Making Healthcare Better: Quality, Equity, and Purpose
Q&A with Don Berwick, MD, MPP (Part 1)
Donald M. "Don" Berwick is one of the leading figures in modern healthcare quality improvement and health policy. He is President Emeritus and Senior Fellow at the Institute for Healthcare Improvement (IHI), which he co-founded and led for two decades, and has also served as a Senior Fellow at the Center for American Progress and an International Visiting Fellow at The King's Fund in the UK. He began his career as a pediatrician at Harvard Community Health Plan before helping establish quality improvement and patient safety as central disciplines in healthcare at IHI, leading initiatives like the 100,000 Lives and 5 Million Lives Campaigns. From 2010 to 2011, he served as Administrator of the Centers for Medicare & Medicaid Services under the Obama administration. He holds an A.B. and M.D. from Harvard, along with an M.P.P. from the Kennedy School of Government, and has held academic appointments at Harvard Medical School and the Harvard T.H. Chan School of Public Health.
Dr. Berwick has authored or co-authored more than 160 scientific articles and several books on healthcare quality, policy, and safety, earning honors including the Heinz Award for Public Policy and the Institute of Medicine's Gustav O. Lienhard Award. In 2005, Queen Elizabeth II appointed him an Honorary Knight Commander of the Order of the British Empire for his work with the U.K.’s National Health Service.
Dr. Berwick’s work has repeatedly returned to a central question: How can healthcare systems move beyond delivering more care to producing better outcomes, greater equity, and better health? Across three conversations with Dr. Berwick, we widen the lens – from healthcare to health to the American Experiment – asking what we choose to improve, what shapes health, and ultimately, what the health of Americans reveals about the nation we are and aspire to become.
Martin Goldstein: Thank you, Dr. Berwick, for agreeing to this interview. We'll begin in 1989, when you co-founded the Institute for Healthcare Improvement (IHI). What was the specific moment or patient case that convinced you American healthcare needed to borrow from industrial quality improvement – and what was the nature of the resistance you faced?
Donald Berwick: There was no single moment that led to the forming of IHI, or to my own thinking. I had come out of an academic tradition at Harvard in Health Services Research, where, among other pursuits, I was involved in studying and developing metrics for health system performance. The RAND Health Insurance Experiment had occurred, and it had birthed a panoply of ways to measure performance. At the health maintenance organization where I was vice president for quality-of-care measurement, I measured – and it became increasingly frustrating. I would spend literally hundreds of thousands, if not millions, of dollars producing reports that ended up on shelves. Worse than that, the reports tended to anger my physician colleagues, who received data on their patient satisfaction levels or complication rates and couldn't do anything about it. I wasn't the most popular person on the block, because I was just making their lives harder. We would measure; we had a lot of metrics – but nothing changed. That all led to a decision to leave that role.
The chief executive of the organization, a man named Tom Pyle, who had come from other industries; his prior job, I think, was as an executive at Elizabeth Arden, when I went to him and said I wanted to quit, he said: "Don't quit. get better. We got to the moon. Products and services improve. Why don't you take some time and study how that's happened?" That began about six months of journeys to companies and organizations where I knew improvement had occurred – Bell Laboratories, which I knew had been the home of the engineering disciplines of improvement since the 1920s, Gillette, Sheraton Hotels, NASA. What I found was a gold mine. I just thought: these people really knew how to improve things.
A new friend introduced me to the work of W. Edwards Deming. I took Deming's four-day course in Washington. That course was probably the closest thing to an epiphany – I realized we had no theory, and Deming had a theory of how to make complex things better. That same friend and I convened a group of about seven or eight people who began meeting every couple of months to study together – reading books, getting together in Chicago, talking, and going back to our day jobs. That group became the founding board of IHI.
Goldstein: You've become associated with the axiom: every system is perfectly designed to get the results it gets. When did you first articulate that idea, and at the time, what system were you describing?
Berwick: I'm not the author of that quote. The same colleague who introduced me to Deming – Dr. Paul Batalden – used it in a speech at one of our early meetings, and it seemed exactly right. The idea is that every system – whether we're talking about the human body, a family, a corporation, or a manufacturing line – is perfectly designed to achieve the results it gets. It's a fundamental idea in modern improvement science: performance is a characteristic of the way a system is configured. It doesn't yield to exhortation, or wishful thinking, or hope. It yields to system redesign. If you don't like the result, you need a different system. The intellectual founders of that idea, I think, were systems theorists – Russell Ackoff, George Box and others who were the scientific architects of modern systems thinking and improvement theory.
