11 Reinvention – Becoming the Change: Leadership Behavior Strategies for Continuous Improvement in Healthcare


The quality revolution is, of course, not the only brink that healthcare stands upon. Massive changes in population and resource distribution are also at the horizon. And the next global pandemic is, perhaps, just now stirring to life inside a bat cave in a little-known region of the world.

We do not know what is coming at us next, but we know that we must not let this massive disruption go to waste.

For many years, we in the medical field have acknowledged major problems and inequities in how healthcare is practiced. We have a system that requires 90-year-old heart patients to drive through snow for follow-up appointments because the clinic is paid by the visit. And health systems lost massive amounts of money during the pandemic, forcing them to lay off nurses, technicians, and physicians, because they could not perform elective surgeries and that’s how they get paid—by the procedure.

More recently, we saw that the fee-for-service model did not reward health systems for being prepared for a pandemic, and so most were not ready. The system was not set up to take care of the population as a whole, and so the whole was not cared for.

There were cases of outstanding response by health systems around the country. We saw hospitals with preexisting daily management systems use daily huddles to solve fast-looming problems. The work was inspiring. But it amounted to islands of excellence.

So, we are reevaluating the systems that failed us during the Covid-19 crisis, including supply chains, management processes, care delivery models, and payment systems. Does it make sense to have states and the federal government competing for scarce resources like personal protective equipment? Should all decisions be funneled through a preexisting chain of command? Is it necessary to have infectious patients report to hospitals and testing centers? Do we even need clinics if we can hand out biometric-collecting wrist straps that tell physicians all they need to know while looking at a patient on a screen? How deeply we address these issues will indicate, to some degree, how prepared we can be to meet the next crisis, whether global virus or regional disaster.

For this chapter, therefore, we will set aside incremental improvements in order to address innovation. Many of our assumptions about how to run a healthcare business are still based on a simple transactional concept. Hospital compensation has been based on “heads in beds” for more than 100 years. But we believe that the move to population health, already underway in some areas despite anemic interest from most traditional insurers, is both necessary and inevitable.

Health systems should be paid to keep people healthy and out of hospitals. If that’s the case, what will a health system look like? What will it look like to keep a whole population healthy, and how will we match resources to needs?


In order to prepare for fundamental change, healthcare organizations need to put energy and resources into the fourth value stream: development. This value stream encompasses the creation of new patient experiences. This is not incremental improvement. Here, we leave the traditional methods of care delivery in order to build new models to deliver better value to patients and new processes that can radically change patient outcomes.

Healthcare has spent the past decade or two focused on the first three major value streams: delivery, demand generation, and support. Leaders were not wrong to pursue changes in those areas first. We absolutely needed to change how we delivered care and supported operations.

Now, the time has come to include the development value stream in our work, to create the infrastructure that will foster creative responses to future needs. Technology, pharmaceutical, and manufacturing industries have led the way here, investing serious resources every year in research and development, creating new products and services at breakneck speed to stay competitive. Workers in these industries have been encouraged to take risks and topple the status quo.

The opposite has been true in healthcare delivery. While we have had major breakthroughs in surgical techniques and drug treatments, the actual process of delivering and receiving care has not changed much since World War II. Patients still go to hospitals and clinics, or wait on hold on the telephone, in hopes of accessing most of what they

In other industries, development work is referred to as innovation, but this is an overused term and too elastic for our purposes. Every way you turn, there are innovation conferences, innovation centers, books, podcasts, and gurus claiming to have the key to using innovation as creative disruption. When innovation experts arrive in healthcare organizations, however, they are usually stifled by the bureaucracy and do little more than erect a fancy idea board, focus on digital apps, or create an incubator for start-ups that may or may not have anything to do with the mission and future of the health system.

What we are talking about is a specific, repeatable process to create or amplify work in the development value stream. This is where teams engage in a fundamental rethinking of the business of delivering healthcare and produce results to be implemented. This process, which we are calling New Care Model Development, is best supported by a principle-driven health system. The bedrock principles here are creating customer value and seeking perfection. The team, and the larger organization, must value the personal traits of curiosity and perseverance to succeed.

