The extra staff you thought you needed may actually be hidden right under your nose.
Perhaps one of the greatest mysteries in the emergency medicine (EM) world is staffing. How many physician staffing hours do I need? How many nursing hours? Am I using the right benchmarks? If my staff is very efficient, why do I have longer-than-desired wait times?
Many hospital executives struggle with these very questions. Unless you have had the experience of managing the emergency department (ED) as a nurse or physician director, these answers to these questions may be quite foreign to you. In order to accurately assess your ED staffing (or at a minimum, speak the same language as your ED directors), there are a several basic concepts you should know and a few specific practices you should understand.
By understanding the following five concepts, you can better prepare yourself to guide your ED leadership towards proper alignment of staff in order to maximize throughput and efficiency. In fact, by applying these strategies, you may actually find that the extra staff you thought you needed is actually hidden right under your nose.
1. Understand Queuing Theory
What is queuing theory? Essentially, it is the study of why lines form. This theory describes why we have waiting and provides mathematical formulas that allow us to predict the amount of waiting time we will have in our system and the numbers of patients likely waiting to be seen.
While we don’t expect you to be experts at queuing theory, it is helpful to understand certain concepts born out of this science. First, in order to predict your wait times and line lengths, you have to know a couple of key elements about your system. First, you must know your arrival rate and the service rate of your key servers. How many patients are arriving by hour of day? How long does a provider spend with each patient?
Secondly, understand the variation in your system, both on the arrival side and the service side. In ED settings, unless there are compelling reasons to the contrary, we assume variation is maximum due to the inherent randomness with which accidents and injuries occur and the extreme complexity of ED patients that requires varying levels of provider time.
For example, let’s say you have an average of four arrivals to triage each hour. And let’s say your triage nurse spends an average of 12 minutes with each patient. This means your nurse can process 5 patients per hour. Well, if you have four arrivals per hour and your nurse can see five, your utilization will be 80 percent.
You’d think this would give you plenty of capacity to provide timely service. The reality is, however, that due to the amount of variation and complexity in healthcare, our systems must be designed with some extra capacity to buffer the periods when arrivals exceed the capacity and when the complexity exceeds our ability to treat.1
In other words, if we staff to 100 percent of the average demand, our system will perform very poorly. In fact, it will meet the demand only 50 percent of the time. There will also be predictable times when the system is severely overrun and long waits will occur.
As the primary source of treatment for critically ill patients, the safety net for uninsured and your hospital’s front door, waiting in the ED must be minimized. In order to minimize these waits, key providers — physicians and nurses — cannot be expected to perform at 100 percent utilization, and variation must be reduced. As utilization and variation increase in the ED, wait times increase exponentially (see Fig. 1 and Fig. 2).
2. Understand the Arrival Rates & Acuity Levels of Your Patients
The first part of the equation is the arrival rate. In order to properly determine staffing, you must understand your demand. Most EDs have arrival rates similar to that shown in Fig. 3. The typical ED has a 4 to 1 ratio of peak to overnight arrivals. Except in very limited situations, we assume there is maximal variation around this arrival rate, since accidents and emergencies typically follow a random pattern.
In other words, we can measure the average arrival rates. If we assume there will be a lot of variation around these arrival rates, we will be better equipped to handle the demand with tolerable wait times.
In larger EDs, it is often necessary to segment patient arrivals into various streams of patients in order to better understand this demand. For example, in a 50,000-visit ED, it is helpful to understand the arrivals in terms of the Emergency Severity Index (ESI) distribution. For example, if we have 10 arrivals per hour and 10 percent ESI 1,2 (high acuity), 50 percent ESI 3, and 40 percent ESI 4,5 (low acuity), we can better understand the demands we will have in the different areas of our ED.
Per hour, in this example, the low acuity area will see 4, the high acuity 1, and the mid acuity area 5 patients per hour. As we will see in the following section, all we now need to do is understand our capacity in order to design the optimal system to provide timely patient care.
3. Measure Your Provider Capacity
Once you understand your demand, the next step is to determine your capacity. In the ED, capacity is essentially your physicians’ and nurses’ ability to treat. This is calculated based on the average productivity and the number of providers you have working at each hour of the day.
Every ED nurse and physician director should be able to provide their productivity figures to you as a hospital administrator. The problem is that not all productivities were created equal. In other words, physicians measure their productivity in patients per hour and nurses measure their productivity in worked hours per patient.