Goldstein: Given that the principle is true and the evidence of 35 years suggests it is, what does it say about a system that consistently produces a 20-year life expectancy gap between wealthy and poor Americans? What is that system designed to do and if the design reflects choices rather than accidents, who has the power to make different choices?
Berwick: A result like longevity is the characteristic of a very large system of interactions. The first consequence of the axiom is that longevity will continue to be what it is unless the system that produces it is changed. But that also raises a prior question: what system are we talking about? There's another Deming insight I love – that aim creates a system, or more precisely, aim defines the system. Tell me what you're trying to achieve, and I can tell you the relevant system. In healthcare, we have a major problem: the actual system that produces health – the full set of elements and interactions that determine whether people are healthy – goes far beyond the formal healthcare delivery system, even though the formal system controls most of the resources, the habits, and the processes. It's a massive mismatch between what produces health and what we've built in the name of health. The second part of your question – who is responsible, or more positively, who could change it intentionally – is the more interesting one, and we'll return to it.
Goldstein: How would you define the health system as distinct from the healthcare delivery system?
Berwick: I have a sort of guru here – Sir Michael Marmot at University College London, a world-class epidemiologist and one of the great intellects in this field. He wrote a book in 2015 called The Health Gap, which directly addresses your question. He observed that in a city like Boston, there's something like a 20-year difference in life expectancy between wealthy and poor areas, and he asks: what's the dynamic? How does that work?
Marmot identified five determinants that go a long way toward answering that question: the conditions of children in their early years; the education system, and how talent is developed and supported; workplace conditions – worker rights, safety, minimum wages; the conditions of elders, especially with respect to loneliness and isolation; and a fifth category he calls conditions of community resilience, which includes, for example, housing security, food security, recreational resources, criminal justice, and environmental factors – what we commonly call the social determinants of health.
In his book, Marmot adds a sixth factor, which he calls the cause of the causes. His inquiry is: why do countries and cities differ in their investment in these five determinants? His answer is fairness – a social contract, or a belief structure in a community, about mutual responsibility and collective action. Countries with a belief system that embraces fairness are more likely to invest in improvements in those five categories. It's noteworthy that formal healthcare does not appear on that list. That’s probably a bit unfair – healthcare does do some things related to the determinants of health – but mostly it's what I call a repair shop, where we try to fix the damage done by inadequacies in the other determinants, see Lord Nigel Crisp's book, Health Is Made at Home, Hospitals Are for Repair.
Goldstein: Early in IHI's work, you focused on clinical quality – the 100,000 Lives Campaign, medication errors, hospital-acquired infections. You set a deadline and met it. Health equity wasn't explicitly central to that work, and you've reflected on that publicly. What's your assessment of why – and what it would have meant to do it differently from the start?
Berwick: I don't want to overstate what the campaign achieved. There was a lot of activity, and I'd estimate results in the tens of thousands of lives saved – but we deliberately chose not to invest in a very precise metric, because that precision would have chilled the energy of change. What doing it differently from the start would have meant is having additional goals echoing in equity. We would have been explicit about gaps in outcomes, health status, and safety by race and by wealth, and we would have put those on the dashboard, made them part of the field of concern. We didn't have race-specific, ethnic-specific, or wealth-specific goals, and we could have. The change packages in the 100,000 Lives Campaign were quite simple and very clear, related to very specific conditions – such as, here are the five things you do to prevent a ventilator-acquired pneumonia. But that relatively linear thinking, I don't think, would have addressed racial and socioeconomic gaps. Those gaps require much more interactive and systemic thinking. You'll only accomplish what you intend to accomplish – and we simply didn't put equity on the screen.
It's also a little bit of a tell that when I led the group that listed the six aims in what became the National Academy of Medicine's Crossing the Quality Chasm report – safe, effective, patient centered, timely, efficient, and equitable – equity was last on that list. We just didn't center it. The growing consciousness in America over the following decade, led by tragedies, has continually elevated equity as a dimension of excellence, and were that report to be written again, I think equity would be much closer to the top.
