I received an excellent and tough question from Randy Ford, a St. Luke’s continuous improvement specialist, following my blog post from April 30, “Why Add Beds?”
Here’s what Randy asked:
Has there been any discussion on when, how much, or where adding additional beds would place us in diseconomies of scale? Otherwise stated, when is more no longer better, and the average unit cost would increase?
Some blog readers might not be very familiar with the concept of “diseconomies of scale,” so I’ll start by way of contrast.
“Economies of scale” refers to the ability to lower costs as a consequence of size.
For example, Walmart achieves economies of scale by being able to achieve larger discounts on purchases of items it stocks than a smaller stand-alone retail store could achieve, and therefore can sell the same item cheaper and still make a profit.
Walmart also is large enough – that is, it has enough scale – to have its own supply and distribution centers and can operate this portion of its operations at lower cost than purchasing the service from another company. Smaller organizations would not be able to do so. Randy is asking whether at some point, St. Luke’s expansion would result in diseconomies of scale, that is, inefficiencies, and as a consequence, average unit costs would increase.
To further complicate matters, the answer changes a bit as we move from fee for service to pay for value. Let’s start in today’s fee-for-service world.
One of the very smart things St. Luke’s has done is to rationalize the scope of services offered in our communities.
I have talked to many of my colleagues elsewhere who have been pressured to replicate services. This process was on steroids in Houston, where there are more than 67 hospitals in the greater Houston area and at least 22 of them performed open heart surgery. This resulted in many examples of diseconomies of scale. Quality and efficiency would have been far greater if fewer hospitals did more of these surgeries. St. Luke’s gets economies of scale by rationalizing the scope of services and not having every hospital trying to perform every service.
An associated problem is costs driven up when hospitals compete to have the same advanced technology and pay whatever it takes to get physicians and staff to move from their competitor to their hospital. This is one of those strange things about health care, as opposed to other industries, where an oversupply of suppliers doesn’t decrease costs, but actually contributes to higher costs and poorer outcomes.
Fortunately, this is not as significant an issue in Idaho. According to the Dartmouth Atlas, Idaho is in the lowest quintile for Medicare payments per enrollee. Costs to the Medicare program for care of Medicare enrollees is lower in Boise and its associated hospital referral area than any other area in Idaho or surrounding states, except for Portland.
Further, the intensity of services to Medicare enrollees as measured by hospital discharges per 1,000 Medicare enrollees is lower in Boise and its associated hospital referral area than any other area in Idaho or surrounding states, except for Bend, Ore.
Economies of scale are also impacted by the design of nursing units. Depending upon the acuity level of patients, a certain configuration of beds and nurses works optimally.
For example, if the nurse staffing ratio for an intensive care unit was one nurse to two patients, it would be far more economical to build a smaller eight-bed unit than a larger nine-bed unit. Furthermore, we have learned that efficiencies can be significantly affected by facility design, and therefore, we intend to use TEAMwork efficiency principles, based on Lean methodology, in our design of future facilities.
There are other opportunities for efficiencies in current hospital operations that can decrease the need for new beds. Length-of-stay reductions can significantly free up beds to decrease capacity pressures, and we can pursue opportunities in this regard. We perform pretty well against benchmarks, but we know that there is room for some improvement.
Frankly, if hospitals wanted to maximize efficiencies, they would operate as 24/7 operations. An example is that there are very high fixed costs associated with operating rooms, yet I don’t know of any hospitals that smooth out their utilization to open up capacity and promote maximum use of this high-cost asset throughout the day and night and over weekends. How many patients wait to have a test until the next day because of a delay in a needed procedure? How many patients could have been discharged at the end of the day instead of the next morning? How many patients could have been operated on over the weekend, or discharged over the weekend? We don’t operate at the same level at night, on weekends, or on holidays that we do weekdays.
To fully discuss the issue raised, I have to address a sensitive subject: critical access hospitals, which are certain small rural hospitals. Are critical access hospitals efficient and the most financially prudent use of resources? Can you achieve economies of scale with critical access hospitals? These are some of the questions that come up.
I have heard some in our industry argue that you don’t really get economies of scale unless the hospital is 200 beds or more. We think our critical access hospitals are run pretty efficiently, but there are critical access hospitals in this state with an average daily census of less than five patients. That makes the cost per occupied bed very high. Is that efficient? Can you then argue that the critical access hospital should be consolidated with the nearest hospital? What if there isn’t another nearby hospital? These are very difficult questions.
There are economies of scale that we bring to our critical access hospitals, and I think that is largely what led each of them to want to be integrated with our System, rather than continuing to be a stand-alone facility. Stand-alone hospitals cannot get the economies of scale through purchasing power of medical supplies, equipment, and medications. Further, we have been able to centralize some services and to decrease the amount of senior management on site through the System.
It’s true that more beds can lead to diseconomies of scale when there is low volume, duplication of services, poor design, and/or oversupply of beds relative to needs.
I feel good about our plans because we are avoiding unnecessary duplication of services by rationalizing scope of service, using TEAMwork in our design, and taking a very conservative approach when it comes to bed demand, building at half the rate of beds per thousand than the benchmarks we previously would have used.
And, unlike Houston, which will have to deal with excess hospital beds in the market, the number of beds needed in some of our communities is consistently projected to be more than the current number of beds available.
But we can’t take our focus off of trying to do everything we can to reduce waste and inefficiency in our current operations. And our front-line staff members often know best where the waste is. We need everyone’s input to optimize our current systems.
With all that said, pay for value will change the way we try to answer this question. Today, hospital beds are profit centers. For the most part, we want them filled. Tomorrow, under pay for value, hospitals will be cost centers, and we will want the beds empty.
Under pay for value, you want patients in the right place at the right time, but in the lowest cost setting where you can still get the best outcome. That often may be at home.
I spoke in my recent employee forums about a new model in which some care that is provided in hospitals today may be provided in the patient’s home in the future. The use of home care and palliative care will increase. Interestingly, this will put some patients in a lower cost setting, but may not be the most efficient way for us to provide the care.
We must carefully plan to optimize the economies of scale. If misjudged, diseconomies can result and the average unit cost can actually increase. That’s another difference under pay for value. Cost will be absorbed by us!
Thanks for your very difficult question, Randy!
David C. Pate, M.D., J.D., is president and CEO of St. Luke's Health System, based in Boise, Idaho. Dr. Pate joined the System in 2009. He received his medical degree from Baylor College of Medicine in Houston and his law degree from the University of Houston Law Center.