Many hospitals in the country operate at capacity, and many patients are boarded in the emergency department. Although there are many ED-based flow initiatives, virtually none of these address the most significant impediment to flow: boarding of admitted patients in the emergency department due to lack of inpatient beds. Only a few interventions really have any lasting and significant impact on boarding and capacity.
This is one of a series of interviews that highlight dramatically effective interventions to reduce boarding and crowding. Eugene Litvak, PhD, is a world-renowned expert in hospital flow who made the remarkable discovery that our problem with capacity is driven in large part by elective scheduling, not by ED admissions. We sat down to discuss his experiences tackling the issue of hospital crowding.
PV: This is one of a series discussing hospital and ED crowding and its impact on patient safety, finance, and staffing. How did you get into the whole arena of hospital capacity and flow?
EL: I came to this country in 1988 from the former Soviet Union. I already had dozens of publications in the United States, and many of my colleagues recommended that I should go anywhere but health care because in this industry, efficiency is not a goal; there is no interest to increase efficiency. That was a red flag for me. I started doing some limited consulting at hospitals and started trying to learn the environment at the hospital, working with the frontline people, ie, nurses, physicians, etc. At that point, I met Dr. Michael Long, an anesthesiologist. At that point, the question that we were trying to address was, what happens with hospitals overcrowding? We found that at the same time hospitals are getting more and more overcrowded, the hospitals’ census and bed occupancy experienced large fluctuations. My initial belief was that everything stemmed from the emergency department. There are two main portals to any hospital. Emergency departments are responsible for over 50 percent of all admissions, and there are elective admissions, mostly surgical, typically responsible for up to 30–35 percent of admissions, the remaining admissions being medical referrals, transfers, etc.
PV: So your first assumption was that this was due to influx of emergency patients?
EL: Absolutely. It was based on the common sense for two reasons: First, the volume is the highest among all admits, and second, it’s unpredictable by its nature. Elective admissions are smaller in terms of the volume, and their schedule is up to us. Unfortunately, our health care delivery is not always based on common sense.
PV: What did you find?
EL: It was impossible for Dr. Long and me to get the data from emergency departments. Nobody wanted to share the data with us. However, we were able to get the data from one operating room. Two transparencies were on the desk in front of us. One of them was bed occupancy, and the other was surgical volume. We found they were overlapping. They had about the same shape. So if you put it up to the window glass and overlap one over another to compare, we found that they practically coincided. That was for me a real aha moment: Emergency department admissions had very little to do with variability. Since then, for years I have talked to many hospital emergency department leaders asking, “Five Tuesdays from now, short of a bus crash or flu epidemic, could you predict approximately how many patients are going to be admitted to your emergency department?” The answer was always yes. Then I asked many operating room managers the same question: “Five weeks from now on Tuesday, how many surgeries are you going to perform?” Given that typically over 70 percent of all surgeries performed are elective, I was very surprised to find out that people cannot answer this question. That, to me, was clear evidence regarding the source of this variability. Of course, this was not just the surgical admissions. This is true for the other elective admissions, eg, cath lab.
PV: Do you find this to be true at most institutions?
EL: Practically everywhere. In dozens of hospitals where I asked this question, the answer was the same. It’s not just in the United States. It’s true in Europe, Canada, you name it. It looks like an international plot against health care cost and quality and the main driver of capacity problems.
PV: In response to this, there were three things that were implemented that we refer to as smoothing: separating out the emergency surgical flow from the elective surgical flow, smoothing the number of surgeries over the week, and also smoothing them to predict the number of ICU beds needed.
EL: That is absolutely correct. Moreover, I would say that’s not an intervention. That is the intervention. We have only two options. The first option is to provide excessive resources to staff at the peak level, which no hospital in the world has resources to do. The other choice is to staff below the peak level, a pivotal way of staffing hospital wards today. Typically, we staff them at the average level that has been documented historically from the last year. About 10 years ago, we received a grant from the Robert Wood Johnson Foundation to study two community and two academic hospitals, and we found that ward bed occupancies changed every hour, if not every half hour. There is absolutely no way that one may have a pool of nurses dynamic enough to address heavy peak volumes because nurses do not live in the hallway to address every hour or half-hour change in the census.
PV: In the places where you helped to implement smoothing of the elective schedule, what was the end result?
EL: The end result was huge, both financial and quality wise at every hospital. Cincinnati Children’s is probably one of the most impressive examples. When we started working with them on smoothing, their census was at the 76 percent level. In order to address peaks, they planned to build a new tower for $100 million in capital costs. Each bed in the United States, in terms of the capital cost, varies from $1.5 million to $3 million in capital cost alone. Plus, the annual operational cost per bed is at least half a million dollars. At the end of our smoothing project, they abandoned their plan to build the new tower. The average census reached 91 percent, a 15 percent increase. Their surgical volume increased dramatically without capacity issues because when we cut off the peak, we filled up the valley. It’s not just the peaks that create quality consequences; when we have those valleys, that’s a waste of our resources. Their surgical volume dramatically increased, and according to their report, their margin improved by over $100 million a year. It’s not just $100 million in avoided capital cost; it’s an additional over $100 million a year margin improvement and quality improvement.
PV: I understand that hospitals had had significant problems with boarding in the emergency department that also disappeared as soon as they smoothed their capacity.
