At the time of our visit, Mayo was just starting to expand an advanced IT system including checks for harmful interactions of medications, order entry, improving communication among physicians, and gaining knowledge about outcomes so that the doctors could make better clinical decisions by designing guidelines and protocols. And even though mutual respect is a core Mayo value, some administrators feel like second-class citizens. Said one, “MDs act like they know everything. They act like everyone else is a dummy.” But others thought only some doctors acted this way and that some administrators invited this attitude. Overawed by the physicians, they were passive when they needed to push back. But all of this seems to me par for the course. Even in the best of organizations, there are complaints, especially about status and bruised egos.
The strengths and weaknesses of Mayo are intertwined, but overall, this is an excellent example of a collaborative craft model moving slowly to a knowledge mode. But it’s doing so without transformational leadership. William Mayo, the visionary leader created such a great organization that it would be hard for a new visionary leader to boost the clinic to another level. It will happen because the Mayo chairs have become convinced of the need for change.
Can Mayo be the model for other healthcare organizations?
Effective though it is, some of its limitations have resulted in its so far remaining unique and, furthermore, confined to particular niches. The most important limitation is that it remains grounded in academic health centers; it has not fully entered the bare-knuckle world of large-scale business competition. Attempts to expand into primary care centers have had mixed success and were abandoned in Arizona. Mayo is most effective where patient problems require complex customized solutions. Although physician interactions go beyond the individual-craft model, they remain quite personal and small-scale. Thus it seems likely that some of the lessons from the industrial model will need to be incorporated into the collaborative Mayo approach.
This fusion, however, is not a simple matter, as we’ll see in the section describing Mayo attempts to learn from Intermountain Health Care.
Intermountain Health Care
IHC, headquartered in Salt Lake City, has been repeatedly rated number one in the nation by independent rating agencies.24 IHC was formed in 1974–1975, when the Mormon Church decided to divest itself of its 14 hospitals. After buying 10 more hospitals and a number of physician practices, IHC comprised 24 hospitals, 26 health centers or neighborhood clinics, and 150 service sites throughout Utah, southern Idaho, and eastern Nevada. When we visited IHC, the system included 400 employed physicians and 1,500 who were directly affiliated. The majority of Utah physicians were impaneled by the IHC health plan. IHC had a 45–50 percent market share of its catchment area, of which 25 percent was paid directly to its system; the rest was contracted out to payers. IHC operated 50 percent of the hospital beds in Utah. The health plan covered 475,000 lives, with only a 6 percent rate of disenrollment after the first year.
IHC’s purpose is to improve the health of the population it serves. It spends millions in direct charity care of over a hundred thousand patients. Intermountain Community Care Foundation supports clinics for homeless and low-income populations with a clinic serving over three thousand children in seven schools. IHC’s policy is to keep premiums low to make quality care as affordable as possible.
IHC leadership can be viewed as a model for any organization of professionals trying to transform itself from a bureaucracy to a learning organization. IHC is moving from a traditional hospital-based bureaucratic system, with physicians essentially operating in a craft mode of production, to a knowledge mode that makes good use of information technology and quality tools to develop evidence-based health care. It’s moving from specialty silos to clinical programs, from treating specific illnesses to solutions that sustain health. To do this, it’s had an exceptional leadership partnership and a leadership system. At the top of the organization, Bill Nelson, the visionary CEO, has a background in finance, and Dr. Charles Sorenson, the emotionally intelligent and interactive chief medical officer, heads operations with the physicians.
At the program level, operational leaders—physicians working with nurses and administrators—work to persuade their colleagues to practice according to protocols (processes or pathways) that reduce variation, lower costs, and improve outcomes. These “doctor” (in the sense defined in chapter 6) leaders are helped in developing protocols and mapping processes by Drs. Brent James and David Burton, who provide the tools of quality management and informatics.
Our team interviewed seventeen leaders of the IHC system, culminating in a two-hour feedback session and discussion of the findings with Nelson and Sorenson. The people we interviewed all believe that IHC is moving in the right direction, up a learning curve. They appreciate Nelson’s leadership and his focus on quality and Sorenson’s interactive style of listening and responding to physician concerns.
A good example of an effective leadership partnership on the operational level of a clinical program comes from Dr. Don Lappé and Susan Goldberg, RN, who run the cardiovascular clinical program. Lappé has a weekly clinical conference at which data is presented on reasons to use a process or therapy. The doctors and nurses together set goals that are easy to measure. Lappé finds that physicians accept clinical pathways when they measurably improve patient care, so he provides timely, relevant, and easy-to-understand feedback. The data and methodology are placed on the IHC intranet.
Lappé is also taking on the task of cutting costs and limiting vendors to those who provide the best value. This means that some physicians will lose the free meals and vacations offered by vendors. However, the savings in the cost of stents, pacemakers, and intra aortic balloons are significant. Once the number of vendors is limited, it becomes possible to negotiate price reductions for quantity.