Not just any skunk works will do. Simply finding the right geniuses, putting them together in a room, and giving them free rein will only produce ideas that work in that room. Some people claim that, in a lean environment, the best new strategies will be generated at the front line, in the course of the work. We disagree.

We believe that a robust development value stream needs to be separate from the daily work of healthcare. It should be focused on the vision of creating care models that continually improve upon patient experience, quality, and cost of care and led by a dedicated, multidisciplinary team of clinicians, operations specialists, engineers, and support staff—all of whom stand out for being willing to experiment in radical ways. This does not have to be a big team. We have seen effective teams of two people working with an outside coach, at the initial stage. What’s critical is that these people practice principle-driven behaviors and that they are allergic to maintaining the status quo or slightly improving on existing ideas.

These personalities can run against the grain of many people in healthcare. Caregivers were not rewarded in medical school for thinking in new ways. We were trained to revere our teachers, to follow best practices and recent research, to find incremental improvements. When looking for clinicians to populate a team of radical innovators, you will be looking for a risk-taker with good interpersonal skills who practices principle-driven behaviors. Deep listening skills are needed for this work.

Our efforts here are still new, but we have seen encouraging results. Before we describe the New Care Model Development process and its early results, we should be clear about the kind of work we are talking about here. This is not a slingshot for every rock.

Think of all the problems you and your organization encounter, and you can visualize all those problems falling into one of four buckets: simple, complicated, complex, and chaotic.1

   Simple problems have a root cause that is easily identifiable and usually have just one good solution.

   Complicated problems have a root cause—sometimes more than one—and have multiple good solutions. Think about addressing the issue of wrong-site surgery, for instance. A team will need to include some technical expertise to address a complicated problem, but repeatable solutions that work every time are still possible in this category.

   A complex problem has multiple variables with no predictable cause-and-effect relationship. This is where we use small-scale probing, prototyping, and experimentation to allow unique solutions to emerge from the system. Think of a complex problem such as moving a health system’s economy from fee-for-service to population health. If a team pondered all of the unknowables while searching for root cause before taking action, it would not be presenting solutions in this decade or, maybe, ever. If a team implemented sweeping changes all at once, on the other hand, it would introduce enormous risk for total system failure. So, teams are encouraged to nail a small boat together, so to speak, and see if it floats.

   The final problem type, chaotic, has many roots and courses of action and is usually fast moving. It is unpredictable. Think about a tsunami heading for a nuclear power plant, or a new coronavirus sweeping across the globe unchecked by immune resistance, and you have a chaos problem.

Health systems using lean thinking or Toyota Production System methods have often turned to 3P—production preparation process—to unleash creative solutions. And there is much to applaud in that model. While we were at ThedaCare, teams used these ideas and tools to completely remake the patient care path on hospital units into something we called Collaborative Care. It was breakthrough work that John highlighted in his first book, On the Mend. And Seattle Children’s mastered 3P while recreating the flow of outpatient surgery in its Bellevue Clinic.2

Some work, however, involves deeper questions than how to better coordinate and streamline patient care within a current environment. For complex problems we need to question every assumption about our current state. This is the work we will investigate here.


New Care Model Development is a multistep, team-based exploration into a complex problem and its possible solutions. Care models, specifically, are systems designed within regulatory boundaries that deliver value to a population, group, or patient cohort. Successful care models must also provide a good place to work for providers and staff and deliver sustainable business results. They are designed through the coordination of six elements: people, processes, equipment, locations, methods, and information.

New Care Model Development is time-intensive at the opening stage, often referred to as research or discovery. This is where the team needs to throw out everything they think they know and go out in search of answers to critical questions. What are the patient needs? What are the clinical requirements? What are the payment contracts? What is the current journey? What are competitors doing? What is the current state, really?

In this research phase, we value divergent thinking. We want the team to remain open to all possibilities. So we broaden the questions even more. What is happening on planet Earth? What are the political, economic, social, technological, and environmental factors that influence our organization and our patients, now or in future? What legal developments are changing the landscape? The team’s attention should be open to all factors.

Diversity of experience among team members and respectful communication are requirements here. Team members will be conducting deep interviews with stakeholders, compiling results, and diving down some pretty deep research rabbit holes in order to do the job. Frank and freewheeling discussions inside the team are expected but are only useful if they are respectful and driven by new facts unearthed in diligent investigations.

