This surgeon and Harvard scholar says better systems emerge when those who encounter problems firsthand design the improvements themselves.
After 20 years as a practicing vascular surgeon, John Kenagy went through an experience as a patient that helped redirect his career. Recovery from a broken neck that was sustained in a fall from a tree gave him a new perspective on the inefficiencies of the health care system. Time off to study at Harvard University, where he earned a master's degree in management in 1998, led to Kenagy's appointment as a visiting scholar at the Harvard Business School, a job he calls "one of the world's most prestigious nonpaying positions."
His research and consulting help health care organizations rethink how they deliver patient-focused services and products within complex, dynamic, highly competitive environments. At the root of his approach is the theory of disruptive innovation, which was developed by a Harvard colleague, Clayton Christensen, in his book, The Innovator's Dilemma: When New Technologies Cause Great Firms to Fail.
Kenagy earned his medical degree with distinction from the University of Nebraska and completed postgraduate training in general and vascular surgery at the University of Washington. In addition to his Harvard post, he is clinical associate professor of surgery at the University of Washington, adjunct associate professor of pharmacy and therapeutics at the University of Pittsburgh, and a Fellow of the American College of Surgeons. He spoke recently with Senior Contributing Editor Patrick Mullen.
MANAGED CARE: How did you move from practicing medicine as a vascular surgeon to helping health care organizations improve performance?
JOHN W. KENAGY, M.D., M.P.A.: As a physician, I've always been interested not only in taking care of patients, but also in how the health care delivery system worked. I've always known that it was necessary for a good clinician to problem-solve the system, and that the best nurses and doctors are those who can solve problems in the system in the best way for their patients. In 1992, when I fell out of a tree and broke my neck, I became part of the system from the other side. I was cared for by hundreds of smart and highly trained people who, in general, were compassionate and interested in taking care of me. They had the best equipment and were supported by health plans that wanted to do the best thing for me. What struck me, though, was that the sum of all those wonderful parts was much less than the total. I became interested in why putting all these pieces together has a negative effect. I assumed it was poor management or maybe nonexistent management, and I figured physicians and clinicians needed to be actively involved in managing our health system. As a surgeon, I worked with a great health system and really smart people, yet I had kind of the same feeling. We were all trying hard, we had the best technology, but I was not convinced that we'd really made our patients' lives better and, in some cases, I thought we might have made them worse. As I looked around the country, I didn't see anybody who seemed to have the best answer, so I decided to take a year out and study the problem.
MC: How did you define the problem?
KENAGY: Health care is so complex. We'd get all these smart people together to figure out how to manage complexity, but by the time we'd pull a lever, things had changed down below. The problem was not just the system's complexity, but also that it was a system that was constantly changing. So I made this connection at Harvard and decided to look at complex dynamic systems outside of health care. I wanted to understand how to manage complex dynamic systems and to identify the lessons we in health care can learn from other systems. I took a class from Clay [Christensen] at the business school. He described how small steel mills took over from integrated steel mills. I wrote a paper drawing a metaphor between small steel mills and outpatient surgery — labs and diagnostic facilities taking business away from general hospitals. He asked if I'd like to develop those ideas in health care. So I started working on Clay's concept of disruptive innovation, and that one-year stint has now extended into its fourth year.
MC: What is the concept of disruptive innovation and how does it relate to managing complex dynamic systems?
KENAGY: Clay was fascinated by the question of why the best firms and the best managers so often fail to manage paradigm-shifting innovations. We humans are naturally good at this — and some companies are great at sustaining innovation — but we can innovate faster than our customers are able to absorb those innovations. We overshoot their needs, which creates an opportunity for a different form of innovation, a disruptive one, to enter the market. Disruptive innovation is a strategy tool. It's a way to look at your ideas, and the environment in which your ideas live, through different lenses — so you can better tailor your ideas to the environment for greater success. Disruptive innovation starts with a worse product in a financially unattractive portion of the market but then steadily improves to the point where it becomes dominant.
