Through the use of a sophisticated modeling technique, investigators at the University of Cincinnati have found that the creation of a so-called "flex track" that includes beds that can be assigned to either high-acuity or low-acuity patients has the potential to lower mean wait times for patients when it is added to the traditional fast-track and high-acuity areas of a 50-bed ED that sees 85,000 patients per year.
Investigators used discrete-event simulation to model the patient flow and characteristics of the ED at the University of Cincinnati Medical Center, and to test out various operational scenarios without disrupting real-world operations.
The investigators concluded that patient wait times were lowest when three flex beds were appropriated from the 10-bed fast track area of the ED.
In light of the results, three flex rooms are being incorporated into a newly remodeled ED scheduled for completion later this spring.
Investigators suggest the modeling technique could be useful to other EDs interested in optimizing their operational plans. Further, they suggest that ED administrators consider ways to introduce flexibility into departments that are now more rigidly divided between high- and low-acuity areas.
The practice of funneling low-acuity patients into a "fast-track" area so that they can be seen and discharged more quickly is commonplace. Patients with higher levels of acuity are cared for in the main ED, an area set up to handle higher-level care. This type of organization has helped to improve efficiency and resource allocation in many departments. But is this sort of division the best way to optimize resources and limit patient waiting?
Not necessarily, according to some intriguing new research conducted at the University of Cincinnati (UC) in Cincinnati, OH. Using discrete-event simulation to model patient flow through a 50-bed ED that receives 85,000 patients every year, investigators have concluded that the creation of an intermediate "flex" track between the high- and low-acuity areas could substantially reduce patient wait times and further improve efficiency.1
Specifically, the analysis found that the addition of a three-bed flex track produced a mean patient waiting time of 30.9 minutes. The traditional division between fast track and high-acuity beds produced a mean waiting period of 40.6 minutes, and a department where all the beds were totally flexible produced a mean wait time of 35.1 minutes.
While a real-world test of the approach has yet to take place, investigators believe that the modeling approach used in this case could be helpful in pointing other ED leaders toward operational models that are ideally suited to their departments.
Recognize system imbalances
Trying to balance staffing and resources with incoming demand is a constant challenge for ED administrators. Indeed, what prompted investigators at UC to delve into the issue was the observation that this balance was at times suboptimal in the ED at the University of Cincinnati Medical Center (UCMC), a teaching hospital and level 1 trauma facility.
"I was spending some time in the ED and saw periods when the main part of the ED was full with waiting [patients] and there were available beds in the fast-track area," explains Lauren Laker, MBA, the lead author of the study and a PhD candidate in operations, business analytics, and information systems in the Carl Lindner College of Business at UC. "We had some unused resources and beds that were there and not being used, so the question became was there a better way to use some of these resources to balance out some of that demand variability that happens throughout the day."
To answer this question, Laker and colleagues decided to use discrete-event simulation, a modeling technique that enables testing of various patient flow scenarios without disrupting the operations of the ED.
One of the advantages of using discrete-event simulation is that you have the ability to include many different variables, observes Laker. "We incorporated service time and a lot of different factors like that in our analysis," she explains. "You can actually incorporate a lot of movement. It allows you to [describe] what is going on, and lets you do a pretty good job with sensitivity analysis as well as testing a lot of different things within one model."
Using this modeling approach, investigators concluded that when a three-bed flex track was created that could accommodate both high-acuity patients categorized as emergency severity index (ESI) 2 or 3 and low-acuity patients categorized as ESI 4 or 5, patient wait times were more than 30% lower than when ED beds were more rigidly assigned as either "high-acuity" or "low-acuity." In the analysis, the flex-track approach also outperformed total bed flexibility, in which any bed could be assigned to any patient regardless of acuity.
Play with different flow scenarios
Specifically, the flex-track beds were taken from the fast-track area, which typically contains 10 beds. Further, while high-acuity patients had priority access to the flex beds, these beds were also available to low-acuity patients when the 40 high-acuity beds were not filled to capacity. The model was based on the patient volume, size, and characteristics of the ED at UCMC.
Investigators tested scenarios, setting aside as many as five flex beds, but found that the optimal number of flex beds for this ED was three. "Three beds isn't this global universal policy that works well all the time. It just happens to work well a lot of the time," explains Craig Froehle, PhD, a professor of operations and business analytics in the Carl Lindner College of Business and College of Medicine at the University of Cincinnati, and a co-investigator on the research. "Then, if you move to a period where you have a slightly different mix of patients or slightly longer or shorter service times ... then you will find that a slightly different number of flex beds will produce the optimal [results] in terms of the lowest amount of patient waiting overall."