So far, given the complex nature of care models, three months has been a manageable lower limit for this initial phase. Beware of strict deadlines, however. The team needs breathing space in order to let their minds depart from today. After all, we are looking for ways to win in a radically new way.

This leads to the phase in which the team develops solution concepts based on what they learned in their exploratory research, pulling together ideas for radical leaps into the future. After the push to diversify knowledge and thinking, this is the phase where the team converges on a compelling vision for the future. (The team should expect to diverge and converge thinking in just about every phase.)

Developing concepts will often overlap with the prototype phase, where ideas are given physical form. Perfection is the enemy of bold experimentation here, and teams are encouraged to fail fast, fail cheap before investing large swaths of resources. This is where a team gets basic proof-of-concept for whether their vision of the future care model is desirable, technically feasible, financially viable, and has the clinical impact they envision. We have used ideas and tools from Design Thinking, agile, and Lean Startup3all popular in the technology sector—to encourage people to sprint toward real working prototypes.

This is the only place, we believe, in a continuous improvement environment where people are encouraged to jump to solutions. That does not mean we want bosses blurting answers. But we do want people thinking about and pursuing workable concepts throughout this process.

Once a team has proven their concepts through prototyping, members must build and test the care model, run it in a clinical environment, and evolve based on real needs and limitations. The team—which has also evolved by this point to include new members with new specialties—remains actively involved, collecting and analyzing data to gauge whether the new care model is meeting its targets.

At the end of building and testing the care model it should be evaluated for scaling. Go-to-market concepts should be incorporated into the enterprise strategic plan to engage broader operational planning and resources for the diffusion of the new model across the enterprise or new market.

The following case studies illustrate how the process works. These are not included to convince anyone that this is the one right way to accomplish innovation. Our work is too new for that kind of certitude. But we hope there is inspiration.

Seeking Innovation in Eldercare: Care in Place

Atrius Health—the Massachusetts multispecialty group profiled in Chapter 6—was already in a good position to reimagine care delivery when it formally launched an innovation initiative in 2015. About 50 percent of its population was operating under value-based payment contracts, so caregivers and staff were already thinking beyond fee-for-service 208models. And their team project—focused on creating a new care model for chronically ill, high-risk elderly patients—was led by a popular geriatrician who was motivated for change.

Eliza (Pippa) Shulman, DO, MPH, had seen the too-often disjointed nature of elder care for her entire career. A patient with multiple issues usually has multiple caregivers who are not always perfectly coordinated. A trip to the hospital can exacerbate the disconnect and lead to poor care transitions, loss of functional status, and risk of injury or insult from the hospitalization itself.

As team members discussed different approaches for their work, they quickly focused on the high level of hospital admissions for the fragile elderly that occur on nights and weekends—outside of regular office hours for a person’s primary doctor. Confusion and fear were too often part of the patient experience, as well as being a large portion of the total medical expense shouldered by organizations like Atrius

At the advice of a consultant, Adam Ward, Atrius Health leaders staffed a small team to begin researching, including George Higgins and John’s son, Ted Toussaint. We watched the experiment with interest from the sidelines.

The first 4 months of this 18-month project were spent mostly in research and investigation. George and Ted found chronically ill patients recently hospitalized for crisis intervention and recorded extensive interviews that they called ethnographies. In one interview, an elderly patient flatly stated, “I would rather die than go to the hospital.”

That statement became a fire lit under the team, driving them to push for ways they could get care to people outside of the hospital environment.

While leaders had been focused for many years on reducing 30-day readmissions in this population, initial research led them to think differently about the problem. Instead of reducing readmissions, they asked, “How can we prevent hospital admissions in the first place?”

Of course, this question was only appropriate for some patients. Emergencies like organ failure and sudden physical trauma require hospital-level care. But what about chronically ill patients—someone with congestive heart failure, for instance—who has an emerging crisis at 4 p.m. on a Friday and calls the doctor for advice? These people were almost always advised to go to a hospital.