MC: What would be an example?
KENAGY: IBM is a perfect example of how disruptive innovation works. IBM focused on making mainframe computers, while Digital Equipment made minicomputers and disrupted mainframe computers. Digital Equipment was recognized as one of the world's great companies, yet within a three-year period they collapsed. Why? They collapsed because they completely missed the personal computer coming into the market. Despite Digital's tremendous technical abilities, and the technical simplicity of a PC, they could never make that jump because they never took it out of their organization. IBM, however, survived. How did it survive? It created an entirely separate organization down in Florida, separated even physically from its parent, and gave that organization the franchise to disrupt IBM by developing PCs. It is characteristic of a disruptive innovation that the leaders can never make the jump to the disruptive innovation, ever. There is a zero possibility of that happening within the confines of their established organization.
KENAGY: The abilities of an established organization to do sustaining innovation become their disabilities to do a disruption. The biggest reason is that organizations are built to deliver sustaining innovations. The organization's values and processes — its DNA — are continually pushing it up one track. The problem is that an organization can't change those processes and values when a disruption comes along.
MC: Without putting new people in.
KENAGY: Not just new people. You need to create a new organization.
MC: How do the ideas of disruptive innovation apply to health care?
KENAGY: I think one disruptive innovation in care delivery can be found in the principles of the Toyota production system. These are new ideas that are just beginning to percolate out. The first article, "Decoding the DNA of the Toyota Production System," appeared in Harvard Business Review last fall, and it's not exactly on every doctor's reading list. I have talked to twenty or thirty organizations, but this is a very new thing.
MC: Explain the Toyota model, if you would, and then we can talk about how it applies to health care.
KENAGY: Toyota has been recognized as the world's best manufacturer for a long time. Its ability to continually improve its product, continually eliminate defects, continually eliminate costs, and continually become more responsive to its customers than anybody else is remarkable. That has made it the most studied manufacturing organization in the world, because everybody wants to figure out how Toyota does it. You could have a six-foot stack of books explaining the Toyota production system and how to replicate it. The consultants who wrote those books recognized all the marvelous tools that Toyota developed over the years to make better products, things like continuous quality improvement, quality circles, and rapid cycle improvement. Management ideas bubbled out of that organization, as did manufacturing ideas like just-in-time manufacturing and zero-inventory and zero-waste programs. People took those ideas back to their organizations and tried to copy them. They discovered that it didn't work. In areas where Toyota would get 30-percent improvement, other organizations would maybe get a 5-percent improvement. If they didn't hold their thumb on it, even those improvements would go away.
MC: What went wrong?
KENAGY: That was the interesting question. Why isn't this thing transferable? Toyota couldn't understand it. It would say that it didn't even know what the Toyota Production System was. Toyota officials knew when they were doing it and when they weren't. They knew it wasn't specific to Japan or to Toyota, because they could teach it to Americans and to their suppliers. That led to a collaborative research project between Kent Bowen and Steve Spear, researchers at Harvard Business School, and Toyota. Steve went to work for Toyota. Eventually, he worked for 35 different companies, on assembly lines, in training, in service areas. Of those 35 companies, 22 used the Toyota production system. He and Kent discovered that it wasn't the tools that made the difference. They saw the same tools used in different companies, but something was always different. At Toyota and Toyota's suppliers, it seemed that people who used the Toyota Production System were following rules, but these rules were not written down. There were no magic formulas or secret handshakes. Steve and Kent codified these unwritten rules, which had developed over 50 years and were part of the culture of Toyota, and then boiled them down to four principles. They concluded that these four principles were what generated all those wonderful tools and ideas.
MC: What are the principles?
