A Conversation with the Dartmouth Atlas Project’s Elliott Fisher, MD, MPH

Q&A: National Data, National Impact

We need to reduce unwarranted geographical variations in care

Elliott Fisher’s e-mail box is full these days, the result of communications with congressional staffers working on health care reform. Fisher and his team at the Dartmouth Atlas Project are intent on making sure the bills going through Congress help create an environment in which integrated care can begin to thrive. Rewarding providers for working together to reduce unnecessary care, which will improve quality and save money, is the key to saving the United States from fiscal catastrophe, says Fisher, who is professor of medicine and community and family medicine, and director of population health and policy at the Dartmouth Institute for Health Care Policy and Clinical Practice. The Dartmouth Atlas of Health Care, founded by Jack Wennberg, MD, has steadily documented the existence of geographical variations in care and demonstrated that more care does not equal better outcomes. Now, its creators are actively working to bring about change. Fisher is helping to lead an effort to establish accountable care organizations, which can be virtual or real provider partnerships that take on responsibility for the care of specific populations of patients. Reform may mean lower revenues overall for health plans, but those that focus on offering better quality at lower costs will come out on top, he says. Fisher earned his bachelor’s degree and his medical degree from Harvard University. He completed his master’s in public health at the University of Washington. He is a member of the Institute of Medicine and is cochairman of the National Quality Forum committee developing recommendations for measuring and improving the efficiency of health care. He spoke recently with MANAGED CARE Editor John Marcille.

Managed Care: You’ve written that slowing spending growth and rewarding providers for better care rather than more care would not require dramatic cuts in income. How are we saving money if everyone is making the same income?

Fisher: I didn’t say everyone. I said physicians and hospitals.

MC: But isn’t that where money goes in health care?

Fisher: That’s about 70 percent of it. Office practices are a huge part of the cost of care, so more efficient office practices would save money. There is a lot of unnecessary care being provided. Hospitals could downsize, and if we protected their margins but not their total revenues, we’d all be better off. Health plan revenue might fall into the category of areas we could save if we were more efficient. But frankly, we could solve the whole cost problem by simply reducing the growth in health care spending to the inflation rate. Even if we remain two percentage points above inflation, that would save $1 trillion to $2 trillion in health care costs and the incomes of most of those in health care would still be able to rise, just not as fast as they have been. Providers and payers could have a huge impact, if we are just willing to not grow as fast as we have been projected to grow. If we reduce unnecessary care, we can cover the uninsured and not have to increase health care spending to do it.

MC: That sounds great if we can accomplish it. Are the current reform discussions in Washington going in the right direction?

Fisher: There’s enough still on the table that we can get something done.

MC: You’ve been quoted in the press quite a bit recently, but people within health care have been paying attention to the Dartmouth Atlas for a very long time.

Fisher: Well, they have and they haven’t. Congress is just now realizing that health care costs actually matter, and that is making the interest in our work suddenly much greater. We are working very hard to try to make sure that all of the bills coming through will include support for the kinds of activities that we think are essential to laying the groundwork for the more fundamental delivery and payment system reform that will get us to value-based payment and integrated care.

MC: So today’s changes won’t be the end of it?

Fisher: We are not going to solve the problem in the short term, but we think it is critically important for the transformation of the American health care system to move toward integration and value-based payment, such as implementing really good performance measures that reassure the public that less can be more.

MC: How long have you been involved with the Dartmouth project?

Fisher: I came to Dartmouth in 1986 to join Jack Wennberg, and about five years later I started working with him to apply methods he developed in Vermont in the early 1970s to the U.S. health care system — that is, to look at the variations in practice and spending across the United States. Jack founded the Atlas project, and I was one of the collaborators from the beginning. About three years ago he stepped down. He continues to work on it, but I am running the project now.

MC: How did the Atlas evolve?

Fisher: Jack published his first study in 1973 in Science because none of the traditional health care journals would touch it. He studied the state of Vermont and found differences in surgical rates across communities and differences in spending. Because most of the care people receive is local, one could attribute practices to providers in that community. As health care reform started to get going, the Robert Wood Johnson Foundation asked our team to map the natural markets for health care services with the expectation that there would be health insurance purchasing cooperatives embedded in the Clinton health plan. When the Clinton health plan did not go through, we took advantage of having done the foundational work to start to compare practices and spending across the United States, much as Jack did in Vermont.

MC: How do you view the project’s effect on health care over time?

Fisher: This work has had a profound influence in many respects. It was the first to highlight the magnitude of the clinical uncertainty that many physicians face and the need for more research on the effectiveness and outcomes of different treatments. The Agency for Healthcare Research and Quality really was strengthened by our work on variations in practice. The second broad area was around the importance of patients’ preferences in treatment choices. Even when the risks and benefits of treatments are well understood, our early work showed that patients’ preferences for treatments vary dramatically because their attitudes toward those risks and benefits can differ. The third insight is around the causes of differences in spending and what we get for spending additional dollars. We found that most of the higher spending is based on how we use the hospital and how many specialists a patient sees, and we found that higher spending is not associated with better outcomes. These insights have been important for policy makers.

MC: How have health plans responded?

Fisher: More and more health plans are recognizing the importance of comparative effectiveness research and are supporting it; they have a major role to play in protecting patients from treatments for which there is no evidence. In the area of informed patient choice, many health plans are now implementing programs to ensure that patients receive balanced information about risks and benefits. Much more important than a second opinion is good opinion in the first place — and making sure patients are the ones who are making these decisions. Health plans, however, have not been very effective at dealing with the problems of rising costs because they have focused on these first two areas. They act as if this is about getting the right answer and prescribing the correct treatment and getting a medical director to approve everything. Having a chief medical officer review requests is not going to save money in the long run. The health system still rewards more care and lack of accountability on the part of providers.