Working with clinical staff, the team created questions for patients who called with troubling symptoms to determine if the patient needed face-to-face diagnosis and treatment that day. If patients were unable to come into the office, in the existing system, they would have been sent to an ED. Instead, the team envisioned a home-based urgent care service, where a nurse practitioner got in a car and went to the patient to see if that dreaded trip to the ED could be prevented.

This prototype was chosen by the team as the idea with the biggest potential impact, with simple implementation and minimal risk. At the end of their 15-week test in a single region, 37 patients had received a home-based urgent care visit, and 13 of those patients had avoided a trip to the ED—an improvement of about 35 percent. The new care model was dubbed Care in Place.

The results were enough to expand upon the idea and, in the implementation phase, Atrius Health spread Care in Place to 20 clinics in the first year. The team selected clinics based on which ones had more patients potentially in need of this new care model and then met with site leaders to communicate and train people on the needed changes for implementation. However, it quickly became clear that the existing Atrius Health clinic operations did not have enough staff to send nurse practitioners out at a moment’s notice. So, they created a partnership with the VNA Care Network in eastern Massachusetts, a visiting nurses’ association.

The nurses’ association was reluctant to try the new process at first, since it was outside of their standard operating procedures. But the team offered evidence from the Care in Place prototype, and the agency agreed to give it a try with the first 60 patients as a test. The test was successful, and Atrius expanded Care in Place to all of its remaining clinics.

In the first three years of this program, Atrius Health saw a 42 percent reduction in ED use for those who received the Care in Place service, and leaders estimated that it actually halved subsequent hospital admissions in that population. In 2017 and 2018, leaders estimated this saved $2.15 million in unnecessary hospital care. The Care in Place model continues to grow and evolve, and in 2019, leaders were planning to expand its operational hours by leveraging the expertise of paramedics affiliated with Atrius Health clinicians.

By the time Care in Place was scaled up and transitioned into operations, the Atrius Health Center for Innovation had expanded to two teams capable of tackling more than one big question at a time. As the team grew and people pushed for bigger ideas outside of their comfort zones, leadership principles and behaviors like humility, curiosity, and respect became even more important.

“Leader behaviors rooted in principles mitigate a number of ways that projects and teams can fail,” Ted says. “People need to feel safe before they can be motivated to push past what is considered possible today.”

Staying Out of a Hospital: Medically Home

Leaders and team members in the Center for Innovation deliberated on a next project to pursue. Still enthusiastic about preventing the pain patients can experience from going to the hospital, and in coordination with senior leaders, they decided to expand upon their original work and focus on home hospitalization.

The idea of creating functional hospitals inside patient homes was not new. But, as far as the team could discover, nobody had yet made it work sustainably. So, they were happy to meet a small start-up company called Medically Home that had experience experimenting with the model and was looking for an established health system to partner with on development.

Following a similar new care model development process over a multiyear period, and this time with an expanded team including the founders of Medically Home, the innovation team researched, prototyped, and piloted a new, fully scalable home-hospital care model.

The new model, also called Medically Home, combined technology and clinical expertise to provide hospital-level care in patients’ homes, including medications, meals, home health aides, and all other services provided by traditional hospitals. Caregivers staffed a medical command center 24/7 at Atrius Health while high-tech equipment and supplies were deployed to patients’ homes. Acute rapid response was available 24 hours per day.

Patients were referred to the program from ambulatory clinics, urgent care centers, or selected emergency departments. As a first step, all patients underwent clinical and social eligibility screening to see if their needs could be met outside a hospital. If patients cleared the screening, they were offered the option of hospitalization at home.

A clinician and a technician then set up equipment such as biometric monitoring and communications, completed a medication reconciliation, and assessed the patient in conjunction with a physician back at the medical command center. The care team included physicians, nurse practitioners, skilled nurses, therapists, and others, providing care that included intravenous therapy, remote vitals monitoring, basic diagnostics, a dedicated phone line, and a video communication setup.

Physicians, nurse practitioners, and nurses located in the medical command center would virtually round on patients, monitor vitals, and document routine and status changes in real time. Various caregivers also visited the patient at home to provide imaging services, meals, and personal care.

