We found Bruce Flamm, a very accomplished obstetrician/gynecologist at Kaiser Permanente, and a couple of other people around the country to bring their expertise as faculty to our first project to reduce Cesarean section rates. Within a breathtakingly short amount of time, we had collaborative improvement going in over two dozen collaborating hospitals. In subsequent Breakthrough Series programs, with other groups of hospitals and clinics, we worked on waiting times, asthma, and cardiac surgery. Over time, the collaborative improvement model has become a mainstay globally. It’s used by many organizations that have made it their own. Today, I would bet that many people engaged in collaborative breakthrough projects around the work don’t know the model’s origins at IHI.
Evidence-Based Care Bundles — A bundle is a set of evidence-based interventions for a defined patient segment or population and care setting. Implementing a bundle reliably can produce significantly better outcomes than if the interventions were implemented separately.
Berwick: There’s a lot we don’t know to do in health care, but some things we do know. For example, strong scientific evidence suggests that a particular way to prepare a site for introduction of a central venous line prevents infections. There’s an evidence-based way to manage a urinary catheter. There’s a way to administer antibiotics safely.
In the past, most of the metrics and approaches to such evidence-based standard procedures looked at each component step separately. For example, when taking care of a diabetic patient, check the foot for ulcers or a pulse. Check the kidneys for function. Examine the retinas. Check. Check. Then, around the mid-1990s, an internist at HealthPartners in Minneapolis said, “Wait a minute. It should be all or none.” That is, if you’re supposed to check the eyes, check the feet, check the blood sugar, and check the kidneys of a diabetic patient, you don’t get partial credit for doing part of it. You’ve either done it all or not done it all. It was an all or none view of excellence. That makes sense technically, since sometimes you have to do the whole package to get the result. A television with 1,000 parts that is missing one crucial part is not 99.9 percent working. It’s a broken television set.
Carol Haraden was one of our experts on patient safety at the time. I think it was Carol — who was one of the greatest faculty IHI ever had — and some of her colleagues who began to develop the idea of defining the collection of things that should happen. Call it a bundle and think about implementation support and metrics attached to the bundle, rather than attached to the individual elements.
The Pursuing Perfection initiative was an eight-year demonstration program (2001 through 2008) whose goal was to learn if and how health care organizations could make dramatic improvements in performance across their organization.
Berwick: In 1999, the Institute of Medicine (IOM) published To Err Is Human. Two years later, the IOM issued Crossing the Quality Chasm. I had the good fortune to serve on the committee that wrote those reports. A few years after it first appeared, the Crossing the Quality Chasm report had become essentially the chartering document for the health care quality movement worldwide. It’s a comprehensive view of the nature of the defects we have in health care, the redesigns needed, and the theory behind those redesigns. It’s a systemic view of routes to excellence using modern quality science. It’s an astoundingly important report in my view, but it’s hard [to put into practice].
About two years after [To Err Is Human] appeared, the Robert Wood Johnson Foundation (RWJF) asked IHI if we could find organizations willing to take the Crossing the Quality Chasm report and make it the charter for organizational redesign. We called the project “Pursuing Perfection” because that was the level we were after. It was like all or none thinking at the organizational level. RWJF supported seven organizations in the US. Four in the UK, one in Sweden, and one in the Netherlands also joined on their own.
These organizations were serious. Their CEOs were committed. They had topflight, executive-level leadership on their teams. The first thing we learned was that, despite being ambitious, accomplished organizations, they didn’t understand improvement basics and didn’t have an approach to systemic improvement. We spent a year teaching the basics and then, one by one, the organizations began making tremendous progress. Nobody got all the way there, but out of it came leaders like Uma Kotagal at Cincinnati Children’s Hospital Medical Center and many others who became IHI faculty and global teachers in this field. It was a wonderful project.
Modeled after an electoral campaign, the 100,000 Lives Campaign was an 18-month initiative launched in December of 2004 that aimed to significantly reduce morbidity and mortality in American hospitals.
Berwick: My son, Dan, had worked on political campaigns for many years. I spent some time with him one weekend while he was working on a presidential campaign. I think he told me they’d mobilized people to knock on 60,000 doors in central Florida in one weekend. I said, “How the heck do you do that, Dan?”
He said there were five basic elements in a campaign: platform, field operations, fundraising, communications, and measurement. Dan talked about the need to focus on clear goals to get results. He was the one who first said to me, “Some is not a number. Soon is not a time” [which became the 100,000 Lives Campaign slogan]. In an election, “some” is not a number because the number is 50 percent plus one vote. “Soon” is not a time because the time is election day. Everything in a campaign is organized around that number and that time.
