An Interview With Paul Ellwood Jr., M.D.
An Interview With Paul Ellwood Jr., M.D.
MANAGED CARE November 1997. ©1997 Stezzi Communications
NEW OPTIMISM FROM THE 'FATHER OF THE HMO'
History has vindicated Paul Ellwood's prediction of a price-competitive health care market. Is he right again in saying the current trend away from integration will be reversed?
Paul Ellwood Jr., M.D., comes closer than anyone else to being the father of modern managed care. He coined the term HMO, and he — along with Stanford University economist Alain Enthoven — has played a key role in shaping the nation's understanding of managed care and managed competition. That doesn't make him an uncritical industry cheerleader. Two years ago Ellwood told Managed Care that the development of viable quality comparisons among health plans had been stalled primarily by resistance from two groups: physicians and health plan executives. Doctors, he suggested, were taking refuge in "precepts of professionalism" that were based on "having a secret knowledge and people trusting you."
At that time, Ellwood was smarting from the failure in Congress of the comprehensive health care reform plan proposed by President Clinton — a plan based generally but not precisely on the formulations of the Jackson Hole Group Ellwood chaired. Pessimism still gripped Ellwood last December, when the New York Times Magazine published twin articles about him and his son David, an economist who had worked on welfare reform planning. Each man, the pieces suggested, had helped to unleash a revolution — and then regretted the turn it had taken. The Times conceded that managed care had brought costs under control — then finished the thought with its headline: "But What About Quality?"
This fall, as he showed in a recent conversation with Editor Timothy Kelley, Ellwood is once again more hopeful about his long-held dream of quality competition in health care.
MANAGED CARE: Let me start by plunging back into the history of managed care, since you're a key figure in it. Do you remember the moment when you coined the term "health maintenance organization"?
PAUL ELLWOOD: We coined the term HMO for the Nixon Administration as a way of describing an organization that would compete on the bases of price and quality and that would combine insurance and health care in a single organization. That happened in May 1970.
MC: Do you remember where you were when you settled on that term?
ELLWOOD: At a meeting in the Du Pont Plaza Hotel with two men from the Department of Health, Education and Welfare — Assistant Secretary Lewis Butler and Undersecretary John G. Veneman. They were responsible for the development of domestic policy for the Nixon Administration.
MC: Were there other terms that you abandoned in favor of "health maintenance organization"?
ELLWOOD: In my case, no. But there were other terms that were applied to organizations that fit the HMO definition. One was prepaid group practice, the typical Kaiser arrangement. And there was another, more like today's independent practice association, called a "foundation for health care." The HMO idea was embedded in the notion of applying market forces to health care. Kaiser and prepaid group practices, on the other hand, originated in order to get medical care to remote places. The latter was also true of the foundations for health care. They were means of delivering care in mining camps, railroads and the like. The thing that was new about the HMO idea was not combining insurance and health care, but having competition in market forces on quality and price in health care.
MC: Was your initial vision more like the Kaiser model or the IPA model?
ELLWOOD: I thought the group practice model would win out in the end, because I didn't think there was any way to control quality without being very selective about doctors and having a highly integrated health system. And I still feel that way. When competition is based to a greater extent on quality and the ability to get superior health outcomes, that will have a profound effect on the structure of existing HMOs, and they will come to more closely resemble group practices. There's a possibility that through advanced information technology we might see something like virtual group practices, so that highly integrated health care organizations like Mayo or Kaiser might not be the predominant model — if we can succeed in communicating without the doctors physically being together. But in any case I think the health plans are going to have to rely more on full-time primary care physicians and to take more responsibility for medical care if they are going to compete on quality.
MC: In fact, the market's movement has been away from the integrated approach.
ELLWOOD: You're absolutely right. The latest InterStudy survey confirms that. It clearly shows that group practice plans are not growing as rapidly as the rest of the industry. Group practice plans such as Kaiser have begun growing again, but they're not growing as fast as nongroups. But we don't have quality competition, either. The whole thing hinges on what it is that HMOs are competing about. If they're competing strictly on price, and if employers continue to offer one or two choices, then the most effective competitors are going to be those that offer easy access to large numbers of physicians and emphasize good marketing and keeping costs down.
MC: So you see the present market preference for access as a kind of substitute for quality?
