Maulik Joshi, president of the Health Research & Educational Trust, interviews surgeon, professor and author Atul Gawande, recipient of the HRET 2011 TRUST Award. In Part 2 of the interview, Gawande discusses controlling health care costs while improving quality, and moving the science of health care back to the bedside.

Atul Gawande, M.D., M.P.H., a general and endocrine surgeon at Brigham and Women's Hospital and the Dana Farber Cancer Institute, both in Boston, has received wide acclaim for his research and writing. He has written three New York Times bestselling books, including The Checklist Manifesto: How to Get Things Right (Macmillan, 2009), and is a staff writer at The New Yorker. He is also an associate professor of surgery at Harvard Medical School and of health policy and management at the Harvard School of Public Health. In 2007 he became the director of the World Health Organization's global campaign to reduce surgical deaths.

Gawande will receive the HRET's 2011 TRUST Award on July 17 in San Diego during the Health Forum-AHA Leadership Summit.

In recent New Yorker articles, you looked at patients with the highest medical costs. What about applying the values of cost and quality to that population? How do you generalize that?

The puzzle we've had is how to control costs while managing and improving quality, not in isolated examples but at a scale large enough to actually lower costs for an entire community. It's a formidable goal, but our economic future as a country depends on our finding answers.

In my recent writing, I hoped to identify the priority situations where we should take those actions. As soon as anyone starts listing all the ways in which care goes imperfectly or even outright wrong in medicine, the list grows so long and daunting. It feels impossible to tackle. And compounding matters, for as much money as there objectively is in medicine, in the middle of our work we feel our resources are constrained. Trying to figure out where to place both energy and resources is hard to do.

The epiphany came for me when I met a young family physician from Camden, N.J., named Jeffrey Brenner. Like a number of pioneering doctors around the country, he started by looking carefully at the cost numbers in his community. He found that in his community, as in every community, a small percentage of patients accounted for the largest volume of costs.

In Camden, 1 percent of patients account for 30 percent of costs; 5 percent account for 60 percent of costs. One percent represents a manageable number of people. It's about 1,000 patients in Camden, or half the size of a family physician's practice. Suddenly, doing something significant about costs became a potentially manageable problem. Brenner started looking at what he could do to change the care for those patients.

He observed that patients with the highest costs often are getting the worst care. It simply can't be adequate care to have patients floating in and out of multiple emergency rooms and going through months'-long hospital stays. In Camden, many of the underlying health issues are caused by poverty and social problems, but Brenner nonetheless found effective ways to tackle them.

In other communities, the highest-cost patients may have inadequate care for terminal illness, severe mental illness or complex chronic disease. As hospitals shift to global payment, we are taking accountability for higher quality and lower costs for large populations. And, it seemed to me, we had to have a concrete place to start. Zeroing in on that top 1 percent or 5 percent of patients who have the highest costs and identifying the best practices to take care of them seemed the obvious, most logical thing to do.

Beginning to scale that work—going from taking care of 100 patients to thousands—physicians have no choice but to put systems of care in place. They have to devise their toolkits for the most common failures in care, the practices that are most successful in making patients healthier, keeping them out of hospitals and emergency rooms and thereby dramatically lowering their costs. These experiments compared with controls are cutting costs in these populations by more than 20 percent. When these populations represent the bulk of costs, it is a substantial gain for the system.

Talking about culture and using toolkits, how do you frame that with the pressure for efficiency and effectiveness?

More and more, cost problems are reflective of quality issues. In the bell curve of wide variations in cost, the places achieving the best results are not the most expensive in the system. They are often among the least expensive. There is a portfolio of things that a health system can do to begin making changes that improve care and measurably manage costs as well.

In every health system we have something that we are proud of doing to improve safety and quality and lower costs. A health system may have a major asthma project, a surgery checklist project or some other project they are very proud of. But usually we are "just hitting," to borrow a baseball analogy from Donald Berwick, administrator of the Centers for Medicare & Medicaid Services. It's like having a team with a fabulous hitter. That's great, but we need more than that. We need great fielding and pitching and scouting, too. We need a whole portfolio of capability to win—not one project but half a dozen major projects targeted toward our highest priority areas. If those projects focus on remedying our areas of greatest failures and greatest cost, then we start to emerge as a winning team.

In my mind, that list of projects almost invariably includes strengthening primary care for the chronically ill population in the top 5 percent of costs, and targeting catastrophic costs in the top 5 percent, which might be cancer or congestive heart failure care, depending on the system. And we almost invariably must have a project in managing better end-of-life discussions and planning for patients, which we now do very poorly.

Each system will have to identify and tackle its priority areas in the data—its areas of greatest failure for patient harm and for cost. If we do that, we can make our health care system demonstrably better from both points of view. We will be able to say we have a health care system locally that is better than it was three or five years before.

What thoughts do you have for the emerging health services researcher or clinical scholar who wants to make a major impact in health care delivery and financing?

It is fascinating watching how the science of health care is moving back to the bedside. A century ago, doctors like William Osler began trying to make medicine scientific by carefully observing and recording what their sick patients experienced. Because we didn't have effective treatments, the patients would die. Then the doctors did autopsies to see what was going on inside them and determine how that correlated with what they had observed. And disease by disease, they began classifying and enumerating the distinct ways in which our bodies go wrong.

As conditions were enumerated, we then moved from the bedside to the bench in order to come up with treatments. And during the second half of the 20th century, scientists came up with everything from transplant procedures to Tylenol. The armamentarium at the end of the century is 6,000 drugs and more than 4,000 medical and surgical procedures. And we're trying to deploy these across the country to every person alive. But we don't know how to do that very well. We don't have the evidence for how to design the most successful system. What's really powerful is using science to answer those questions.

So, it has come back to the bedside. It may be a checklist. But it may be far more sophisticated—information technologies, redesign of the flow of care. Health services research has been mostly diagnostics focused—that is, we tried to diagnose the system problems, the social disparities, the flaws in care and where errors appeared. But it is time now to move to the therapeutic side. Having found these patterns, what are the changes that can be made? What are the trials we can run to improve care demonstrably and not destroy our economic resources in the communities where health care is practiced?

My answer for where emerging researchers can have the greatest effect and ensure their work is applied: Start moving to the solutions side. It is an uncomfortable place. Systems are complex. Doing clinical bedside experiments and trials can be messy. Large system experiments are difficult to do, but they are desperately needed. It is where change will come from.

Editor's note: In Part 1 of the interview (June 28), Gawande discusses reducing surgical complications and deaths using a checklist and instilling a culture of humility, discipline and teamwork.

Maulik Joshi, Dr.P.H., is the president of HRET and senior vice president for research at the American Hospital Association. He is also a member of Speakers Express.