Goldstein: So the racial and economic composition of the lives saved wasn't something the campaign was designed to capture whether those saved were Black, brown, or poor wasn't on the screen, as you put it?
Berwick: That wasn't central on our screen at the time. There had been plenty of research, even preceding the campaign, about these inequities. But it wasn't in the zeitgeist, I might say, to put it front and center. Should have been – should have been. But it took a few more years of realization.
There's also a developmental dimension to this. I've recently given some talks on the phases of development of the improvement field, and back then we were still in what I would call Phase 1 – conceiving of improvement in project terms. You designate and name a challenge: reduce ventilator-acquired pneumonias. You organize a project team, develop concepts, test them serially, and make progress project by project. Joseph Juran, one of the great thinkers of the last century, actually wrote that all improvement happens through projects and in no other way. So there was a very strong belief that a project orientation was correct. The 100,000 Lives Campaign was essentially a list of six projects that seemed to have leverage. The second phase, which was already emerging then and has grown since, takes a much more systemic view of change – thinking of excellence as a property of an organization, not just a project, and beginning to reach outside the direct purview of the system toward social determinants. The third phase, which I call collective action, addresses challenges – equity, climate change – that no single organization could ever solve on its own. It's the problem of the commons, with a whole different set of dynamics and demands on leadership. The fourth phase – the one I'd say we're in now – is political. If you're going to work on collective action and improvement at scale, you've got to be involved in politics. There's no other way.
Goldstein: Given that equity wasn't centered in those early decades – that it was, as you put it, simply not on the screen – what does that tell us about what quality improvement, as a field, understood its purpose to be? And what did it leave out?
Berwick: I'm not sure I have a good answer, but here is how I think about it. Quality improvement mechanics, project-by-project improvement, like how to prevent pneumonia – are a technical set of tools that lie outside moral space. They're not moral declarations. The moral declaration lies in picking the aim. Once you've picked the aim, the techniques are very powerful – but you could use quality improvement to rob a bank well. The same process thinking would make you a great bank robber. The decision about what to improve is a reflection of moral, ethical, purpose driven commitments.
The gap in focus on equity – for centuries in this country – has been a moral problem. There is a kind of waking up to: Black people are starting off with a really poor set of cards, and is that my problem? And the answer is yes – it is my problem. There's a quote I use from Saint-Exupéry, the author of The Little Prince. His quote is gendered, but I think it means: to become an adult is to accept responsibility for the correction of problems you did not cause. That's a belief system. You could say, I didn't do that, it's not on me. Or you could say: this is a problem, I live in this place, I'm a member of a community, and it is my problem. That's not a quality improvement framing. That's a moral framing.
Goldstein: Let me come at this with some sobering facts. When IHI published its landmark white paper on achieving health equity in 2016 – 27 years after founding – the foreword, written by your successor Derek Feeley, opens with a striking admission: "As a leader of IHI, I have to admit to a frustration with our failure to help move the needle on health equity." The paper begins with the story of Tommy Cannon – a Black man in Perry County, Alabama, who waited hours in a segregated waiting room in 1973, was told to travel 50 miles to a hospital, and died the next day from a ruptured appendix without ever being seen by a physician. It documents that the life expectancy of Black Americans in 2010 was equal to that of white Americans in 1980 – a 30-year gap that persisted across the entire arc of IHI's founding and growth. And it arrives 15 years after the Institute of Medicine named equity as one of six essential aims in Crossing the Quality Chasm – 15 years during which it was still being called "the forgotten aim." We've also heard the story of Dr. Susan Moore – a physician who documented her own mistreatment in real time in 2020, and still couldn't get the care she needed. How do you make sense of that arc?
Berwick: The persistence of the gap.
Goldstein: Yes. From 1973 to 2020 – that is the gap.
Berwick: We are still a country experiencing the consequences of its racist background – and it's, in my view, the most important blot on the American landscape. It competes, perhaps, with our neglect of the tragic conditions of low-income countries around the world. But I think it all comes down to a fundamental question: are we responsible together, or not? And it's been hard to establish that we are.
The other dimension is that closing this tragic and unacceptable gap requires coordinated action among many, many players who don't work well together – who are not paid to cooperate, but paid to work separately. We are still a country in thrall to the Chicago School view that competition and profit are the drivers of excellence. But I don't believe you can ever achieve the kind of distributional equity we need through a market system. It just won't work, as long as we are hooked on markets rather than collective action. Under a pure market logic, the poor get the short end – and that seems to be viewed as acceptable.
