, for a surgery always performed electively on stable patients, CMS repeatedly makes it clear that this exception can be used.
As more surgeries are removed from the Inpatient-Only List and advances in surgical techniques, anesthesia, and pain management lead to shorter lengths of stay, there will be more need to determine which patients have a higher risk, and which surgeries are intrinsically high-risk, warranting inpatient admission under this case-by-case exception. As an example, the transcutaneous aortic valve replacement (TAVR) procedure is often performed with discharge anticipated the next day. Every single patient having a TAVR has severe aortic stenosis, and the procedure itself involves inserting large catheters into the heart and then placing an artificial valve. When this procedure is no longer inpatient only, would it not be reasonable to consider every patient as high risk due to their life-threatening valve disease, and every procedure as high risk due to its invasive nature, with catastrophic consequences if things do not go well – and therefore perform all TAVRs as inpatient? Would the same not also apply to carotid artery procedures, or use of a left ventricular assist device for cardiac stenting?
What determines if the risk is high enough to warrant admission? Is this not a judgment best left to the physician who is caring for the patient, as long as that decision is clinically valid? Going back to total knee arthroplasty, CMS made it clear that they expected the majority of such procedures to continue to be performed as inpatient, yet no surgeon would perform such a surgery on an unstable patient. That means that risk must be assessed relatively, and not simply based on whether the patient has a chance of dying on the operating room table or is likely to decompensate post-operatively. A recent study demonstrated that diabetic patients having joint arthroplasty have a 450-percent higher risk of in-hospital mortality than non-diabetics. That’s a significantly increased risk, even if the possibility goes from 0.02 to 0.09 percent.
The perioperative risk for a patient should be judged relative to the average-risk Medicare patient. The risk for a 65-year-old patient who takes one blood pressure pill having spinal fusion is going to be less than that of a 65-year-old who takes two or three medications. And their risk is lower than that of a 90-year-old, even one who takes no medications at all. CMS has made it clear that they will not set guidelines, as they “believe that the surgeons, clinical staff, and medical specialty societies who perform (the procedures) and possess specialized clinical knowledge and experience are most suited to create such guidelines.” Some have looked to the American Society of Anesthesiology classification scale as a guide, but this has almost no utility for risk stratification for status determinations, with a wide gulf between an ASA II patient, who has a mild systemic disease, and an ASA III, who has severe systemic disease. A diabetic patient could be classified an ASA II or ASA III, yet, as the study above shows, all diabetics are at higher risk. Where would one classify a patient with chronic lung disease who uses a daily inhaled corticosteroid, or the patient with an elevated body mass index? What about the patient with several “mild” systemic diseases?
One could also ask if a higher-risk patient requires any special care to justify being deemed “higher risk.” CMS sets no special requirements, but common medical practice should dictate what extra care, if any, is provided. And that care may be intrinsic to the patient’s normal peri-operative care. For example, when administering anesthesia to a patient with hypertension, even if that patient is classified as ASA II, the anesthesiologist is aware of the diagnosis and will be closely monitoring the blood pressure. If the patient takes a beta blocker for the hypertension, the heart rate will also be watched more closely. Likewise, a patient with sleep apnea will warrant not only careful monitoring during surgery, but the nursing staff caring for that patient will know that this patient’s respiratory status must be closely monitored, especially if opioid pain medications are given. Does there need to be an order for “close monitoring of respiratory status?” There does not, because that is part of the complex assessment and care planning by the nurse. Yet that should warrant inpatient admission, because despite the extra monitoring, the patient is at higher risk of respiratory depression and respiratory failure.
It is important to note that CMS allows physicians to base their decision on the patient’s risk, not on whether the risk has been realized, a standard adopted by other payors. In other words, the Medicare patient with sleep apnea is at higher risk of perioperative complications – and, in my opinion, likely warrants inpatient admission for a surgery in which the expected length of stay is one midnight. But a commercial or Medicare Advantage payor may not acknowledge this risk, and allow inpatient admission until the patient’s risk has been realized – once the patient has post-operative respiratory failure. It must be remembered that CMS has no jurisdiction over the status and payment to providers for Medicare Advantage patients if the providers are in-network with the plan, so the admission status can vary for two patients with the same risk, but different payors.
It is also worth noting that this exception would only apply to patients who are expected to stay one midnight, and CMS does not expect it to be used for patients who are not staying past one midnight. In the 2016 OPPS Final Rule, CMS said “we would expect it to be rare and unusual for a beneficiary to require inpatient hospital admission after having a minor surgical procedure or other treatment in the hospital that is expected to keep him or her in the hospital for only a few hours and not at least overnight.” If surgical techniques improve to the point that patients undergoing TAVR could be discharged the same day, it will need to be clarified if the intrinsic high-risk nature of the procedure would warrant inpatient admission.
While the case-by-case exception became a hot topic when total knee arthroplasty was removed from the Inpatient-Only List in 2018, on an Open Door Forum call, CMS noted that the exception was adopted in 2016, and they never specified that it could not be used for other surgeries with a one-day expected length of stay, such as hysterectomy, prostatectomy, or cholecystectomy. These surgeries are not now on the Inpatient-Only List, and as such, are not restricted from status audits, so providers would be wise to use the exception here thoughtfully.
When total knee arthroplasty was removed from the Inpatient-Only List and the Beneficiary and Family-Centered Care Quality Improvement Organizations (BFCC-QIOs) audited some of these one-day admissions as part of their short-stay probe-and-educate audits, many providers noted that the reviewers were denying these stays, stating that the patient did not stay two midnights, so the Two-Midnight Rule was violated. A prominent healthcare lawyer also noted that many of their clients faced the same issue, leading them to recommend that providers view all elective knee replacements (and later hip replacements, and soon, all orthopedic and spine procedures) with a one-day expectation, as surgeries to be performed as outpatient and the patient only admitted as inpatient if they remained in the hospital past the second midnight for necessary care.
This is not the intent of CMS, and in my personal opinion, misinterpretation of the rule should not guide behavior. If the speed limit is 35 miles per hour and you are driving 34 miles per hour on a clear, sunny day, and you get a ticket for driving over 25 miles per hour because the police officer decided that was the speed limit they wanted to enforce that day, you should not pay the ticket, nor should you start driving no faster than 25 miles per hour on that road. You should dispute the ticket, appear at court with evidence supporting that the speed limit was 35 miles an hour and that the weather that day did not warrant lowering the speed limit due to safety concerns, and argue for dismissal.
CMS established the exception for one-midnight stays in which the physician determined that inpatient admission is warranted, and codified it in the Code of Federal Regulations. In the 2021 OPPS Final Rule, they also commented that they do not expect significant financial implications for hospitals from the removal of the nearly 300 surgeries from the Inpatient-Only List, many of which have an expected length of stay of one day. I take them at their word. The exception is there, and should be used when it is appropriate and not in “rare and unusual circumstances,” with the documentation to support the decision. That’s the way laws are supposed to work.
Ronald Hirsch, MD, FACP, CHCQM-PHYADV, CHRI, FABQAURP is vice president of the Regulations and Education Group at R1 Physician Advisory Services. Dr. Hirsch’s career in medicine includes many clinical leadership roles at healthcare organizations ranging from acute-care hospitals and home health agencies to long-term care facilities and group medical practices. In addition to serving as a medical director of case management and medical necessity reviewer throughout his career, Dr. Hirsch has delivered numerous peer lectures on case management best practices and is a published author on the topic.