ELLWOOD: Yes, but you know, when the HMO idea was introduced, we did not know whether demand would be elastic on the basis of price. We didn't know whether people would go to less costly health plans or not. Clearly, we've learned that people are very price-sensitive, and that very small differences in price cause large movements of patients. We still don't know what role truly accurate, clinically based quality information will play in moving patients from one health plan to another. If consumers don't respond to precise, in-depth information about quality, then price will continue to rule and emphasis will be given to patient satisfaction and that sort of thing, which you can manipulate without having a full-time, multispecialty group practice.
MC: You said HMOs will need to depend more on the efforts of the primary care physician.
ELLWOOD: I still believe primary care physicians are essential to coordinating care and communicating an overall understanding of a patient's problems and how they might be resolved. You need some kind of generalist to pull all the pieces together and function as the patient's advocate in a health system. One of the unfortunate things about the idea of gatekeeping is that it has come to be thought of as a way of limiting access rather than promoting the best possible set of services for patients.
MC: How do you feel about the present tendency of HMO advertising to favor "warm and fuzzy" images indicating great concern, with minimal detail?
ELLWOOD: That's what I call the commodification of medical care. It's being sold like beer and cereal and other undifferentiated consumer products, and I guess it works. [Laughs.] They keep on doing it.
MC: Yet in some advertising, consumers are pretty sophisticated. One hears about rack-and-pinion steering in a car, for instance.
ELLWOOD: That's an interesting question, just how sophisticated consumers can become about health care, particularly when the system is always going to have to rely to a certain extent on trust. There is good evidence that patients who ask more questions and are more aggressive in dealing with doctors actually get better results. But patients at this point aren't terribly sophisticated consumers; the average patient asks fewer than four questions during a doctor visit. Whether quality is going to count depends, I think, very, very heavily on patients becoming better informed.
MC: When we spoke two years ago, you made some pretty scathing comments about why viable quality comparisons among health plans had been delayed. You blamed it on resistance from doctors and HMO executives — the two groups who read this magazine.
ELLWOOD: I'm a lot more optimistic about the likelihood of our relying on really meaningful comparisons of quality now than I was even as recently as six months ago.
MC: What accounts for the change?
ELLWOOD: The demands of purchasers, the anxiety generated by all the adverse publicity and probably a recognition that if we're going to continue to control health care costs we're going to have to have better insights into what's actually working in medical care. A whole lot of things have come together to move the idea of genuine quality comparison. I also have modified my views on how to do it. I used to feel very strongly that, like financial accounting, health care quality measurements should be uniform, that there should be one generally accepted set of standards. We're not moving toward that goal very rapidly. But there are many systems out there that are in use or ready to use that will allow for comparisons — maybe not the perfect ones that I hoped for, but a start in moving health plans in the direction of being truly clinical organizations.
MC: What about an alternative vision? What about a world in which the role of the HMO recedes somewhat to that of a third-party administrator and the multispecialty physician group takes on the functions of quality control, credentialing, outcomes study and things like that?
ELLWOOD: It comes down to who is at risk financially. If a health plan is going to be a third-party administrator and do marketing and reinsurance, the bulk of the financial risk will shift to providers. I think it's inevitable that the organization that takes the financial risk will be much more clinically responsible than the typical carrier-based HMO of today.
MC: Employers were interested in quality two years ago, when you helped to found the employer-backed Foundation for Accountability (FAcct). Yet you say your new optimism is more recent than that. What specific events have inspired it?
ELLWOOD: Now, a number of major purchasers are saying, "If we're going to buy from you, we want you to collect the following outcomes information." They include American Express, IBM, Marriott, Sears and Pfizer. You have the Federal Employees Health Benefits Program writing its desire to obtain outcomes information into its contracts. And the Health Care Financing Administration is now moving to do the same. At the same time, health plans are realizing that their public image depends on their being able to demonstrate real quality.
MC: Aren't health plans up against a tough challenge, even if they have abundant quality information, in arguing for the numbers and against anecdotal reports of individual tragedies?
ELLWOOD: It's very hard to argue epidemiologically against individual cases, but that's also why we have quackery. If we based all of medical science on the idea that one patient got well or didn't get well, we'd be back in the dark ages in medicine. But I don't think the managed care industry has even used what information it has to defend itself. There was a piece in JAMA recently about how lenient utilization review is. I was appalled. Reviewers were approving something like 80 percent of the procedures that should have been denied. They've got to get in and actually manage care.
MC: Is it fair to conclude from what you're saying that you have limited sympathy for physicians who complain about micromanagement and the "800"numbers they have to dial?
