How poverty affects cost and outcomes
We need a new road map for health care reform,” says Richard “Buz” Cooper, MD, professor of medicine and senior fellow at the Leonard Davis Institute of Health Economics at the University of Pennsylvania. Some of the ideas driving legislation — that more health care spending does not yield better outcomes and that primary care physicians achieve better outcomes at lower cost, for example — distort the real picture, and some of the solutions being proposed are unachievable because reform advocates are failing to grasp the fundamental dynamics that drive the system, he says.
Critical issues such as the pervasive effects of poverty and the growing physician shortages are losing out. “Health care reform is a work in progress, and everything should be open to question to try to search for the best answers,” he says. Cooper, originally trained as a hematologist and an oncologist, conducts research on physician supply and demand. He is a co-author of the recent “Physicians and Their Practices Under Health Care Reform: A Report to the President and the Congress,” sponsored by the Physicians Foundation and available at http:// www.physiciansfoundations.org/.
Cooper has served on the faculty at Harvard and the University of Pennsylvania and was dean of the Medical College of Wisconsin, where he also founded the school’s Health Policy Institute. He received a bachelor’s degree from the University of Wisconsin–Madison, earned his medical degree at Washington University in St. Louis, and completed his residency and a fellowship in hematology at Harvard. Many of his ideas can be found on his blog, http://buzcooper.com. He spoke recently with MANAGED CARE Editor John Marcille.
MANAGED CARE: You have been publishing your arguments against some of the assumptions being made in health care reform legislation. Are people listening?
RICHARD “BUZ” COOPER, MD: I believe they are. For example, I recently wrote to Speaker Pelosi because a group of academic leaders brought these issues to her attention. Almost 60 congressmen have done the same. So someone is beginning to wonder if the framework of geographic variation that they are all operating under, which has been pushed by Peter Orszag [director of the Office of Management and Budget] since the beginning, is correct. Is it because of me? I have no idea. But there is no question that more and more people both inside and outside of Congress are beginning to behave differently, and many are wondering if we are doing the right thing.
MC: Are you acting as an advocate for physicians?
COOPER: I’m just an academic, as are my co-authors. We wrote the Physicians Foundation report based on research that was published previously. We just tried to examine it in the context of health reform. We are actually not lobbying for anything. We are trying to describe how health care works and how one might approach reforming it. There is nothing particularly prescriptive in the report.
MC: What is wrong with the proposals being discussed in Washington?
COOPER: One thing that’s wrong is that they’re based on measuring health care expenditures and quality using Medicare as the yardstick. If you take all the counties in America and look at which have the lowest Medicare spending per enrollee, and then reward those that are low or penalize those who are high, you are no longer looking at the performance of providers, you are looking at tremendous differences in patient characteristics, and most of those relate to poverty.
MC: What would be a better way to look at this?
COOPER: Total expenditures, not just Medicare, should be measured. The outcomes for individuals don’t relate strongly to their insurance status. If two patients are in the same room, one of whom is uninsured and the other of whom has great employer-sponsored insurance, and they both have congestive heart failure, the treatment is the same and the cost of care is the same. The payment to the physician or to the institution will be different because it depends on what is being paid on behalf of this particular patient, but physicians don’t change what they do in caring for patients based on the patient’s insurance status. On the other hand, care may change a lot based on the aggregate insurance status of everyone being cared for because that determines the aggregate revenue. So a hospital that has mostly patients who are well insured will have sufficient funds to hire a very good nursing staff, have high nurse-to-bed ratios, attract a broad range of physicians, and so forth. At the other extreme, if a hospital’s patients are mainly on Medicare or Medicaid or uninsured, at the end of the day when you add up all the revenue, there’s less the care system has to operate under. Those are the realities of how one practices. The ultimate result depends on aggregate revenues. That’s how the system actually works.
MC: How does that relate to outcomes?
COOPER: It relates directly to outcomes. Studies by Linda Aiken at our institution have shown that nurse-staff ratios make a huge difference. The educational level of nurses makes a difference, and baccalaureate nurses tend to be attracted to better-funded institutions. Obviously, there are shades of gray here, but across the spectrum, hospitals ultimately have to deal with total funds available. Medicare spending alone does not correlate with clinical outcomes. For example, we hear about Minnesota all the time. Well, Minnesota is a lower-cost state for Medicare beneficiaries, but it is in the top 25 percent of states for total health care spending per capita, and outcomes are good, as would be expected. You spend more and you get more.
MC: I wouldn’t have expected that difference.