Goldstein: If healthcare has become – as you've just described it – an engine for redistributing wealth from the poor to the rich, what would it take to reverse that? What does the alternative actually look like?
Berwick: It's about values, but also theory – social theory. Naomi Oreskes's and Eric Conway’s book The Big Myth is a very important text for me. Their argument is that the intellectual, evidentiary foundations of market theory are very weak when it comes to solving certain social problems. If you want better computers, markets and profit may be the right mechanism. But there are certain collective goods that markets will not improve on their own – at least not without a very strong governmental hand. We are addicted to market structures that don't work for healthcare. I don't think healthcare is a good place to enact market theory as the primary driver. It's a mutual responsibility that we need to discharge together, and the way we discharge responsibilities together is generally through government.
I think we set it up wrong. It's even worse now, because we have this unhinged, completely out of-control profiteering going on, in which healthcare is probably the biggest engine in America for the regressive redistribution of wealth from the poor to the rich. You are not going to solve these racial gaps in health if that is the dominant strategy.
Goldstein: So what would it take – practically – to change that?
Berwick: It's too complicated to give a soundbite answer, but forced to the wall – here is what I think is needed. First, a collective budget. There has to be a consolidated budget so that investment can flow toward the underlying determinants of this enormous gap in well-being by race and by wealth. That can't be done in a fee-for-service, widget-based payment system. We have to be able to invest where the investments would actually matter.
Second, consolidated leadership. If a hospital regards victory as increasing its top line or its bottom line, rather than closing racial gaps, that is what it will optimize for. There is currently nobody in our system for whom closing that racial gap constitutes success – not from the point of view of a board of trustees, not from the point of view of a Wall Street investor. We need to change that financial thinking into a holistic view of healthcare budgets. I would prefer to see the insurance system be a government insurance system. I don't believe the commercial insurance intermediaries are adding value of any substantial level.
Single payer is a step in that direction – though not a simple one. After all, I ran Medicare and Medicaid. I was a single payer for 110 million people, and it's not pretty. There's a lot of politicking going on, a lot of uncertainty. When elections happen, directions can change. But overall, I think it would produce a better result than an investor-owned, for-profit payment system.
We also need to enact activities across policy disciplines. The WHO calls it health in all policies. And following Marmot's work, you need a leadership table where all of the sectoral leaders are together. If you're a mayor, it means that when you're sitting with your cabinet – transportation, housing, environment, policing – there is a shared sense that this constitutes a team, with shared responsibility for policies that will affect the racial gap, altogether or not at all. We don't have that kind of leadership structure. But I think we could, at a municipal level. One thing I'm interested in now is whether we can find a few cities in this country willing to take the science and close that gap.
Goldstein: And underlying all of that – the budget, the leadership, the municipal commitment – you'd still have to get at Marmot's determinants themselves. The conditions of children, of workplaces, of elders, of communities. That's the actual work, isn't it?
Berwick: Yes. Marmot has developed the notion of what people call Marmot Places – the idea of asking: what would a city like Worcester, Massachusetts do if it wanted to take those determinants and the principle of fairness and actually design its civic activities around a collective view of the determinants of health? I think it's rather easy to describe, but sociologically very difficult to achieve.
Goldstein: Is that what you mean by the moral determinants of health – that it's ultimately a question of shared responsibility?
Berwick: That’s my belief – it's a shared responsibility. An interesting question to me is at what level of aggregation you can actually sustain that shared responsibility. In a family, perhaps. In a neighborhood? Could you do it in a city? What about a state? A nation? Except in wartime, we don't seem able to generate that kind of shared responsibility at the national level. Some states are better at it, and many cities are better at it – so finding the right level of action is an important and interesting challenge.
Goldstein: Are there cities you'd point to that are actually doing this – not just aspiring to it, but putting it into practice?
Berwick: There have been examples. Grand Junction, Colorado, actually declared universal access to healthcare as a civic norm and called on the corporations and other players in the city to come together and make it real. They did it for a while, though I believe it fell apart for reasons I'm not entirely clear on.