The point is that the optimal number of flex beds for EDs with different volumes and characteristics is likely to be different, but the research suggests that some level of flexible capacity could well be beneficial in many EDs, notes Froehle.
In fact, in light of these results, hospital administrators at UCMC have incorporated room for a flex track of sorts in a new structure scheduled to open this spring. "The redesign of the front end and the intake area [of the ED] has three of what they are calling flex beds that will be able to accommodate both your traditional low-acuity/fast-track patients as well as patients of moderate acuity, ESI 3 or ESI 2," says Froehle. "The planning is already underway to make sure those [beds] are integrated into operations as fluidly as possible."
Incorporate more flexibility
Froehle believes that discrete-event simulation as a technique has a lot to offer hospital administrators who are in constant search of better ways to match resources with demand in their EDs. "The strength of discrete-event simulation is really in terms of its flexibility," he says. "You can apply it to a vast number of scenarios, including very simple problems as well as very, very complex problems."
Froehle adds that discrete-event simulation is much more scalable than cuing analysis or a mathematical approach, but he stresses that it requires good operational data, which are challenging in health care environments. "One of my biggest frustrations whenever I go into any study is the lack of good, high-quality, operational data," he says. "Electronic health records have gone a long way toward collecting better clinical data, but in terms of time-stamping actual events, we still have a long way to go."
Indeed, to carry out this study, investigators relied on time stamps from ED operational data in order to model patient flow through the ED, beginning with the arrival of the patient and proceeding through triage, patient waiting, service time, and discharge. For the results to be meaningful, the data must be reflective of what actually occurs.
In addition to the challenge of obtaining high-quality operational data, some hospitals may also find it challenging to access the specialized expertise needed to conduct discrete-event simulation exercises, acknowledges Froehle. However, he notes that even if this type of specialized expertise is unavailable, hospitals can still take steps to incorporate more flexibility into their ED operations.
"Think about what it would take to even have just one room that might flex between a fast track and a traditional role as needed, how it would be staffed, what physical changes might be needed, and what kind of rotation implications it might have if, for example, it is a teaching hospital," suggests Froehle. "Working through some of those managerial issues would be a useful first step [for ED administrators], even if they only want to use one room as a pilot."
However, for hospitals that do have access to the necessary expertise, Froehle sees a number of advantages to using discrete-event simulation to improve operational efficiency. "You don't disrupt day-to-day operations, you don't invest a lot of money, and if you have data, then they can be a fairly cost-effective and less disruptive way to evaluate what the best call should be," he observes.
Focus on core concepts
The idea of using flexibility as a technique to improve operational efficiency is just the latest tool that health care has borrowed from other industries, offers Froehle. "We have been working on flexibility as a technique in other industries for many, many decades, so it is rewarding to see that the same lessons that we have learned from manufacturing and hospitality service industries are just as applicable to health care," he says. "And the reception and the response have been very good."
While there are no current plans for further study of the "flex track" idea, Froehle notes that it would be interesting to explore how the flex beds are implemented in the ED at UCMC later this year. "The physical redesign of the ED is quite substantial, so the learning curve is going to be significant for everybody,"says Froehle, although he acknowledges that it would be very difficult to separate the effect of the flex capacity vs. the effect of the overall system change.
Nonetheless, with increasing pressure on EDs to conserve resources while caring for higher volumes of patients, the authors are engaged in further work on the larger issue of how best to make use of operational flexibility. "The core concept of flexibility is not novel," notes Froehle. "But what we are seeing is that we are now getting to the point where we can have conversations with informed components of the health care delivery system to say: Here are some core concepts. How might we employ them to make the care delivery system better?"
Laker L, Froehle C, Lindsell C, et al. The flex track: Partitioning between low and high-acuity areas of an emergency department. Ann Emerg Med 2014;64:591-603.
Craig Froehle, PhD, Professor, Operations and Business Analytics, Carl Lindner College of Business and University of Cincinnati College of Medicine, Cincinnati, OH. E-mail: [email protected].
Lauren Laker, MBA, PhD Candidate, Operations, Business Analytics, and Information Systems, Carl Lindner College of Business, University of Cincinnati, Cincinnati, OH. E-mail: [email protected]