Patient feedback on the new model has been positive. The pilot conducted with 68 patients received perfect scores in both willingness to recommend the program and communication with doctors and nurses. As one patient expressed, “Every stage of this whole experience, everyone came and helped whenever I needed, for whatever I needed. I never felt alone in this.”

As of fall 2019, Medically Home had successfully completed over 100 in-home inpatient episodes. Episodes were, on average, 30 days in length from admission to discharge and involved two phases of care: acute and restorative. Patients in the acute phase are equivalent to hospital inpatients, receiving all the services outlined above. Patients in the restorative phase are those who would have been discharged to home or a skilled nursing facility. The focus during restorative phase shifts to prevention of further exacerbation, advance care planning, teaching, and addressing the social determinants of health such as transportation, food security, and the like. 213Patients receive comprehensive, patient-centered services tailored to their needs at an average savings of 15 to 30 percent when compared to equivalent hospital-based care.

So far, quality and safety outcomes have been equivalent or better than traditional hospital care. The pilot population had an overall 30-day readmission rate of 8.1 percent versus 22 percent in an equivalent population with normal hospital care. (Patients treated for acute exacerbation of heart failure had a 6 percent readmission rate.) While careful clinical screening seeks to ensure that patients are appropriate for this care model, 7.4 percent of patients needed to be escalated to a higher level of care, such as a hospital ICU.

Many patients served by this program were multimorbid and met criteria for palliative or hospice care. All patients had advance care planning and almost 20 percent were discharged to palliative or hospice services.

A key breakthrough in Medically Home was a scalable economic model enabled through creative staffing, centralization of physician resources, and new technology. The team estimates that this home-hospital model can replace 20 to 30 percent of all inpatient hospital stays, and it is designed to be financially viable at a national scale. While the original pilot was launched in Massachusetts, Medically Home has begun expanding to other parts of the United States including the Midwest and Pacific Northwest.

If scaled to its full potential, Medically Home could make a substantial contribution to the current crisis of national healthcare spending.


Building a New Care Model Development process requires some new thinking as well as new infrastructure. Ideas that underpin this work have been pulled from agile, Design Thinking, and lean product and process development, as well as from organizational excellence, so leaders must be prepared to incorporate new concepts. And since this work is about the strategic direction of the organization, the senior executive team will need to be front and center.

For instance, consider secrecy. Usually, we are proponents of transparency in healthcare operations. But this work actually requires protection from prying eyes. For this team to feel safe to play with wild ideas, rumor mills must be avoided. The last thing people on the front line need are speculative whispers going around about the closing of a hospital.

Also, a New Care Model Development team does not begin the work by studying current conditions in a target area. A clinician or engineer with decades of experience cannot help but bring their knowledge of processes into the room. And we value their expertise. But we are not studying current processes and inviting frontline opinions here. We are trying to imagine and create new cloth. So, we need to lay the groundwork.

What follows are the five initial steps an organization needs to take to prepare for a New Care Model Development team.

Create a Team

Selection criteria for this team must favor traits and behaviors over professional accomplishments. Team members should display humility and curiosity above all else. Also, they should be known for having an entrepreneurial spirit—the kind of people who are happy to try out new ideas or different methods and want to rope others into their efforts.

Youth can be an advantage, as well. A certain amount of neuroplasticity, at any rate, is required to imagine a world and human interactions that are beyond what is known.

The core team does not need to be large, but it should be dedicated to this work. There can also be half-time people on this team, or people with functional knowledge of IT, EHR,4 or finance who come in to consult. But the core team should be completely invested.

The leader will be a kind of chief engineer, with deep knowledge about product (type of care) and customers (patients). This role was filled by Pippa Shulman in the Care in Place example above. In that case, Atrius senior leaders knew they needed to rethink aspects of geriatric care, so they selected a clinical leader who was passionate about the field and dissatisfied with the care her patients received.

Pippa’s dissatisfaction with the existing care model was crucial to that team. She was willing to disrupt the status quo, even if it could create conflict with her peers, to help her patients. That is the kind of leader the team requires.

Dedicate Space

Technology companies have embraced the concept of the obeya,5 which was popularized at Toyota Motor Corp. The obeya is often referred to as a war room or a brain. It is a space with enough room for the team to work and to visually capture and display their process and progress. There should be enough wall space for the team to build the story of their concept and plan, giving them a way to introduce newcomers or functional consultants to the full story.