Lappé spends one-third of his time in clinical practice. He believes this is important if he is to maintain his credibility with physicians and, with the continual changes in practice, his knowledge of the field. Physician leaders say that once they stop practicing medicine, their colleagues no longer consider them real physicians. In a meeting on this report in Stockholm, Dr. Carola Lemne, CEO of Danderyd Hospital, remarked wryly that physicians will accept only other physicians as leaders, but once these doctors have become leaders, they’re no longer considered to be physicians. Or as one chair said of a colleague who became a hospital head, “He’s gone over to the dark side.”
Brent James affirmed the outline I showed him of the movement from craft to manufacturing to knowledge modes of production in health care: “We can show that the craft style actually harms patients. Caring concern can lead to the wrong solution.” For example, a surgeon defines quality in terms of a nurse following his orders. But if twenty surgeons give a nurse twenty different processes, she is likely to make a mistake. Working within standard processes reduces the possibility of confusion, and at the same time, leaves physicians free to exercise craftsmanship and to be caring.
Why hasn’t the movement to evidence-based medical practice not gone more quickly? According to James, “We are dealing with relatively new knowledge. Physicians are still arguing about the scientific facts.”25 He went on to say that the tools to manage quality of care are new. So is the mind-set needed—systems thinking, since 50 to 75 percent of errors are not human error, but system problems. Physicians need to understand the elements of total quality management: root cause analysis, and statistical process control of variability. Furthermore, it’s a mistake to try to impose pathways for all 610 common medical processes when 65 account for 95 percent of the variance. James also said that leaders are just starting to learn how to produce change, not by dealing one-on-one with doctors, but by gaining group consensus to protocols and persuading doctors of their value—changing mind-sets.
Another reason why hospitals are not driven to change, James noted, is that national surveys show that people choose providers on the basis of cost, con
venience, and friendly service, not quality. The outcome study in New York State on coronary artery bypass surgery had no effect on patient volumes in hospitals with good or bad results. Patients typically stay with bad doctors who develop a good relationship with them—when you are worried about your heart, you’re more likely to feel comforting transferential trust in the doctor you have who may not be the one you need.
Another important reason for slow change is disincentives for the providers. While protocols may lead to savings of 30 to 70 percent, the organization loses money because of reduced fees for service. James said: “You can only make a business case for quality if you can link quality to a payment strategy.” This can be done in a capitated health plan (the patient pays a fixed amount each year to cover all services) with little turnover or a partnership with a company that’s willing to invest in the long-term health of employees.
The Mayo-Intermountain Dialogue
On the surface, the Mayo Clinic and IHC organizations are very different. IHC is a large integrated delivery system with a health plan. To maintain its huge market share and not-for-profit status, it must demonstrate a commitment to “the best clinical practice” at “the lowest appropriate cost.” Its hospital charges are 15 percent lower than the national average. Mayo-Scottsdale, by contrast, is part of a unique academic health center in which research and teaching are vital to the mission. Compared to IHC, Mayo-Scottsdale is a medical boutique for episodic treatment; its charges are among the highest in the Phoenix area.
Despite their differences, these organizations appeared to have complementary competencies, and it seemed that that they could provide valuable lessons for each other. IHC is a national leader in evidence-based medicine, and Mayo wanted to learn how IHC leadership went about this. In turn, some of the IHC physicians had asked us how they could create more of a Mayo-type group practice, with a patient-focused cross-disciplinary culture.26 We thought that by bringing leaders of the two organizations together, they might learn from each other, and the CEOs of both organizations liked the idea.
What emerged from this encounter between two of the best healthcare organizations in the world is that they are both approaching knowledge modes, but from different directions. The IHC leadership came from the world of bureaucratic hospitals—they are comfortable with large scale and with practices like rule-based management and performance incentives—but sought to reinvent themselves in a way that would create greater involvement of physicians and staff. The Mayo tradition comes originally from the craft world and has remained generally within a smallgroup, personalized framework. Mayo and IHC were interested in each other because both see that their approaches need to be further developed to deal with complex interdependence on a large scale. The Mayo physicians are interested in and wary of formalized process, while the IHC leaders are interested in, but also wary of, encouraging flexible local teamwork in place of some of their top-down rules.
These differences emerged most clearly around two issues.
Evidence-based medicine: IHC leaders believe that EBM leads both to more consistent clinical excellence, reducing the rate of medical errors, and to maximum cost-effectiveness. In the dialogue, however, the Mayo doctors expressed resistance to EBM; they pointed out that they often deal with complex health problems that require cooperation across disciplines and do not fit standardized pathways. Although some Mayo chairs have begun to address systematic variability, the physicians as a whole are not yet convinced that the scientific literature is strong enough to provide consistent answers to many situations; they trust more in their peer interactions and mutual criticism focused on the patient to maximize quality. They will move to EBM only when their own research supports it. However, this will probably happen more quickly with IT systems that provide the medical staff with systematic information on both process and outcome performance.