ELLWOOD: Well, from what I can see, it seems there are very few instances where utilization review denies the opportunity to do what is appropriate with patients.
MC: How about the hassle of even just having to ask? And especially having to ask a bunch of different plans for a bunch of different policies?
ELLWOOD: How did we get to the point where health care was so expensive? We were doing too much with patients in the wrong places. Now I think it's going to take a better rationale for doing less than we have, and I guess I'd always hoped that that could be accomplished intellectually rather than through utilization review. But what I hear in talking to doctors who try to stimulate other physicians to use outcome measures is that doctors are remarkably uncurious about how their patients turn out. It varies a lot by specialty. Orthopedists are very interested in whether their patients can walk without pain and without deformity. Apparently, there is very little interest on the part of your typical obstetrician in health outcomes. That's why the discoveries in hormone replacement after menopause have come from other than obstetrics. They have come from things like the nurses' study conducted by the Harvard School of Public Health. At any rate, I misjudged the degree of intellectual stimulation that might accrue from being able to follow large numbers of patients and feed the information back to doctors. It was clearly a mistaken impression that I had as a physician, and I think it stemmed from the specialties I was in. I was in physical medicine and rehabilitation and pediatric neurology. Pediatrics emphasizes growth and development, and physical medicine emphasizes functionality and activities of daily living, so following patients and charting their course is very important to those specialties.
MC: Do you think primary care physicians are as interested as they should be in outcomes?
ELLWOOD: I don't know. I haven't seen any data on it. I had hoped they would be interested, because I felt it would bolster their position as patient advocates and integrators of care. Now if primary care physicians are just general doctors posing as specialists, then of course it doesn't carry as much power for them either. Do you understand what I mean?
MC: If they are acting on an episodic basis rather than providing patients with a complete care experience that promotes a better outcome in every sense?
ELLWOOD: Yes, right. But if they want real power, then their best opportunity is to be in a position to pull together all of the pieces and know more about the overall course of a patient than anybody who's looking at his gut, or his eyes, or his ears, or his heart.
MC: The model that appears to be dominant right now is that, instead of HMOs with exclusive teams of doctors, we have doctors all over the place, usually in groups, with each group having contracts with each of the HMOs. Physicians are asked to do one plan's bidding — and also the bidding of its competitors. Doesn't that diffuse and complex structure make quality measurement and standardization very difficult?
ELLWOOD: Yes. That's why I think that once you begin forcing health plans to compete on quality, the present predominant structure of managed care will have to be changed. It isn't worth the investment in training, integration, information systems and so forth in the present system, where quality per se doesn't necessarily pay. I must say that one of the initial assumptions about HMOs was that if you didn't do it right, it would cost you. I think that's true to a large extent, but not so much that it has forced health plans to be very sophisticated about how they manage quality.
MC: Well, for one thing, there's the turnover issue.
ELLWOOD: Yes, but that's not so great. That's a part of it, certainly, but it isn't as extensive as you might think. One thing that happens with chronic illness — if you're in fee-for-service and have a chronic illness, like people on Medicare, you're reluctant to join a health plan because it means changing doctors. The same thing happens in HMOs. If you're sick, you want to preserve your existing medical relationships if you're reasonably satisfied with them. Part of the turnover, of course, is attributable to the fact that Plan A isn't any different than Plan B. For instance, Alain Enthoven was telling me the other day that he thinks that one of the reasons why in CalPERS or at the University of California you get so much movement with slight differences in price is that the patients realize that they're not going to any different set of doctors when they change. So if you can get the same product, or what is apparently the same product, for less money, why not switch?
MC: That casts doubt on people's supposed willingness to switch plans for a tiny premium savings.
ELLWOOD: Yes. That sort of demand elasticity is really pure economic demand elasticity because the product is the same — or at least they're assuming it is. That's all part of this commodification phenomenon, with the reliance on soft, fuzzy advertising. I think there is a lot more — I'll call it "taste" — in medical care than is recognized. A lot more subtle, personal preferences between patients. I can remember when I was actively practicing when you'd take another doctor's practice when the doctor went on vacation or whatever — their patients seemed to be alike. They seemed to have similar personalities, similar expectations. One doctor's patients would be very, very demanding; another's would be very passive. There are those kinds of differences between people that a health plan can, I believe, take into account and base itself on.
MC: Despite your new optimism about quality measurements, you've been having some pessimistic thoughts lately about the way physicians are viewed by patients. Isn't that right?