KENAGY: They sound somewhat sterile when you recite them. The first rule talks about how the work of individuals is specified for the content, sequence, timing, and outcome of their activities. The second rule, for customer-supplier connections, or any other connections between two individuals, is that these connections are simple and direct, with specified ways to send requests and receive responses. Because two people can't do everything, you need multiple people, so the third rule governs the pathways over which complex goods or services are created. Where you have multiple activities and connections linked together, the rule is that you have to specify all the suppliers to be connected to the pathway, and you have to simplify the pathways for all goods and services. The fourth rule, the improvement rule, states that you do all improvement at the lowest possible level of the organization, which means that improvement happens where the work is done, under the direction of a specified teacher.
MC: How are you testing the theory to see how well it applies to health care?
KENAGY: We're doing an experiment sponsored by the department of pharmacy at the University of Pittsburgh Medical Center (UPMC). I want to emphasize that we don't claim to have an answer. This is not a silver bullet. We're doing an experiment to test a different way to manage health care. While the principles of this approach are the same between manufacturing and health care, health care is not a stamping plant or an assembly line. Methods that can be easily transferred within a manufacturing organization are not necessarily those that will make for good transferability within health care. It's a more complex process. You have the emotional attachment of people to their current organizations. Identifying those differences is one of the things that I'm working on.
MC: Describe the Pittsburgh experiment.
KENAGY: One company that has taken the ideas of disruptive innovation and used them is Alcoa, which is based in Pittsburgh.
MC: And was, until earlier this year, run by our new treasury secretary.
KENAGY: Exactly. Paul O'Neill is very interested in health care. He saw that it's a wreck and while at Alcoa was very interested in having Pittsburgh lead the world in doing health care differently. He was instrumental in bringing Kent, Steve, and me to Pittsburgh and creating a nonprofit organization, called the Pittsburgh Regional Health Care Initiative, to begin to try to see how these ideas could be inculcated into health care. From that original meeting the spark came to start these ideas in the University of Pittsburgh health system — particularly within the department of pharmacy, where Associate Dean for Research Randy Smith took a leadership role in making the experiment happen. Paul O'Neill gets credit for being the instigator and Randy Smith for putting the wheels on the experiment. They just believe in this stuff. At the start, we want to improve health care in Pittsburgh, but Pittsburgh is huge and complex, so we just take the University of Pittsburgh medical system, one piece of Pittsburgh that is a $5 billion system with 17 hospitals. How can we do experiments in that system? We started with Presbyterian Hospital and one community hospital, Southside. Presbyterian Hospital is a 600-bed academic medical center that administers four million doses a year. They want us to do medication administration. So we decided to start by just looking at medication administration for the liver transplant unit within Presbyterian Hospital and connecting it to the pharmacy. That's a classic way to approach these issues, by continually working down to smaller and smaller units until you find a place where you can start to work.
MC: What's the next step?
KENAGY: It's very much like the health care system. Before you treat, you have to diagnose and understand what's going on. To do that, you must actually go and observe. You don't get reports, you don't get process or procedure manuals, you don't ask people what they do. You physically go and observe. As we observed, we found that it's an unbelievably complex system. We found over 700 different ways to get a Tylenol #3 to a patient. The next step is to break it down even further and, as you understand how it works, there are two questions you can ask to help orient yourself from the Toyota system's view: What is the patient's specific need, and how does the organization meet that need in terms of activities, connections, and pathways? When you understand how it really happens, you start to make small improvements as an experiment under the guidance of a teacher to continually start to move health care closer to the ideal. The goal of this system is to be able to do that quickly. So you can have small experiments, immediately get a response, and then design the next experiment. Since there are a million problems, the problem you attack is the next one you face in the course of doing work. So we ask nurses to tell us the next time they don't have what they need to meet the needs of patients. We take that one highly specific problem, develop a root cause for it, and then design an experiment. We involve the caregivers in designing a new way to work that would solve that problem and in identifying the specific steps we will take to get to that targeted condition.
MC: How are physicians and nurses on the ward responding to your work?
