Why Lean Fails: The Challenge of Variability

Posted on Tue, Sep 22, 2015
Why Lean Fails: The Challenge of Variability

Part Two of a two-part series.
By Joseph Twanmoh MD, MBA, FACEP, FAAEM

In Part 1 of this two-part series, we discussed the history of the lean process improvement methodology, its basic tenets and how lack of leadership commitment can doom a hospital’s efforts to achieve efficiencies with lean projects. This post explores the differences between manufacturing and healthcare, and how these differences can adversely affect lean projects without mitigation.

Leadership commitment is crucial to reaping the rewards of lean. Teams need the appropriate support to plan and execute lean projects. Time is required to meet, collect data, test ideas and make any necessary changes prior to implementation. This is the classic Plan-Do-Check-Act Cycle of process improvement. When leadership doesn’t understand lean, they believe that they can just add on these additional responsibilities to managers and front line staff; dedicated time and resources for the lean improvement process isn’t viewed as a priority.

When lean is used as a tool only, piecemeal implementation can occur. Hospitals may do a rolling start, where lean is introduced in one department at a time instead of house-wide. This leads to communication failures between the departments and hampers lean implementation. An example is when the emergency department begins using lean, but the registration department does not. The ED team re-engineers their patient intake process and later informs Registration of some changes that are planned. The ED Lean team explains to Registration that the new process will revolve around the needs of the patient, add value to the patient, and reduce waste and non-value-added activities. Registration’s response is that the old process is fine because “this is the way we’ve always done it.”
Another phenomenon that occurs when lean is simply used as a tool is discontinuous improvement. Lean is all about continuous improvement. The quest for quality is never finished, with teams constantly moving from one project to the next. What happens all too often in healthcare is the “one and done” mentality. The team works on the particular project, finishes that project, and then considers their job finished. They never monitor the effectiveness or go back for a second or third round of improvement. The result is initial gain followed by backsliding, so the conclusion is that lean doesn’t work.
Choosing the Right Lean Project
Internal lean teams often work without the help of expert guidance. The result is that teams can pick the wrong project. If they pick projects that are too small, it results in meaningless outcomes.
For instance, an emergency department could have very long patient lengths of stay and one aspect of that problem could be the turnaround time for CT scans with contrast. However, if a team chooses to work at reducing CT turnaround time with a goal of reducing overall length of stay, it’s unlikely that its work will have any meaningful impact, because CT turnaround represents only a portion of the patient’s length of stay, and the number of patients receiving CT scans with contrast represents an even smaller percentage of the total number patients in an emergency department. Consequently, any time savings on CT turnaround is diluted when calculating overall length of stay. If a team chooses a project that is too large, such as reducing the length of stay for both discharged and admitted patients in the emergency department by 50 percent in the next six months, then the number of processes and workflows to be re-engineered becomes overwhelming and the team fails.

The Challenge of Variability: Manufacturing and Medicine Are Not the Same
Lean has a number of inherent shortcomings when applied to healthcare. First, lean delivers incremental change. As a result of incremental change, it can take a long time to get the desired outcome.
For instance, shortening ED length of stay is a very complex process. There are hundreds of steps in different processes that happen from the time a patient arrives until the time of discharge or admission. Taking on this project from a lean perspective requires breaking down the patient’s ED visit into small, manageable pieces. Therefore, it would take a series of lean projects to make a meaningful impact on length of stay. Particularly when leaders are not fully committed to the lean process, this incremental change is perceived to take too long, and therefore efforts often are abandoned.

Lean may be helpful to get one from Point A to Point B more efficiently. However, lean may not be helpful in telling you where Point B is. In most hospitals, the staff of nurses and physicians may have only worked in one or two other institutions. Consequently, if you’ve never worked in or been exposed to a highly efficient system, how would you know how to design one? If you take a group of people who are not experienced bakers and give them a recipe for an apple pie, what’s the likelihood that they will produce a world-class apple pie? Probably fairly low. Yet we put together teams of healthcare professionals who may never have worked in an operationally excellent facility and expect them to come up with a world-class process design using lean. Without appropriate expert guidance, their likelihood of success is fairly low.

