The Relationship between GMLOS and Revenue Leakage

Many healthcare facilities have implemented processes that require clinical documentation integrity (CDI) staff to review all patients as quickly as possible to establish a working Medicare Severity Diagnosis Related Group (MS-DRG) so the associated geometric mean length of stay (GMLOS) can be used to support discharge planning and improve patient throughput. While some hospitals continue this process, other hospitals are implementing technology to attain the desired outcome.

An article by Becker’s Hospital Review centers on the success Baptist Health has had using predictive tools to identify patients likely to discharge the next day. The article reports that “the system saw a 34-percent reduction in geometric mean length of stay (GMLOS) variance.” Many hospitals are working to decrease their average length of stay (LOS) as they better understand the relationship between hospital reimbursement under the MS-DRG methodology and GMLOS. Implementing technology to better predict and organize patient discharges could be a fantastic way to promote clinical efficiency that bridges the gap between clinical practice and the clinical revenue cycle.

Monitoring an organization’s GMLOS is a more sound and contextual approach than focusing on the average LOS, but mastering this approach requires an understanding of MS-DRG methodology. My goal with this article is to explain some of the nuances associated with hospital reimbursement and GMLOS.

Even though hospitals may be using GMLOS as a tool to help healthcare teams know if they are on target with discharge planning for one particular patient, compared to other patients expected to use similar hospital resources, it is a statistical average. There will be some variation among patients within the same MS-DRG. GMLOS is most powerful as an indicator of revenue leakage when viewed holistically within a MS-DRG population, rather than applied to individual cases.

I have always been a proponent of comparing a patient’s actual LOS to the GMLOS associated with the working MS-DRG to help determine the most appropriate timing to conduct follow-up CDI reviews when it is not determined by other factors like a symptom principal diagnosis, outstanding diagnostic result, or surgery. If a patient is hospitalized longer than the GMLOS associated with the working MS-DRG, it can be indictive of a documentation issue or a discharge issue. This is where it is helpful to have a collaborative process between CDI professionals and case managers. The premise for this approach is that if the patient’s acuity is accurately captured by the working MS-DRG, there should be alignment between the actual LOS and GMLOS.

Each MS-DRG has an associated GMLOS, a relative weight used to determine reimbursement, and an arithmetic length of stay (ALOS). The GMLOS removes outliers from the arithmetic or average LOS calculated from all claims with the same MS-DRG each fiscal year, which is why it is more reliable than the ALOS.

In general, a hospital technically loses money when the patient’s actual LOS is longer than the GMLOS associated with the working (or billed) MS-DRG. A decrease in a hospital’s GMLOS may:

• Improve patient throughput;
• Increase patient satisfaction; and
• Reduce revenue leakage.

Although the above statement is correct, the nuance is more complex, because the relationship between GMLOS and relative weight is not the same for all MS-DRGs. Additionally, as GMLOS decreases, there is less wiggle room for early discharge, especially because the median GMLOS across all MS-DRGs is four days.

The reason revenue leakage is associated with an LOS that exceeds the GMLOS is that room-and-board costs, which include nursing costs, are one of the most expensive resources used during an inpatient admission.

Extended stays beyond the associated GMLOS not only contribute to revenue leakage for that particular claim, but also lost revenue from an operational standpoint in a facility at or beyond bed capacity. If the bed is operating at a loss, it is not only losing revenue from its current occupancy, but it is also losing revenue from loss of the potential of being occupied by a patient early in their admission, where the opportunity of a profit is still possible. In other words, instead of generating revenue, the bed is losing revenue. Every hospital has a limited number of beds, so the more beds filled with patients where reimbursement is below the actual cost of services, the greater likelihood of revenue loss.

The first couple of days of hospital care, regardless of patient status, are the most expensive, because this is when the most orders are placed and treatment is being initiated. As the length of stay increases, the primary cost becomes room and board, because fewer additional resources are required to diagnose and treat the patient. In fact, when a patient is transferred as an inpatient between hospitals, the payment is not equally divided between facilities, even if the patient spends the same number of days at each facility.

But as average LOS decreases and GMLOS follows during annual MS-DRG updates, GMLOS does not have the same direct impact on reimbursement as it once did. To better understand this concept, I will compare a couple of MS-DRGs:

• MS-DRG 885 Psychosis has a relative weight of 1.3968 and GMLOS of 6.5 days. A patient admitted for a psychiatric diagnosis is less likely to require diagnostic tests and expensive medications like antibiotics; therefore, the biggest resource they require from the hospital is room and board. If the actual LOS is less than the GMLOS, it should result in a profit.
• MS-DRG 083 Traumatic stupor and coma > 1 hour with CC has a relative weight of 1.3958 and a GMLOS of 3.2. This patient is more likely to need a head CT and other medical services in addition to increased nursing care, hence a similar relative weight with a shorter GMLOS. The costs associated with this admission are more equally distributed across a variety of resources.
• MS-DRG 358 Other digestive system OR procedure without a CC/MCC has a relative weight of 1.3979 with a GMLOS of 2.5 (an even shorter GMLOS because the cost is primarily associated with the surgery).

As a reminder, GMLOS is reported using fractional days, but Medicare counts inpatient admission from midnight to midnight. Day one is the day the patient begins receiving hospital services, regardless of the time of day. Any part of a day before midnight counts as one day. Once midnight is crossed, the patient is on hospital day two, even if care is initiated at 11:58 at night. It is also important to note that Medicare does not include the day of discharge when counting Medicare inpatient days.

Another factor that can affect profitability is the type of patient bed. Because nursing care is included in room-and-board costs, patients who require more intensive nursing services are placed in step-down units or intensive care units (ICUs). MS-DRGs are not impacted by patient location within a hospital. Technically, the same MS-DRG rate is paid whether the patient requires an ICU bed or a general medical/surgical bed. However, patients with a diagnosis classified as a major complication/comorbidity (MCC), which results in a higher-weighted MS-DRG and associated GMLOS, are more likely to require more intensive nursing care.

Tools that can guide appropriate and timely discharge can improve hospital throughput, which can in turn minimize revenue leakage associated with operational inefficiencies. It can also allow hospitals to treat more patients, which can also increase the likelihood of profitability. This is just another example of why hospitals need a coordinated clinical revenue cycle approach to address financial challenges.