KENAGY: They are incredulous. They've had so many flavors of the month come rolling down the pike. Yet what makes this attractive is that we're responding to specific problems. We're not attacking medication errors. In our concept, the medication error as a problem doesn't exist. It's as amorphous as world peace. We're attacking the thousands of things that happen that create medication errors. We respond directly to meet caregivers' needs, solve problems for them in the course of their work, and involve them in the solution. We're just learning how to do this. While people are attracted to these ideas, it's hard sometimes, because these ideas are new.
MC: Your goal is not to provide solutions, but to build a framework that will help the people who do the work create solutions.
KENAGY: That's it exactly. That's the genius of the Toyota system. The people who do the work are the people who design and redesign their own work. They do it in the course of work, with no meetings, committees, or task forces. It's part of work, so it's very different. We're discovering how this works in health care. It's tremendously promising to me, because it answers this complexity issue.
MC: What's your goal for the experiment at Presbyterian?
KENAGY: That we discover that this experiment works and also discover methods to make our learning unit transferable within the Presbyterian system and the UPMC system. The next step is to create transferability to other systems. We start with one learning unit, then one learning unit teaches two, two teach four, four teach sixteen. So you get logarithmic growth. Our goal is to foster rapid change that grows from the bottom up in an organic logarithmic fashion.
MC: What are some things that could trip you up?
KENAGY: The biggest challenge is that this is a different management system. It will require disruption of our present way of managing, because we have traditionally managed from the top down in health care. Health care adopted the notion of doctors telling patients what to do, and we've carried that approach along. We continually look for a higher authority to solve our problems, rather than realizing that the authority is within ourselves as individuals. We just haven't had a way for the people doing the work to make changes. The other challenge that we face is that there are tremendously deep silos in health care. This is such an interactive process that we have to bridge the gaps between those silos. One perfect example is the connection between the physician writing an order and the nurse who carries it out. Is it legible, free of defects, direct? We need physicians to experiment with that.
MC: Is the fact that most orders are still being written by hand on paper a problem?
KENAGY: It's actually an advantage, because paper is extremely flexible and computer systems are tremendously inflexible. One of our biggest challenges is computer systems that have been laid over the top of our systems by experts, smart people trying to do the best they can. These systems often don't match the needs of the people who use them, and they lack the flexibility needed to make changes. Disruptive innovation takes an entirely different approach to innovation and technology. Systems are pulled in to meet specific needs as your flexibility increases, rather than having an expert's solution imposed from the top.
MC: In health care, to what degree are the large established players — either hospitals or managed care companies — capable of making the changes necessary to survive? Or will they be replaced?
KENAGY: They are incapable of doing disruptive innovation. They have a zero percent chance because, to quote Clay, "The leader has never been able to make the jump within the confines of their present established organization." That's based on studying over 300 technological, service, and manufacturing organizations. So none of the leaders right now will be able to do disruptive innovation using the processes and values and resources that they presently use. That's the important part, that today's organizations can't do disruptive innovation within their present confines.
MC: So they can do it if they're willing to rethink their organization?
KENAGY: Exactly. It will take new partnerships, new organizations. It's a wonderful opportunity. What partnerships will make the new systems in the future? One tenet of disruptive innovation is that our capabilities become our disabilities. The capabilities of our organizations are our disabilities in terms of doing disruptions. If you can adopt that philosophy, it's tremendously liberating.
MC: Although I would think it's also pretty scary.
KENAGY: It's scary but also liberating. We don't have to defend the fact that we can't come up with the answer within this organization. We have an opportunity to build new relationships and partnerships to create new businesses. Yes, it may cannibalize our present organization, but we're part of the cannibalization process rather than being cannibalized from the onset. I hope that some of the leaders I've talked to from these organizations are beginning to use these strategy tools and ideas in their work. The ideas are nothing unless they're rooted in real life. They have to be tied to the problems of active and real organizations, but it's up to the organizations to do that. I can help teach them the principles and help make the transfer, but they have to take them and use them. I think it's going to create great opportunities.
MC: Thank you.