Not every patient who comes to the hospital needs to be admitted, but it can take time for a physician to decide whether they do or don’t. And so, many patients are put on observation status and kept in the hospital (generally up to 23 hours) either in an observation unit, in the emergency department (ED), or in an inpatient unit. 

Because there are various observation bed models, and because observation status is an important part of patient care, this topic inevitably becomes significant in the pre-design and programming phases of an architectural project.


What is observation status, anyway?
“Observation care is a well-defined set of specific, clinically appropriate services, which include ongoing short-term treatment, assessment, and reassessment before a decision can be made regarding whether patients will require further treatment as hospital inpatients, or if they are able to be discharged from the hospital,” according to the Centers for Medicare and Medicaid Services (CMS).

Under Medicare insurance, observation care is considered outpatient care. In some cases, an inpatient stay can be reclassified to outpatient observation during or after admission. CMS states that observation care, in most cases, should not last longer than 24 to 48 hours.

However, there are cases where observation status has exceeded that guideline. If care or monitoring is required beyond that timeframe, inpatient status is ideal. Under commercial insurance, 23 hours is typically the maximum time allowed for observation status. 

Some healthcare advocates argue that observation status is solely for billing purposes and has no impact on care delivered—that whatever is appropriate care at the moment would be provided whether on observation or inpatient status. Others argue that observation status puts patients in “care limbo” while hospitals determine what’s most economical.

According to Advisory Board Co., a global healthcare research, consulting, and technology firm, observation care typically costs the hospital less and may even increase revenue for the hospital—which is commonly misunderstood, as many people assume inpatient care generates more revenue. It’s also debated whether patients get a lower level of care in observation status, although in many cases, appropriate care may be a less-acute level of care. 

Regardless of the debates, outpatient observation status/care is a necessary process. So how should hospitals make care decisions during this transitional status? 


Models for observation care
Observation care can be provided in a consolidated or a distributed model—that is, in a dedicated unit or distributed among the available beds in other inpatient units. Both models have pros and cons. 

Consolidated models:

There are several consolidated observation models, with subtle differences between them. These models are commonly called observation units, clinical decision units, short-stay units, and chest pain units. They’re typically located adjacent to the ED as separate units, and sometimes they’re dedicated units within the inpatient unit floors.

Having the unit located next to the ED makes sense, as many observation patients have presented there. However, there may be concern with this location as the ED often needs these beds during peak times, therefore breaking down the observation management system. 

Conceptually, situating patients needing observation in these consolidated units helps free up ED and inpatient beds for incoming patients clearly admissible to inpatient care or in need of ED treatment.

Dedicated units also provide tighter-focused observation and, in theory, facilitate better decision-making on whether to admit or discharge a patient. Additionally, data from Advisory Board Co. suggests that dedicated units can have a positive impact on efficiency, with surveys showing decreases in length of stay by more than one day. 

Distributed model:

The distributed model is less organized in terms of location, with observation patients placed throughout the hospital. Patients in this model could be held in observation status in the ED or in an inpatient bed. While this may seem inappropriate compared with consolidated units, in some cases, it may be suitable for a hospital.

The need for inpatient beds has a tendency to change annually, even seasonally. Similarly, ED volume has predictable peaks day to day and hour to hour. Capacity and demand may not always align; therefore, inpatient or ED beds could be available as a resource for observation. 

As long as the infrastructure for providing distributed care is understood and managed accordingly, this could be a successful and economical model for managing observation within available resources.

But if not managed with discipline, this model could put observation patients out-of-sight and out-of-mind. And if ED and inpatient demand eventually exceed capacity, housing observation patients in the inpatient unit or ED could cause system compression and backlog. 

When deciding which location (ED or inpatient units) is a better place for patients in the distributed model, ED physicians often contend that keeping observation patients in or near the ED contributes to a shorter length of stay because of a more efficient delivery of observation care. 

Unique models:

There’s no one-size-fits-all system, and unique models of observation do exist. For instance, Seattle Children’s Hospital in Washington created an observation model called High-Intensity Ambulatory Care (HIAC) that consolidates prep, infusion care, and observation stations. HIAC rooms flex services depending on need.

This consolidated unit achieves efficiency in shared space and staffing. Due to healthcare reform, observation units adjacent to surgical services may become more prevalent in the future, focusing on post-operative patients suited for observation status. 


Understanding demand
HKS’s Clinical Solutions Group is often charged with analyzing ED and inpatient volumes, then translating the results into future bed projections for those departments. Reviewing inpatient days, ED volume, and the subsequent demand is typically straightforward, until the topic of observation patients is introduced.

These precise inpatient and ED projections often rendezvous with a rule of thumb for estimating observation stations. 

The generally accepted rule for determining observation admissions is to use 10% of a hospital’s admission volume as a basis. “Emergency Department Design: A Practical Guide to Planning for the Future” (2002) by Jon Huddy, AIA, NCARB, MArch, of FreemanWhite Inc., suggests two observation stations per 10,000 ED visits for ED observation. To inform those estimates and develop a model, we typically ask the following questions: 

Who clinically manages these patients? It’s important to understand who currently operates a client’s observation system, and how it’s managed. For example, how much do case managers and physician advisers collaborate in the designation decision? Who is determining status: ED physicians or hospitalists? Are there protocols already established for conditions that are related to observation status? Are the inclusion/exclusion criteria clear? 

On average, how many 23-hour/observation patients are treated in addition to the inpatient and ED volume? If this data is available, it illustrates the impact observation patients
are having on the hospital system—either in the ED or inpatient units, or both. 

What’s the average length of stay (LOS), mode LOS, and max LOS of observation patients? The mode explains the most typical length of stay, max shows the most extreme case, and the average helps us make room/bay need projections with the volumes. 

Are these patients in an inpatient unit or the ED? Knowing where observation patients are located currently begins the discussion of the pros and cons of this location. 

Does the hospital want them to be in the inpatient unit with this status? If so, the inpatient bed need will increase. 

Does the hospital want them in the ED? If so, the ED bed need will increase. 

If not in the inpatient units or ED, what consolidated model is preferred for managing observation patients? 

What is the process for admission, when needed? It’s important to know if the admission process melds well with the current or desired observation system. 

What is the outpatient volume for surgery, catheterization, infusion, etc.? There may be opportunity or desire to create a multipurpose unit that can benefit multiple patient types. 

Is there a high volume of medically unnecessary, one-day inpatient stays? The actual observation volume may be higher after reviewing inpatient length of stay more closely. 

Regardless of the model, outpatient observation status requires beds in addition to those needed for ED visits and inpatient stays. Most important to determine are the number of beds needed and the clinical responsibility of these beds.

This analysis will help formulate the best decisions on status and will keep other patient flows moving efficiently. Statistical simulation of volumes could help facilitate this decision.

Additionally, it’s important to weigh the perspectives of all concerned parties: the business office, the health information management office, physicians and nurses (from inpatient units and the ED), hospitalists, administration, and performance-improvement officers. 

While consolidated units are not the norm, they’re on the rise around the country (about 30% of hospitals have an observation unit of some kind, according to Advisory Board Co.). Data suggests that dedicated units reduce the length of stay for observation patients and free up beds elsewhere. Still, the “right” model depends on the volumes, capacity, and culture of each organization. 

Making a definitive decision for an organization regarding observation models leads to a better design solution during the planning of a project, and a more operationally sound way of taking care of patients on observation status.  

Ashley Dias, AIA, ACHE, LEED AP, EDAC, is an associate with HKS Inc. in Dallas. She can be reached at