COOPER: Minnesota is able to spend more on average because very few people in Minnesota are uninsured; they also have good Medicaid coverage, and the companies in Minnesota are strong employers with strong benefits. While Medicare spending is a bit lower, spending for everyone else is higher. But take a state like Nevada, which has high Medicare spending per enrollee but has among the lowest levels of overall per capita spending in the country and the fewest number of health care workers per capita of any state in the union. So if you were going to implement some standard that says we are going to penalize the high-spending states based on Medicare payments, that would destroy a state like Nevada, while Minnesota would benefit. And remember, even with Medicare, the government only pays about half. We hear all of this talk now about how the Mayo Clinic is this wonderful “do it cheap” place, and that is based on Medicare spending. But nobody knows how much the Mayo Clinic collects per Medicare patient. All we know is what Medicare pays. With few poor patients, Mayo Clinic is able to go to the other side of the ledger and get the rest paid by someone else, a private insurer or the patient. But even if a hospital in Nevada or Texas that is full of poor people gets paid a bit more by Medicare, it receives less overall. Then what happens is the Mayo Clinic patients go home and take care of themselves because they know how to. They have families; they have prescriptions. They have everything they need. But in places where there is a great deal of poverty, a patient goes home and may or may not have family support systems and may or may not have prescription drug coverage that is adequate, and may or may not have the education required to deal with the complexities of getting well. And two weeks later, that patient is right back in the hospital with a catheter infection or any number of things that we know go wrong all the time, and the costs that occur yearly for that patient are higher. Put the Mayo Clinic in Birmingham, Alabama, and let’s see how well it does.
MC: How should poverty affect the way we study geographical variance?
COOPER: People always say that we spend more and get less when compared to the rest of the world. Well, the U.S. is a nation of nations. We have a lot of spending in the North, and lousy outcomes in the South. When you average the two, the South is bad enough that it drags down our mortality average and the North spends enough that it drags up our spending average. So poverty is what gives America a bad rap in terms of health care spending and outcomes. We actually divided the country into six regions and looked at this. The mortality rates varied across the regions, largely in proportion to health care spending but also in proportion to how much poverty is in the area. The more poverty, the higher the mortality rate irrespective of spending, but adjusting for poverty, the higher the spending, the better the outcomes.
MC: Is the answer to fund poverty-stricken states and cities in a different way?
COOPER: Obviously the best thing to do is to get rid of poverty. But let’s look at it from the hospital’s perspective. Imposing a requirement upon hospitals that they reduce their rate of readmissions is like an unfunded mandate. I would say: Set standards for a hospital to reach, but fund the mandate. Allow hospitals that have a lot of foreign-speaking patients to hire interpreters. Allow hospitals that have a lot of poor people to create the associated systems that they need, such as for home care. The health care system is geared toward people like you and me who are smart enough to deal with it, who speak English, and who have families. But we are taking care of millions of people who, by those measures, are not like you and me, and they just need other stuff. You can’t penalize hospitals for that. You should reward them if they can get costs down, but you don’t want them to get costs down by not giving care. You want to fund the process. If President Obama stood up and said, “One of our biggest problems is that the poorest portion of our society costs double the average to take care of, and we need solutions,” people would get out and see what could be done. Like energy. Energy’s a big problem, and you’ve got all these little companies now trying to figure out how to provide energy cheaper and make efficient cars and everything else. So my answer is, identify poverty as an issue, create goals, and provide the necessary funding to accomplish those goals.
MC: What are some examples of the programs that could be tried?
COOPER: My daughter is a medical interpreter. Ask her how much money can be saved with proper medical interpreting. That alone would pay for itself 25 times over. Mary Naylor, a professor of nursing at our school, has studied home care programs for the poor elderly in Philadelphia and found that they cut down expenses enormously. There are lots of things to do. Don’t put pressure on institutions to do it cheaply if they are doing it more expensively right now because they are dealing with poor people who get readmitted or who stay an extra day because they can’t go home. You can’t discharge patients if there’s no one at home to take care of them, and you can’t fault the hospital. Create money for such programs and people will come out of the woodwork figuring out what to do. Make that the goal instead of assuming that providers are spending this money because they are ripping off the system. Physicians are not the bad guys, as the health care reform rhetoric would have you believe. The people who work in hospitals are not bad guys. The folks in managed care companies are not bad guys. They are your neighbors, your children, your wives and husbands. They are just people trying to do the right thing. That’s how the system works. Help them. When they are caring for the poorest patients, don’t bash them for failing to do things as cheaply as they could if they were caring for more affluent patients.
MC: Two of the areas you have questioned are the emphasis on primary care and the reliance on the Dartmouth Atlas studies. Can you discuss primary care vs. specialty care and the work of Professor Barbara Starfield at Johns Hopkins?
