Portal de Periódicos da CAPES

Costly Emergency Department Expansions Are Ineffective to Improve Flow Without Addressing Culture and Process Efficiency

2014; Wiley; Volume: 21; Issue: 5 Linguagem: Inglês

10.1111/acem.12372

1553-2712

Jesse M. Pines, Khaled Alghamdi,

Healthcare Policy and Management

In this issue of Academic Emergency Medicine, Mumma et al.1 describe the effect of an emergency department (ED) expansion on metrics of ED flow, at the University of California–Davis, a large academic medical center. After expanding adult bed capacity from 33 to 53 beds, a 60% increase, two notable things happened: ED volume increased by 13%, and there was an increase in ED boarding hours. Importantly, there was no effect on other measures of ED flow, such as left without been treated rates or overall length of stay. This was an unexpected result, as ED expansions typically are implemented with the goal of improving flow. Theoretically, extra space from an expansion means additional treatment rooms should be available so patients will not be required to line up and wait in the waiting room. Instead, they should move directly into an available treatment room—one that was not there previously—to be seen by a provider, have the work-up started, or be more rapidly treated and discharged or admitted to the hospital. However, this study and another describing similar findings from a doubling of the ED2 demonstrate that the results of an ED expansion on flow metrics do not always go as expected. After expanding the size of the Vanderbilt ED from 28 to 53 beds, daily volumes increased, and both average length of stay and the boarding time increased by about an hour.2 There was also no effect on the rates of ambulance diversion. To understand why ED expansion does not have a long-term effect on ED crowding, we must first explore why EDs become crowded in the first place. A good way to conceptualize ED crowding is as a mismatch between the demand for patient services and the supply of resources available to process them. When demand exceeds supply, ED crowding—ultimately a line of people waiting for services—worsens. We must also recognize two additional factors. The first is that ED functioning is highly dependent on hospital flow—in the case of boarding and hospital bed availability. The complexity of the service delivery in the ED also allows staff in the ED and hospital to adapt—ultimately changing their own workflows—to any structure or process change. On the demand side, both the UC Davis and Vanderbilt EDs noted higher volumes immediately after expansions. Fundamentally, demand for ED care is related to accessibility of its services (e.g., ED supply). However, the nature of this relationship is complex. Certainly, much of ED demand is exogenous—determined by the community acute care needs as people become ill and get injured. Yet expanding the size of EDs in some communities may cause more people to seek care and can be endogenous to the structure of the ED, as likely happened at both UC Davis and Vanderbilt. This is especially true if community needs are not being met, which probably was the case at both sites. However, there is almost certainly an upper limit on community needs for acute care services. Say the UC Davis ED was to expand its ED more from 53 to 100 beds. It is possible this might not affect demand because at that level of supply, demand may be "topped out." Both EDs were likely undersized for their yearly volumes, which both led to the decision to expand and to the community demands to increase in response. A similar phenomenon is also likely playing out on a larger scale with the recent, dramatic expansion of urgent care centers across the United States. While some may believe that these urgent care centers will lower ED visits with a 1:1 substitution effect and lower overall health care costs, the more likely scenario is supply-induced demand and higher overall health care costs. This is similar to what is observed during ED expansions. The adage "if you build they will come" holds: some may choose to use urgent care centers over EDs, but ultimately overall demand will increase along with the supply of available facilities.3 The supply side is a similarly complicated effect in the ED. In discrete event simulation computer models, holding all other factors constant, increasing ED size should reduce length of stay.4 Treatment space number is a model "constraint," and by increasing it, theoretical patients spend less time waiting for an available bed. However, all other factors are not held constant as they can be in a computer simulation model. The first issue is that increasing size increases demand. Another major factor that also cannot be held "constant" is human behavior itself: ED staff—and hospital staff—likely change their behavior in response to expansions. In EDs and hospitals, staff are required to be flexible to increase the speed of work during episodes of high workload. On any given day in the ED, independent of any expansion, both ED volume and the average work required per patient vary tremendously. After an ED expansion, ED staff may have more treatment spaces available, but may also dynamically change the speed of their work in other ways to shift back to what they may see as a suitable level of performance. This may be why length of stay and left without being seen rates were not affected in the study by Mumma et al.1 On the hospital side, hospital staff similarly may have changed their workflows to the newly expanded ED space. Moving patients to inpatient beds rapidly from the ED may have become less of a priority to inpatient teams with the knowledge of the new ED space. Alternatively, there may have been less pressure from the ED to push patients to inpatient beds because of the additional space. Regardless, the effect in both expansions was to increase ED boarding as a behavioral adaptation to greater ED capacity. In the literature, there have been many published studies describing successful interventions to improve ED flow.5 Common threads among successful interventions involve several critical elements.6 The first is a close examination of ED processes and identification of wasteful steps or duplication (e.g., Lean).7 The second is the presence of an ED leadership team who effectively communicate the priority to improve flow and gets buy-in from staff. Leadership aims to change the ED culture so that faster, more efficient care is a priority, which prevents some staff from adapting their own personal workflows to match "usual" performance levels. The third is a supportive hospital leadership that provides encouragement and resources to the ED team and, where possible, tries to prevent the interdependent resources in the hospital, such as inpatient teams, from adapting to counteract improvements in ED performance. The fourth element is the ability to use valid data to choose and monitor progress. The final factor is commitment over a long period of time: ultimately what creates better performance is culture change and the ability to reinforce new processes when, inevitably, the staff in the ED and hospital tries to adapt their own workflows and slide back to "usual" performance. So, should we conclude that expanding EDs is completely ineffective at improving ED flow? If an ED is undersized for flow, expansion may be necessary to accommodate existing need to meet community demands. However, results from two large expansions in large academic medical centers clearly demonstrate that expansion alone is insufficient to improve flow without addressing some of the other factors that cause crowding, such as ED and hospital culture and inefficient processes.