The physician–administrator relationship: At IHC, administrators and physicians often clash: the former focus on controlling costs and quality, while the latter continue to try to maximize their individual professional autonomy within the framework of a large organization. At Mayo, administrators explicitly serve physicians. The positive element in this approach is that the focus on patient welfare is unclouded; the negative element is that there is little opportunity to create a unified set of processes that would standardize the best learnings throughout the hospital and lead to continuous improvement of the whole.
One way of putting the problem is that the Mayo approach of physician interaction and consensus works very effectively to bring together varied resources around a specific problem; on a case-by-case basis, it is unmatched. At Mayo, groups have relied on peer review and discussion to work out consistency of procedures. The problem is to scale this capability to a larger system where doctors can’t be expected to learn from each other informally.
IHC leaders seemed to be struggling to define an approach that reconciles the need for physician involvement and commitment with the need for procedural consistency and efficiency on a large scale. They were moving, with difficulty, toward a process in which medical protocols are developed through group consensus across the system. This may point the way to a resolution of the tension between collaboration based on casefocused discussion and more formalized processes.
Thus the two organizations show in different ways the difficulty of making the transition to a knowledge mode of production for the healthcare system as a whole. The Mayo example involves high levels of collaboration but on a small and personal scale, with much use of informal mechanisms; the IHC case involves large scale and an attempt to create greater involvement, but continues to struggle with the tensions between administrators and physicians. Both are trying to move beyond individual craft performance without getting stuck in an industrial mode.
As a result of the dialogue between these two systems, Sorenson continued in the role of “doctor,” using the gap survey to lead the dialogue with physicians, responding to legitimate concerns, but emphasizing the logic of change. Mayo-Scottsdale planned to study IHC’s approach and to design a version of EBM that fit its culture and values. In both places, the leaders recognized that they had to create collaborators, not followers, and this required them to persuade, using data and encouraging open discussion.
Visionary Leadership at Vanderbilt University Medical Center
What kind of leader can move traditional academic health centers with their craft baronies to a knowledge mode of production? Certainly, the ombudsman-type leader won’t motivate change. I asked Roger Bulger for a good example, and he recommended a visit to Vanderbilt University Medical Center (VUMC), where Dr. Harry Jacobson was attempting this kind of transformation.
Jacobson is a visionary leader and a successful business entrepreneur. But he faced some tough opposition as he attempted to transform VUMC medicine to the knowledge mode of production. He has put together a first-rate leadership system with one of the leading medical IT innovators in the world, Dr. William Stead, Director of the Informatics Center. When we visited, Dr. Paul V. Miles, the Chief Quality Officer and Dr. Robert Dittus, Director of Internal Medicine, were teaching evidence-based medicine. But many faculty members were skeptical.
Some faculty argued that Jacobson was trying to force them into practicing “cookbook medicine,” that his purpose was cutting costs, not doing what was best for patients. Others were wary that they’d be pushed into a group practice, like Mayo, where they’d become salaried and lose the right, as one orthopedic surgeon put it, to “eat what you kill”—meaning, get the payments from the patients you treat.
Jacobson recognized he needed a strategy for transformation. Education and dialogue were essential elements. But Jacobson knew he was better at strategizing and visioning than at managing. One big step he took was to hire an effective chief operating officer, Dr. Steve Gabbe, from the University of Washington, as dean. He also recruited a couple of physician ex-astronauts who helped teach faculty, nursing staff, and students/ residents
to understand systems-based practice. Given their backgrounds, these physicians had instant credibility. They adapted airline crew resource training (CRT), and they were quite successful in getting their students to see the role systems and teamwork play in the quality of care. And Jacobson also made use of his leadership power to incentivize physicians to tackle variability and to place supporters in department chairs.
THE LESSONS OF LEADERSHIP FOR KNOWLEDGE WORK
Leaders are essential to transform business bureaucracies and healthcare organizations to the knowledge mode of production. One type of leader alone can’t do this. A system of leaders is needed: visionary strategists who are generally productive narcissists, obsessive systematic operational leaders, and interactive network leaders.
These leaders will face resistance, rooted in threats to self-interest and the bureaucratic social character. But they’ll also find supporters among the Interactives. Resistances won’t be overcome by warm, fuzzy leadership. But narcissistic visionaries also won’t succeed by forcing knowledge workers to follow them. As one VUMC chair said, “If Harry forced people, they’d become disaffected and leave.” Furthermore, as the power of paternal transference wanes, the key is to educate interactively, creating collaborators by persuasion and the generous use of supportive data as was done at Intermountain. Leaders must be “doctors” before they can become “democrats.” They need to understand the people they lead so they can address resistance and recognize real support. But as doctors, leaders also need to use their powers of promotion and incentives to back up the message. Dr. Kenneth Kizer, who led the transformation of the VA system, said at a seminar I attended that some of the bureaucratic physicians and administrators never got the message: “I had candid discussions with them and facilitated their leaving.”27
The Leaders We Need, And What Makes Us Follow Page 15