I remember coming back to the IHI office and saying, “I think we could do this.” We put together a platform of six changes focused on patient safety. We had a team of young people headed up by the brilliant Joe McCannon. Alexi Nazem headed our field operations. Madge Kaplan organized communications. Andy Hackbarth began a measurement program.
One of the key aspects of the 100,000 Lives Campaign was that we invited organizations to join us. You don’t order someone to be in a campaign. You invite them to join you. And the response was electric. I’ve never seen a bigger outpouring of abundance, of motivation, and interest as when we announced the start of the 100,000 Lives Campaign.
We also took a lot of heat. When we finally did our estimate of the Campaign’s results, some critics in the scientific community said it wasn’t well enough supported by the data and the results weren’t audited. While it’s true that we don’t know the exact number of lives saved, I’ll tell you the results culturally and in terms of energy toward improvement were astonishing.
The Triple Aim is framework for optimizing health system performance that simultaneously pursues three dimensions: improving the experience of individuals in care; improving the health of populations; and reducing the per capita cost of health care.
Berwick: In 2005, I was approached by two of our most important faculty of all time, John Whittington, a physician who was deeply involved with Pursuing Perfection, and Tom Nolan, probably the most significant science mentor I ever had in my life. Tom, Lloyd Provost, Ron Moen, Kevin Nolan, Jerry Langley, and Cliff Norman were colleagues at Associates in Process Improvement [developers of the Model for Improvement] and they provided the intellectual core of IHI’s scientific approaches to improvement over about a 20-year period.
Whittington and Nolan came to me and said, “We’re not thinking systemically enough about aims.” The Crossing the Quality Chasm report laid out aims for what health care should be: safe, effective, patient-centered, timely, efficient, and equitable. Those six aims had become canonical, but Whittington and Nolan said, “All of that is about treating the individual when they have a need. But where did the need come from? Why did they have a heart attack? Why are they depressed? Why did they break an arm? Why did they have a stroke?” They said we have to do more than just treat what goes wrong. We need to move upstream to address the causal system, which today we call social determinants of health. They said we need a network of three components: better care for individuals, better health for populations, and lower per capita cost. I thought then, and I think now, that the Triple Aim is the correct collection of goals or aims for our system. For health care, they define our reason for being.
Later, Tom and John gave me the honor of letting me join them in authorship of a paper in Health Affairs that introduced the Triple Aim and it caught fire. Today, you can go all over the world, and you’ll find the Triple Aim as the framework being used in organizations and even by nations. The credit goes to John Whittington and Tom Nolan, the authors of probably one of the most important frameworks IHI has ever produced.
The IHI Open School is a collection of dozens of online courses and over a thousand face-to-face chapters around the world that provides learning opportunities for well over 950,000 students, residents, and health professionals.
Berwick: Back in 1989 and 1990, the group of friends who were discovering Deming and working on the demonstration project [that eventually became IHI] developed a strategic plan to target four sectors: those who lead health care systems, the public at large, researchers, and teachers. We said, “If we change those four components of health care, we will begin to shift the system.” We made fast progress with people that provide health care. We didn’t have much success with getting the attention of the public. We had plenty of research going on, but medical schools and nursing schools were very slow to pick up on modern improvement science. Reasons? Their curricula are already full. The improvement sciences don’t have the classic pedigree. There’s no Nobel Prize in improvement science, and you don’t get to be professor in that field. You do in engineering, but not in health care.
We beat our head against not quite a stone wall, but a pretty hard obstacle for years. And then, in around 2008, it started to dawn on us that we could start our own virtual school. At that time, digital campaigning became the way that you had to organize for elections. So, we said, “Let’s start a school. If the medical schools won’t change, let’s start our own. In fact, let’s start our own nursing school, pharmacy school, engineering school, and management school!”
We were discussing this at that year’s IHI/BMJ International Forum on Quality and Safety in Healthcare in Paris with a bunch of students. We said, “What would happen if we organized a virtual school to study improvement science?” We had surveyed 1,700 professional students and asked them simple questions: Are you interested in improvement science? Are you learning it where you are now? Do you want to learn about improvement science by discipline or together with students from other disciplines? Overwhelmingly, 85 or 90 percent of respondents indicated that they wanted to learn more about improvement, they weren’t learning it in school, and they wanted to learn it with people from other disciplines, not in disciplinary silos.
The Open School took off like a rocket. Today, there are close to a million people in almost 100 countries enrolled in the Open School. And now IHI is entering a strategic era when I believe the Open School kind of architecture can be more than just a program. It could be the core of IHI. I think it represents the low-cost modernization and democratization of learning.
Editor’s note: This interview has been edited for length and clarity.
You may also be interested in:
White Paper — IHI Innovation System