Inpatient versus observation status: the Medicare dilemma

The review process is flawed, and physicians should take action, three experts say.

A physician must consider a variety of factors when deciding whether to admit a patient to the hospital. Every hospital has its own admission protocols and criteria, and case managers and utilization review committees use proprietary screening tools, such as McKesson's InterQual or Milliman.

All of these are part of the decision-making process, and that decision between admission and observation will directly impact the level of care provided, the hospital's reimbursement, and, in many cases, the post-hospitalization course of the patient. However, the contractors at the Centers for Medicare and Medicaid Services (CMS) are beholden to none of these factors. The ambiguity in CMS regulations, and the inconsistencies in how these regulations have been applied, have created an apparent battle involving Medicare, its contractors, physicians and hospitals regarding inpatient versus observation (outpatient) services.

Photo by Thinkstock
Photo by Thinkstock.

CMS has been aggressively reviewing and auditing hospital admissions (primarily for one- to two-day stays) in an attempt to reduce the rate of improper payments to hospitals and physicians. At this time, the agency's Recovery Audit Contractors (RAC), Medicare Administrative Contractors (MAC), ZPIC (Zone Program Integrity Contractors), and Comprehensive Error Rate Testing (CERT) employees are all reviewing hospitals' inpatient admissions to determine if decisions to admit were correct. Unfortunately, there are some fundamental flaws in this review process.

The error rate

The CERT program was constructed by CMS to measure and correct the incidence of improper payments. An error rate (estimating how often patients are admitted when they should be observed) was calculated by the CERT contractor. For fiscal year 2010, the CERT calculated a 9.5% Part A acute inpatient error rate. The CERT's annual error rate serves as the impetus for advanced audit scrutiny, prepayment probes, provider education and more. It is not an idle number, and it is time that physicians and hospitals begin to question the methodology used to establish the error rate.

In an effort to reduce this error rate, multiple CMS contractors have proceeded to audit and review hospital admissions. MACs were instructed to lower the error rate by denying payment for improper Medicare Part A or B services.

Of the roughly 80,000 claims reviewed by the CERT in fiscal year 2010 (the most recent year with fully available statistics), only 2,400 involved inpatient admissions. In choosing the 2,400 admissions, administrators did not consider the geographic location of the hospital or the demographics of the patient population. This behavior unfairly subjects a small hospital located in a rural and underserved area to provide the same resources as a hospital in a major metropolitan area. Not all hospitals are able to develop specialty observation units, and hospital policy might dictate an inpatient admission for some patients who would be placed in dedicated outpatient units in other areas of the country.

Also, there are close to 1,000 Diagnosis Related Groups (DRGs), so a sampling of only 2,400 hospital inpatient cases, or 3% of all claims subject to the CERT review program, cannot possibly address the nuances of each classification across multiple geographic and demographic boundaries. Medicare's DRG payment system was created to adequately reimburse hospitals for inpatient care. The calculation of the DRG takes into account the weight of a given diagnosis based on the expected resources to be provided as well as the expected length of stay. It also takes into account the comorbidities of the beneficiary, complications, procedures, geographic location of the hospital, demographics of the patient population and whether the hospital was a training center for new physicians. In contrast, the outpatient (observation) payment pays only for the itemized services and is adjusted on only very limited factors.

Little guidance

Each MAC has taken a turn at developing language to define what constitutes an appropriate inpatient admission. Three MACs have crafted Local Coverage Determinations to clarify their position on inpatient admissions versus outpatient (observation) services; two MACs created decision trees to visually represent the issue; two MACs have only required adequate documentation and have not spoken to the medical necessity requirements; and, several MACs have excerpted portions from the Medicare Benefit Policy Manual in describing their rationale for inpatient admission, avoiding more specific guidance.

The Medicare Benefit Policy Manual provides vague guidance to an admitting physician regarding the factors to consider when making an admission decision and leaves quite a bit open to interpretation by a physician or a reviewing Medicare contractor. These guiding factors include the patient's signs and symptoms, the predictability of an adverse outcome and an expectation for 24 hours of care.

For physicians, these factors have to be weighed in the moment, and not three years later, which is often when the admissions are reviewed by the contractor. Although physicians have been granted the responsibility of determining which patients require an inpatient admission, their decision does not carry much weight in the eyes of the Medicare contractor. As part of the Medicare appeals process, the doctor's decision is being second-guessed, often by nurses who do not have admitting privileges themselves, but do have the luxury of knowing that the admission had a favorable outcome.

Medicare contractors make blanket statements that their reviews are not retrospective, but with the benefit of hindsight and the knowledge of a favorable outcome, this is clearly not the case.

Inpatient = outpatient?

When it denies a hospital the DRG payment for inpatient services, CMS is simply not covering the cost of care the hospital provided.

At least two of the MACs have claimed that inpatient and outpatient services are the same and therefore should be paid for at a lower outpatient rate. However, CMS stated in the Federal Register (Nov. 27, 2007) that inpatient and outpatient services are not the same level of care, and they are not considered the same by hospitals either. The denial of an admission does not lower the cost of health care; it only shifts the payment burden to another party or fails to adequately reimburse the provider. When the hospital is only reimbursed at an outpatient rate, and expected to provide the same care to the beneficiary as an inpatient, it does not cost any less for the hospital.

As a result, hospitals are going to have to shift more of the cost of care to beneficiaries. As an outpatient, a beneficiary not only has to pay a higher co-pay, but also has to pay for all self-administered and chronic medications administered during the hospitalization. Additionally, when classified as an outpatient instead of an admitted patient, the beneficiary may also be denied Medicare coverage for follow-up care, like that provided in a skilled nursing facility.

The consequences

While hospitals and clinicians struggle with the complex decision to admit patients, observation claims have increased by more than 46% since 2006 and observation stays longer than 24 hours have increased by more than 300% from 2006 to 2010. The reason for the increase is obvious: it is a direct result of more audit scrutiny, ambiguous CMS criteria, and the misconception that inpatient admission and observation services are identical.

Physicians need to become more involved and informed in the process of CMS reviews and denials. They must learn to document in the medical record the information to support an inpatient admission if one is necessary. Medicare contractors also should involve a physician at every level of review, and should use transparent and agreed-upon evidence-based guidelines. The trend to eliminate short-stay admissions can affect how hospitals provide care and are reimbursed for services rendered. Most of all, these changes directly impact, and may be harmful to, patients.