A more famous example was Rochester, New York, in the 1960s, where the then-dominant corporations – Xerox, Kodak, and others – came together with government and the private sector to guarantee primary care for everybody. They really redesigned primary care to achieve that aim. For whatever reason they did it – perhaps because a healthier city would produce more profits – the aim was clear: everyone here is going to have a doctor. End of story. And then they worked to make it a reality, and it was. It became quite well known as an example.
Goldstein: And what happened to it? Does it still exist in some form?
Berwick: These things are fragile. I was in Rochester a couple of years ago and tried to figure out what had happened. I don't remember the full story. But it's very hard to sustain durable collective action. The Nobel Prize-winning economist Elinor Ostrom – the first woman to win the Nobel in economics, from Indiana University – won her prize for work on essentially solving the problem of the commons that Garrett Hardin had described. To research for her book Governing the Commons, she scoured the world for examples of communities that had successfully managed shared resources – fisheries in Turkey, forestry in parts of Europe – that Hardin would have predicted would collapse. She interrogated what the common characteristics were and found eight recurrent features wherever the commons was successfully managed. I see a direct parallel to healthcare. If a community could think that way – about health as a shared resource, a commons – maybe we could begin to manage this beast that is otherwise consuming us.
Goldstein: You've just described Ostrom's work – communities that found a way to manage the commons successfully, with eight recurrent characteristics, and the possibility that healthcare could be thought of the same way. That's a powerful frame. So let me ask the founding question in that light: if you were starting IHI today – and perhaps that's a newly relevant question given that you've just departed – the values would be there from the start, equity at the center, not added later. But the harder question is what changes about the work itself: the methodology, the measures, the definition of which harms are treated as urgent enough to name, count, and act on by a specific date. What would that founding look like? And I'm interested in the smile.
Berwick: The smile is because I just had a conversation on exactly this topic in the last hour. My colleague and I were talking about whether we could recruit five American cities to a three-to five-year commitment to achieving massive improvements in health status through collective action of this type – using Marmot's action framework, with the participation of payers who would agree to change payment systems so that it would be affordable. I think the answer is yes. I can't wait to be invited to give a speech to the National League of Mayors and put out the challenge: we're looking for five cities. Just five. Five Marmot Places.
Somebody has to convene the activity. My bet would be a mayor, though mayors come and go. My best guess is that this would work at the municipal level, because there can be genuine civic cohesion – love for a place – that you can build on. Pride and an intrinsic sense of purpose. But you still need a stable organizer.
The best example I know right now is in Sweden, where a county I work closely, Jonkoping Region – has been at this for 20 to 25 years without giving up, even though five changes of executive leadership. They have a very strong sense of purpose. But that's a completely different economic and cultural context.
I also work a great deal with Wales. I'm on a national commission – the Bevan Commission – advising the Welsh government on healthcare, and just a couple of months ago they decided they are going to try to become a Marmot Nation. Three million people, their own devolved National Health Service, and that is the language they are now starting to use. I'm quite excited about it.
Goldstein: Before we close today, I want to ask you to hold a thought we'll return to in our next conversation. You've said that every system is perfectly designed to get the results it gets. You've also described a 27-year journey within IHI toward centering equity – and we've just sat with two stories, 50 years apart, that suggest the journey isn't over. So in your own words: what do you now believe the American healthcare system was designed to do – and who was it designed to serve?
Berwick: I think the answer has changed over the past decades. But the current answer is: the healthcare system is designed to preserve and expand the wealth of incumbent stakeholders – hospitals, insurance companies, pharmaceutical companies, medical guilds. It is an engine for the accumulation of wealth through the production of care, but not with its primary goal being the best possible care for individuals or the best possible health for communities.
Goldstein: We'll take that up in our next conversation.
Berwick: I reserve the right to change my mind – but that's what I think.
Goldstein: I'll hold you to that answer when we meet again. Dr. Berwick, this has been interesting and informative. It's been a privilege, and I look forward to the next two conversations.
About the Author:
Martin H. Goldstein is a 2020 Harvard ALI Fellow and Managing Editor for the Social Impact Review. Martin has thirty plus years in biopharma, initially at Hoffmann-La Roche, followed by Genentech. Subsequently, as the founder and CEO of ViroLogic, Inc., a clinical laboratory guiding therapy of HIV-infected patients. More recently, he has been working with venture to build biotech companies pursuing cutting-edge science to develop therapeutics to treat disease.
This Q&A has been edited for length and clarity.