The space should be secure to avoid cross-contamination. Think of it as a negative pressure room. The outside organization should not be hearing about any early speculation or untested ideas. And the team should be protected from pressure and presumptions from the larger organization.

Plan the Process

As we outlined above, there are five essential steps in the process: research, concept, prototype, test, and evaluate for scale. There are many ways people codify design and development processes, and over time, many organizations develop their own, tailored to their specific culture and needs. Some organizations may have many of the right steps already in place but using different words. In that case, the above list exists as a check to ensure all elements are in play.

These steps are the arc of the work but are not always rigidly followed. Team members should respect the creative process and allow for overlapping of some steps and diversions at any point.

Establish Governance

Oversight of the New Care Model Development team should reside in a small committee that acts as both buffer and yoke between the team and senior executives. Communication is important here because nobody wants to get three months down the line on a concept just to have the CEO say no. On the other hand, the team should not be listening to a lot of opinions from senior executives, or anxiously trying to read their facial expressions for approval. Even if senior executives are requesting innovation in specific areas, the governance committee still acts as a buffer.

The governance committee will create clear definitions of what success and failure look like. Those definitions may very well change from one project to the next, but it is important to have clarity here so that the governance committee can devise its own decision filters to guide the work and team.

If there are barriers in the team’s way, the governance committee is there to run interference, as well as acting as an extra layer of security, protecting the team and its work from the rest of the organization. Because this team is solving tomorrow’s problems—not today’s—members of the governance committee may also be chief interpreters of the work.

Provide Funding

The team will need access to resources, of course, and it is useful to think of funding as three buckets.

In the research and discovery phase, the team should feel empowered to make small purchases. Nobody is buying solutions here. But if the team wants to experiment with pieces of technology or offer honorariums to patients in exchange for lengthy interviews, that should be available. This bucket might contain $5,000.

The second bucket is empty until approved by the special governance committee, and this is for costlier items needed to flesh out new care models. For instance, the team might need a nurse practitioner for a month to assist in developing and testing workflows for home visits. An organization should set aside something like $20,000 to $50,000 in anticipation of these expenses.

The third bucket is also a negotiation between the team and its governance committee. After creating a business plan for the new care model and pitching it to the organization, the team will be expected to present a budget to bring the idea to scale. So, the third bucket is like the wallet of a venture capitalist—empty until approval.

The future of healthcare, we believe, rests on the creation of development value streams. Covid-19 exposed major weaknesses and opportunities for us as caregivers. We have been needing new care models for years. But now, the lives of our grandchildren depend on our ability to innovate.

We can push ourselves past immediate fixes, go further than a cool new app. If we can think and act differently, if we can build repeatable process to foster creativity as outlined here, we can get there.


The other new piece of our work going forward is a continuation of this book. We all learn from the stories of others. We have seen this in the network and at every conference and summit where our colleagues across the world tell the stories of what they tried, what failed, and what became something greater than they imagined. True stories help illuminate our own paths.

Changing leadership behaviors is not a one-off project. Learning how to best communicate with others cannot be accomplished in a single workshop. Actually working at these behaviors and becoming better leaders is an ongoing practice that requires regular reflection and communication. Like any regular practice—spiritual, musical, athletic—our understanding of what we are doing deepens and changes over time.

So, we will be including regular updates of some of the stories you have read here, plus new case studies. We intend to roll these updates up into a second edition of this book in 2022.

We are deeply humbled by the work of the people highlighted in this book and the many others we have been so privileged to work with throughout the years. Our job is to continue to facilitate this remarkable community of committed learners as we learn, share, and connect with each other to make the world a better place. Our mission is no less than that. We hope this book can add to our goal to keep the conversation going.

1. With thanks to Dave Snowden and his Cynefin framework. For more information: www.cognitive-edge.com.

2. See Management on the Mend (ThedaCare Center for Healthcare Value, 2015), pages 42–47.

3. Eric Ries, The Lean Startup (Crown Business: 2011).

4. Electronic health records.

5. Obeya is Japanese for “big room” or “great room.”