In addition, there is variability in healthcare that lean has trouble addressing. Toyota believes that the right people following the correct processes will result in desired product. This is not the case in healthcare. Not all patients with pneumonia will respond to antibiotics, even when appropriately chosen and administered in a timely fashion. Not all patients with heart attacks survive even when life-saving cardiac catheterization is available in 90 minutes or less.
In manufacturing, one source of variability is materials. Toyota works hard to reduce the variability of materials. It works closely with suppliers to ensure that parts are free of defects. However in healthcare, our materials are our patients, and patients are rarely alike. Age alone is a major variable. Caring for a newborn is much different than caring for a 5 year old. Caring for a 20 year old is much different than taking care of an 80 year old. A morbidly obese, hypertensive, diabetic 50-year-old male with an ankle fracture is a much more challenging patient to care for than a healthy one.
The other main sources of variability in manufacturing include man, methods and machine. Man represents the workers. At Toyota, its workers are the associates on the assembly line. In healthcare, man represents physicians, nurses, technicians and the rest of the staff who directly impact the patient. Physicians and nurses have independent licenses to practice, and their licensing boards hold them to standards above and beyond that of their jobs. Consequently, changes involving patient care are much more difficult to implement than changes around the assembly line. Furthermore, not all doctors and nurses have the same background and training although they may hold the same degrees. As a result, there is variability around the healthcare worker that is more difficult to address than that of a manufacturing worker.

Then there are methods. Toyota tries to standardize methods as much as possible. And while there is certainly room to improve standardization in healthcare, there is much variability that is hard to overcome. Not all orthopedic surgeons perform joint replacement surgery using the same technique or same prosthesis. Take weight loss. What is the best diet? What type of exercise and how much should one do? Should dietary supplements be used? When should bariatric surgery be used? The evidence is not always clear.

Regarding machine, Toyota believes that all equipment should be in good working order and standardized wherever possible. However, in most hospitals, equipment is replaced over time. As an example, it’s not uncommon for hospitals to have multiple types of stretchers. Sometimes the stretchers are from different manufacturers. Sometimes they’re simply different models from the same manufacturer. Nevertheless, controls and functions may differ. For instance, some have scales built in and others don’t. The standard workflow for weighing patients may vary depending on the stretcher used.
Lastly, lean attempts to produce “single piece flow” whenever possible. Batch processing is considered bad and should be avoided. However, in the emergency department, patients rarely arrive in an orderly manner. Frequently, they show up for triage in groups of five or 10 or maybe 20. And while this flow of patients is considered natural and may be predictable, the predictability is not always precise.
For instance, Mondays always tend to be busy days, but can you predict with certainty how many patients will arrive two weeks from now at 10 a.m.? We know that volumes rise during flu season, but the severity of the flu in a given season is never known in advance. Lean doesn’t provide the tools to deal with this variability.

Joseph Twanmoh MD, MBA, FACEP, FAAEM, is the Senior Vice President of the Mid-Atlantic Division of MS2, Providence, R.I. He is a Lean-certified physician with expertise in change management. Dr. Twanmoh has more than 20 years of leadership and management experience turning around emergency departments, redesigning hospital processes, and improving patient flow. Dr. Twanmoh previously served as a Medical Director, ED Chairman and Director for Health Systems Innovation for EmCare’s North Division. He is a graduate of Robert Wood Johnson Medical School and received his MBA from the Johns Hopkins University Carey School of Business. He is past president of the Maryland chapter of the American College of Emergency Physicians, serves on the Practice Committee for the American College of Emergency Physicians, and is a fellow of the American College of Emergency Physicians and the American Academy of Emergency Medicine.


Blog post currently doesn't have any comments.