Key Considerations in Developing an Observation Unit

By Bartho Caponi, MD 

The recent ACPA Observation Committee Town Hall on Observation Units answered many questions about how to set up such a unit, how billing works in an observation unit, and discussed many different models for a unit. I wanted to follow up on what I consider to be a few key points that were not addressed in detail. 

It is important to identify exactly who and/or what cause your Observation Unit serves. Besides questions during the Town Hall, a colleague at another organization reached out to me because a consultant firm reported that his organization’s observation rate was “too high,” and that an observation unit was not only necessary, but the only way to solve the issue. I counseled caution. 

As we have heard before, there is no global benchmark for observation. When you are deciding on an appropriate hospitalization status, you must be guided by the patient-specific situation and the payor-specific policies you are subject to—Medicare’s Two-Midnight Policy, for example, or your state’s Medicaid policies, or managed/commercial payor contracts. Social determinants of health (SDOH) are also important when considering use of observation services—impoverished patients with poor access to care are placed under hospital observation more frequently and are more likely to be re-observed. In my colleague’s case, his organization serves a large proportion of Medicaid-eligible patients, and we can surmise that a higher number of observation stays appear necessary to the clinicians, driven by access to care issues and payor policies. 

Would a dedicated observation unit solve this problem? That also depends on many factors. Patient selection and adequate resourcing of an Observation Unit are critical to its success, and a unit is not automatically a solution to “lots of observation patients”. I identify two specific populations of patients who get hospital observation care. The first group consists of patients with specific, often straightforward presenting conditions where a shorter period of further monitoring, an expedited, protocolized evaluation, or a limited period of targeted therapy is most appropriate. That population of patients benefits from structured, cohorted, properly resourced specialist observation care, and a dedicated observation unit is likely to benefit them (and the organization caring for them). The second group of observation patients consists of those patients who do not meet payor criteria for inpatient status and are not subject to time-based determinations like the 2-Midnight Policy. That second group may be older and more chronically ill than the first group and may require prolonged hospitalization regardless of status or medical acuity. While ostensibly receiving observation services, these patients are less likely to present with straightforward, medically acute issues that would benefit from expedited evaluation and are more likely to need extended stays. Many of us remember the pre-Two Midnight Policy era, when patients with medical needs who required hospital care but did not satisfy a commercial tool’s determination of inpatient status languished in observation for extended periods, never qualifying for rehab benefits but unable to leave a care setting. While Medicare Fee-for-Service beneficiaries are now much better off in that regard, Medicare Advantage, Medicaid/Managed Medicaid, and Commercial patients are not, and those commercial plans, generally administered by for-profit corporations, have a decided financial incentive to push patients towards hospitalized observation status. 

Another issue identified during the Town Hall was management of post-procedure patients in an observation unit. Again, one must consider what purpose the Observation Unit serves when deciding what patients are appropriately placed there. Observation services are provided to hospital outpatients while prospective evaluation and management is ongoing, leading to a medical decision point—discharge, or admission for further care. Extended recovery cares provided to an outpatient surgical patient are not the same as observation services and should neither be considered nor billed as such. Sometimes deciding whether a patient should be observation with a procedure or a surgical patient with extended recovery is not perfectly clear, but if the patient’s stay is expected to require a longer period, or they need therapy assessments or assistance with ambulation, they are not likely observation-unit appropriate. Though both may be “observation” based on payor criteria, there is a world of difference between a healthy 20-year-old Managed Medicaid patient with uncomplicated appendicitis and a 90-year-old Medicare Advantage patient with spinal compression fractures for whom someone is considering vertebroplasty pending response to medical management (a scenario posed during the Town Hall). 

In summary, a dedicated Observation Unit can be a powerful tool to help you get the right care to the right patient at the right time. With proper patient selection, an appropriately resourced Observation Unit can increase efficiency in care delivery and increase throughput, and even effectively increase your capacity. However, not all hospitalized outpatients are suitable for dedicated observation units, and only ongoing national advocacy and pressure on payors hold any hope of positively addressing the second type of observation patients.