How did this come to be? We don’t know. However, the good news is that good old math can save the day. By taking the inverse (dividing 1 by your worked hours per patient), you get your nursing patients per hour. We prefer to express productivity in these units, as they are the same units we use to calculate queue times (usually patients per hour).
What if my ED directors can’t provide productivity figures? First, I’d take a hard look at your ED leadership and make sure they are the right people to run your department.
In the meantime, a quick shortcut is to take the average daily arrivals — over a month, for instance — and divide this by the total daily physician staffing hours (to get physician patients per hour) or nursing staffing hours (to get nurse patients per hour). In our experience, a good range for physician productivity is 1.8 to 2 patients per hour and for nursing, 0.5 – 0.8 patients per hour.
If you’re outside these ranges, you still might be OK. Things that tend to decrease productivity include high acuity mix, high admission rate, excessive numbers of midlevel providers or inexperienced physicians and confounding variables such as boarding or process-related issues.
Higher productivity ranges can be indicative of efficient processes in the ED and support areas, or favorable acuity mix. If you have high provider productivity accompanied by long waits or poor patient satisfaction, you may likely have queuing behind your providers. While high productivity is desirable, it should not be pursued at the expense of quality or service.
4. Aligning Capacity With Demand
Once you understand your arrival rates and your physician and nurse productivity, it’s time to put them together to properly match your demand to capacity. At a high level, you should take your average hourly arrival rate and overlay your physician and/or nurse capacity (number of providers on multiplied by the productivity).
Better yet, you can divide your arrival rate by your nurse or physician productivity and overlay your absolute staffing numbers as shown in Fig. 4. This will give you a general understanding of where you may have problems.
Once you have assessed the demand and capacity of an average day, if there are issues, then you should do the same for every day of the week, as there are few EDs that don’t have significant variation in day of week arrival rates. The average ED, for instance, has a 10-20 percent difference between Monday and Saturday average arrival rates. Interestingly, most EDs we have worked with have the same staffing patterns 7 days a week.
Another consideration is boarding patients. It is important to understand that boarding patients consumes nursing resources, but not physician resources. Also consider trauma and other infrequent but high-demand patients when attempting to accurately align your demand with your capacity.
Fig. 5 shows an analysis in a 70,000-visit ED, which had the same pattern of staffing 7 days a week yet had significant variation in boarding hours between the weekdays and weekends, and also had a 20 percent difference between Monday and weekend arrivals. By aligning the staffing with the demand from all sources, we found the equivalent of approximately 10 wasted nursing full time equivalents (FTEs).
This waste was hidden in either too many or too few nurses staffed by hour of day throughout the week. Having too few nurses at certain times means you cannot care for the arriving volume, having too many nurses is equally harmful since this excess is frequently hidden in reduced productivity levels on these days.
5. Continuously Reassess Your Staffing
Once you have appropriately aligned your staffing with your demand, you must sustain this system through continuous measurement and realignment. A major flaw in the way our EDs function lies in the annual budgeting of human resources. By annual budgeting practices, we cannot respond to changes in demand in a fluid manner.
In other words, if the volume of patients increases by 10 patients per day (as we have seen in some EDs over a period of just 2 months), the subsequent increase in nurse hours to accommodate this demand would be on the order of 15 daily nursing hours and 5 physician hours. We know from queuing theory that if we do not accommodate to this demand, wait times will increase dramatically, but if we use traditional staffing budget allocations, it may be 10 months until we can modify staffing to match the increased patient demands
In the upcoming years, we will have to rely on more fluid administrative processes to handle these fluctuations in demand, both increases and decreases. Many leading EDs are now using their emergency department information systems (EIDS) to provide real-time feedback as to their staffing levels and adjust up or down based on real time demand information.
If you chose to use real-time demand/capacity matching, realize that this should be done on top of a solid planning platform or you will frustrate your staff and require excessive numbers of as-needed (PRN) and temporary staffing. This may actually cost you more money and be harmful to your staff culture and morale.
To properly understand an address potential staffing issues, we must understand queuing theory and apply it to the ED. This requires us to properly assess our arriving demands and our capacity to treat based on current staffing levels by hour of day and day of week. We then must properly align our resources and monitor these staffing allocations closely in order to meet the dynamic changing environment that is the emergency department.
Noon, C. (2003). Understanding the impact of variation in the delivery of healthcare services. Journal of Healthcare Management, (48)2.
Crane, J. & Noon, C. (2011). The definitive guide to emergency department operational improvement. New York: Productivity Press.