COOPER: I respect Barbara. Her international work is outstanding. But her work in the U.S. is based on comparing the geographic distribution of family practitioners and mortality rates. It is biased by the fact that there are a lot of family practitioners in the upper Midwest and all along the Canadian border, but fewer in Southern New England and along the East Coast or in the South. So family practitioners are concentrated where there are big, wide open spaces, no big urban centers, no complex urban issues and few minorities. That is family-practice land, and everybody knows that in the land of sky blue water, mortality is less. It’s not less because of family practice. It’s less because where family practitioners are located is a lot different from the South Bronx. And the correlation only applies to family practice, not to general internists who give very similar care. That’s because they’re in the large urban centers, with lots of poverty and racial diversity and higher mortality. The whole thing is a statistical artifact. Even the family practice comparisons don’t compute. You can’t expect a family practitioner to have a major influence on mortality from heart disease, stroke, and cancer. It is unrealistic. The primary care community is an important part of American medicine. Barbara’s studies simply don’t prove that. But that doesn’t mean there’s not a problem. There is. Primary care physicians are not compensated correctly. We have to deal with what they currently do, what they should do, and what the payment should be. The system needs primary care physicians — really, generalist physicians — to expand their panels. But to accomplish that we have to bend the compensation curve to more adequately reward generalist physicians for this kind of high complexity. Generalists deserve more compensation for an extended visit with a complex patient than with a less complex one. All of the extended visits by geriatricians are complex. They can’t make a living with the current compensation system. If we’re really going to address chronic illness, complexity has to be rewarded. That’s what we should be talking about.
MC: And the problem with the Dartmouth Atlas studies is that they are based on Medicare?
COOPER: That’s only part of the problem. Many people are critical of their work. It starts with their use of Medicare data, but their whole methodology is invalid. We’ve insisted that poverty is a major factor, but they minimize the effect of poverty, which is amazing. Even using their own data, it can easily be calculated that if all Medicare enrollees cost as much per year as those with incomes above $25,000, Medicare would spend 34 percent less. It’s a huge difference. And one thing we know about geographic variation is that poverty is very geographic.
MC: It’s important to question these studies, then?
COOPER: My argument with Elliott Fisher and his associates at the Dartmouth Atlas and with Barbara Starfield is really with their doctrinaire view that this is the world and you better believe it. I’ve been a physician for almost 50 years, and the health care system that Barbara and Elliott describe is not one I’ve ever worked in. There’s a lot to do in America, and I think the readers of your magazine are trying to do it. They are looking for a road map. They want to do the right thing. They want to decrease the rate of spending growth and they are searching for how to do it. What I am trying to say is, here’s how it works. When you make decisions on what to do, base it on how the system really works.
MC: You also have reservations about pay for performance?
COOPER: I don’t think there’s an argument with the principle — who would ever argue with the principle? — but the argument is, how does one measure performance? I am not against the notion that we should try to apply quality outcomes measures to the compensation system. I would not be unalterably opposed if it could be shown to lead in a positive direction that was not intrusive in care. But there are real problems. Medicare spending is the wrong measure, for reasons we have discussed, particularly at the regional level, whether state, county, or even hospital. And applying it to individual practitioners is even more fraught with problems. One is that the necessary ability to risk adjust just doesn’t exist. Another is that an insured patient in a hospital will get better care if the other patients in that hospital are also insured and worse care if they are not. It’s the “spillover effect” due to the issue of total resources that I discussed earlier. So the outcomes that an individual practitioner can achieve depend on more than that practitioner. Similarly, the outcomes that a hospital can achieve for Medicare patients depend on the entire payer mix in the hospital. The proposed systems of modifying payment based on value have the potential of penalizing or rewarding a hospital or a group of physicians when in fact the reason that care was better or worse was that there was either better or worse funding as a consequence of the total universe of patients under their care. Even determining “better or worse” value is a problem.
MC: You’ve done quite a bit of work on the size of the physician workforce, but that does not seem to be an area of concern in reform discussions.
COOPER: Without increasing physician supply, we are not going to have a functioning health care system, with or without reform. But you can’t even talk about it because everyone is talking about this other stuff. It is something we will have to deal with once the current legislation is off the table. It is a huge problem, and it is worsening. The number of people reporting difficulty seeing a physician within the time frame that they need to see a physician is creeping up a couple of percentage points a year. But under the best of circumstances we are not going to be able to solve the physician shortage problem for 15 years or more, and when you extrapolate out 15 years, you’ve got 30 percent of the population saying that they have difficulty seeing a physician. We’ve missed a huge opportunity to deal with this in the context of health care reform, but I am a realist. I don’t see getting it on the table now. It’s for another day.
MC: Thank you.
We hear about Minnesota all the time. Well, Minnesota is a lower-cost state for Medicare beneficiaries, but it is in the top 25 percent of states for total health care spending per capita.
Managed care companies are not bad guys. They are your neighbors, your children, your wives. They are just people trying to do the right thing. That’s how the system works. Help them.
You can’t even talk about the physician shortage because everyone is talking about this other stuff. But extrapolate out 15 years and you’ve got 30 percent of the population saying they have difficulty seeing a physician.