MC: So one thing that private health plans should be looking at is payment reform and getting away from the fee-for-service model?

Fisher: Absolutely. We are not going to solve our problems until we move away from fee-for-service payment. Bundled payments are not going to get us there, because they still require figuring out whom you pay. We really need to move quickly toward systems of local provider groups, multispecialty group practices, independent practice associations, or hospitals and their associated physicians that can be held accountable for the overall cost of care.

MC: These are the accountable care organizations you have been talking about?

Fisher: Yes. We can be pretty agnostic about what organizational structures could support integration and be accountable for the care of a population. In some communities, it can be an independent practice association. Those are virtual organizations, and they are quite strong and thriving in many communities. In some communities, small community health plans may play the role of integrator of care, pulling together all of the physicians and hospitals to work together. In other communities, a hospital or academic medical center that employs its own physicians may well be the right model. But scale will be important. Many different kinds of structures will support accountability, but they need to be big enough that they enable us to look at the overall cost of care for their patients and measure outcomes.

MC: How will communities figure out which is the best way to approach this?

Fisher: The first thing is to begin a discussion about how hospitals and physicians can come together. We are seeing a lot of interest in these accountable care organization models. We will do best if we get several early pilots off the ground so communities can see an example of something that they could emulate.

MC: You have written about organizations that are already doing this well, such as Intermountain Healthcare and the Geisinger Health System. Do diverse examples of integration exist today, especially recent integration?

Fisher: Geisinger is a pretty good example of recent integration, and in California independent practice associations have come together effectively over the last 10 years. We have also seen some physician-hospital organizations and multispecialty group practices that are good models. Health plans can help them by offering a different set of rewards for measurably slowing the growth of costs.

MC: Can health plans do that and still maximize their profits?

Fisher: In the long run, health plans that are able to provide better care at a lower cost should thrive. But if, as many of us believe, overall health care spending could fall by 20 or 30 percent, the total amount of money to be made in the insurance industry may fall as well. My hypothesis is that the plans that will succeed will be those that are better able to support local integration and coordination and therefore reduce total costs. And the way they will win is by getting more members to their health plan, which is offering a better deal. Health plans may not all be standing at the end of the game.

MC: What is your advice to Medicare Advantage plans? Hasn’t most of the research been on Medicare patients?

Fisher: The Dartmouth Atlas is focused on the fee-for-service Medicare population because that’s where we have complete utilization data, but the insights from the Dartmouth Atlas are equally relevant to Medicare Advantage plans. Many of those plans are fee-for-service plans, so they will be subject to the same incentive structure that fee-for-service is, where if you reduce the price to primary care physicians, you’ll get more visits. Medicare Advantage plans will also benefit from thinking about how to foster accountability at the physician group or the physician-hospital level for the overall costs of care. We are trying to get private payers, including Medicare Advantage plans, into the accountable care organization pilots.

MC: How many pilots do you have going?

Fisher: We are working with four sites very closely, and we will have two launched by the end of this year. We have another 50 sites across the country that have expressed an interest, and 20 have joined a learning collaborative designed to get sites ready to go in the next year or two. We want to get the pilot programs up and running so we can kick the tires and get the details about how to do this wisely.

MC: Where is the Atlas project headed?

Fisher: One of our goals is to gain further insights into the causes of variation in practice across hospitals and physicians, so we have a major project observing physicians in different communities. The second broad area of our work is to try to improve the measures of performance that we make available, and we will be releasing some reports over the next six months that will provide detailed information at the physician-hospital network level around how their performance compares to others in their communities. We’ve done this for end-of-life care and chronic disease, but we will be doing it for ambulatory care. I think that will be a pretty important step forward in helping physicians understand how they are practicing in their communities and what is happening elsewhere in the country.

MC: Is this different from report card efforts that are more local?

Fisher: We will try to provide reports that allow everyone in the country to compare themselves to the national average and to others in their community. Most of our focus is on the epidemiology of clinical practice, so interpreting good quality or bad quality is less important than highlighting how different the practice is in one place compared to another. That provides an opportunity for physicians to start asking the question, “Why are we different?” We may not know which rate is right, but it may highlight either a problem of scientific uncertainty, a problem of inadequate information being provided to patients, or a problem of a physician group doing an awful lot of a particular procedure compared to everyone else, but because no one was really aware of it, the problem was invisible to them.

MC: How do you view the role of the advanced medical home? Is this a major or a minor development?

Fisher: We need to strengthen primary care to reduce the cost of health care and improve care for patients, especially those with chronic illness. The advanced medical home can markedly improve quality of care. The pilots are starting to show that in those systems, patients are more satisfied and get better care and the physicians are more satisfied. The question of whether the medical home on its own can reduce the cost of care is less clear-cut.

MC: You stopped seeing patients about five years ago, but when you did practice as an internist, did your research activities affect your practice, and did your practice affect the research?

Fisher: I could never have done the research I carried out without having been a practicing clinician. It was tremendously helpful to understand how difficult decisions are in routine clinical practice, how much judgment is required. One of the great myths that people are struggling with is that there is a right and a wrong answer. Health plans ask whether a procedure is appropriate or inappropriate. But that evidence rarely exists. Only a tiny fraction of what is being done in medicine is inappropriate. We have to turn the question around and ask it in a different way: “Is this care absolutely necessary or might it have been done in a lower-cost way that was better for the patient?” If you ask that, you immediately find that a lot of health care in the United States is unnecessary.

Health plans may not all be standing by the end of the game.