EL: Boston Medical Center is a Level I trauma center. Their emergency department was constantly overloaded. After surgical smoothing, their ambulance diversion decreased by 20 percent. Their waiting time dropped to 2.8 hours compared to five plus hours at other academic hospitals in Boston. Improvement in the emergency department overcrowding was not at the expense of the surgeon. Due to their nature of being a Level I trauma center, their cases were frequently bumped by the emergent surgeries with gunshots, etc. The number of cancelled or rescheduled cases dropped by 99.5 percent, from the average of 700 a year to about six a year.
PV: What are the upsides and the downsides?
EL: Ottawa Hospital is a large academic hospital. They reported a $9 million margin improvement, and they reported 40 lives were saved in the first year. Why? Because they documented that when the hospital is overcrowded and the operating rooms are overcrowded, the waiting time to get emergent or urgent surgery could become prohibitive, resulting in an increased mortality rate.
PV: Hospitals are doing so many different things to address crowding. Few have been effective or sustained. Would you consider this intervention just one of many on the list of things that hospitals can do?
EL: As long as we have those peaks, we are going nowhere. Let me give you another example from 2009 publication in Critical Care Medicine. At the Johns Hopkins neurological ICU, authors found that during peaks in admissions, the hospital readmission rate increased by 500 percent. What does that mean? I believe that [the Centers for Medicare & Medicaid Services] suggest that there should be a 20 percent reduction in the hospital readmission. If you do not smooth, you could report a success, with 400 percent instead of 500 readmission rate during those peak days. When I say that Cincinnati Children’s was able to improve their margin by $100 million a year, hospitals of similar size that do not do that will waste $100 million a year. In terms of safety, cost, readmission rates, and mortality rates, it’s dangerous to the patient and the financial well-being of the hospital to ignore these peaks and troughs. I consider this an absolutely essential part of any effort to address crowding. Without it, you will not solve your problem.
PV: In summary, you have found that smoothing improves the hospital’s available capacity to decrease emergency department boarding of admitted patients, and you’re going to have steadier nurse-patient ratios without peaks and valleys. It’s going to be safer for the patients, it’s going to be better for the hospital’s financial line, and it’s going to actually be better for the doctors, particularly those that are doing elective surgery, because they don’t get their cases bumped, their patients are placed in the preferred rooms, and one can run the elective operating room with a higher capacity because it’s much more predictable.
EL: That is absolutely correct. So what is the alternative today? Let’s build more beds. The average hospital bed occupancy in the United States is much lower than in any industrialized countries. In the US, it’s about 66 percent on average. One-third of our hospitals are empty, and yet we are overcrowded. That’s everyday life compared to Canada, for example, when their average bed occupancy is 90 percent. We have this luxury of having a lot of beds, and yet we are overcrowded. Building more beds would not solve the problem.
PV: Some hospitals reading this will say, though, that they run at an average occupancy of 85–90 percent. Would this apply to them?
EL: Cincinnati Children’s census is about 90 percent. That’s the same as in Canada. In Canada, when we started working with the Ottawa Hospital on these issues, they reported their census in excess of 100 percent. If your average bed occupancy is 85 or 90 percent, then every peak in census hits the ceiling. Every peak means that emergency patients are going to be boarded, quality of care is diminished, and yet the next day’s valley will result in waste. In short, hospitals lack capacity because of the way they choose to do business.
PV: What does it take to make this happen? Why isn’t every place adopting this?
EL: That’s a key question. The answer is multifactorial. First and foremost, if the hospital does not have an inspired and committed leadership, it’s not going to happen. If the hospital CEO, personally, is not supportive of this intervention, it’s not going to work. Second, surgeons do not realize that if they agree to smoothing, they would increase their volume, reduce their overtime, and improve their and patient satisfaction. At Cincinnati Children’s, despite a one-third reduction in waiting time for emergent and urgent surgery, they increased the number of cases and yet the overtime dropped by 57 percent.
“When I say that Cincinnati Children’s was able to improve their margin by $100 million a year, hospitals of similar size that do not do that will waste $100 million a year. In terms of safety, cost, readmission rates, and mortality rates, it’s dangerous to the patient and the financial wellbeing of the hospital to ignore these peaks and troughs. I consider this an absolutely essential part of any effort to address crowding. Without it, you will not solve your problem.” —Eugene Litvak, PhD
PV: I think the principle could be said, by a surgeon, that you don’t cure constipation by adding more colon.
EL: Absolutely. Leadership and education are critical. Surgeons should be educated to appreciate the benefits of this intervention. The third reason is that in order to accomplish smoothing, hospitals should do pretty intense data analysis. Not all hospitals have these resources, and the government should do its job to invest in hospitals getting the necessary technical support. Last but not least, I think emergency physicians must do a better job of explaining to the public the real cause of overcrowding and boarding. No matter what you do in your emergency department—and I am not suggesting that emergency departments are flawless—you alone cannot resolve overcrowding. That message should be known by the public.
PV: What would happen nationwide from implementing this intervention?
EL: The return on investment would be huge. In 2012, two leading US health policy experts, Dr. Arnold Milstein and Dr. Stephen Shortell, in their piece “Innovations in Care Delivery to Slow Growth of US Health Spending” in the Journal of the American Medical Association, estimated that national diffusion of patient-flow optimization—optimally managing patient demand and health care capacity—has the potential to reduce total US per capita spending by 4 percent to 5 percent, which is $120–$150 billion a year. This intervention does not require capital investments. Quite the contrary, hospitals that implemented this approach saved millions of dollars and many human lives.