ELLWOOD: Yes. In some ways I see a very negative kind of interaction developing. People are not trusting their physicians, and the untrusted physicians are feeling unhappy about what they're doing and conveying that unhappiness back to patients. It's kind of a vicious cycle of distrust. These systems where doctors work for several different organizations and aren't necessarily loyal to any particular one, and where the doctor exerts very little influence on the organization, lead to a depersonalized relationship between the doctor and the organization and the undermining of doctors' trust and loyalty in the organization.
MC: Yet what we hear about some traditional HMOs is that they're ceasing to be completely staff-models or group-models, they're asking their doctors to work more hours for the same money or the same hours for less money. They're casting off groups of doctors into semi-independent arrangements.
ELLWOOD: That's the way the Harvard Pilgrim Health Care is behaving. But Kaiser Permanente is moving toward integrating its medical groups to a much greater extent, and part of the reason is a reaction on the part of Kaiser's physicians to the leadership of the health plan itself moving in a Harvard Pilgrim-like direction. And I don't hear of the traditional multispecialty group practices falling apart. Groups of doctors have always been difficult to manage, and there are always going to be problems around how much people get paid.
MC: Kaiser, too, has required doctors to work more hours for the same pay, hasn't it?
ELLWOOD: They're having to work harder to compete. That's the big problem with salaried physicians, motivating them to stick around if the waiting room is still full at 6 p.m.
MC: So trust is better if you have an exclusive relationship, as with salaried doctors, but for trust you pay a price in motivation for productivity?
ELLWOOD: Well, that's been a problem with a lot of organizations, and most group practices in the past paid doctors on the basis of productivity, the amount of money they brought in. The sacrifice you make when you pay a salary is, perhaps, a decline in motivation to work hard. But you know, it's funny. You don't hear that so much about executives in industry.
MC: Do you take a dim view of what Harvard Pilgrim is doing in terms of opening up the staff model and cutting some of the doctors loose from full-time involvement?
ELLWOOD: Yes. The thing that was admirable about the Harvard Community Health Plan was how well integrated it was. There was a sharing of responsibility and information, a sense that "We're all in this together to practice superior medicine." That's what I think we need to strive for.
MC: But you see Kaiser — and Group Health of Puget Sound also, since it's merging with Kaiser — as still living up to that ideal?
ELLWOOD: Yes, but underlying all of these generalizations are specifics that apply to these individual cases. Group Health of Puget Sound, ever since I've known about it, has had a certain amount of difficulty between the doctors and the lay board. And Group Health's latest problems are economic. If it does merge with Kaiser, I'd say the merger is based more on economic considerations than on some peculiar affinity of the two organizations. It gives Kaiser a bigger entry into the Seattle market.
MC: Do you see a Kaiser-Harvard merger someday?
ELLWOOD: I don't know. I've never talked to the Kaiser people about that. Years ago, I worked and worked on getting Kaiser to expand across the country and, at first, it was very hard to convince them to do so. I tried to get them to go into Detroit, Cleveland, Washington and many of the places where they ultimately went. This business of growing group practices rapidly and maintaining their culture is a problem that we haven't fully solved. The great group practices, in general, train their own people and grow relatively slowly. The thing about managed care, you know, is that it has grown explosively, and it took more expeditious arrangements to grow it as fast as it's grown. Stepping back from that isn't going to happen overnight.
MC: What's your proposed remedy for the "vicious cycle" you see in physician trust and morale?
ELLWOOD: I think the remedy is organizations where physicians work pretty much exclusively for one health plan and have a much greater voice in the governance of that plan, and all of the systems in the plan are built around superior patient care and providing the doctor with a rationale for working in a group. The thing that has made group practices work in the past is communication — the opportunity to share concerns about how to manage a patient with other doctors. Health plans are going to have to justify their organizational existence by making the practice of medicine in an organization superior to practicing by yourself. And it's got to be something more than having somebody to take calls for you.
MC: Do you think the shift you're predicting toward exclusivity and integration, which as we discussed is almost the reverse of some present trends, is going to happen by itself in response to more effective quality measurement?
ELLWOOD: Yes. I think the principal driver will be competition on quality, and the same thing will also drive computerized medical records systems. They're not going to happen just because they're a good idea. They're going to happen because they're necessary to do a better job of practicing medicine.
MC: Someone semifacetiously suggested I ask you why the things you predict don't come true.
ELLWOOD: Well, we have a new American health system, don't we? [Laughs.]
MC: We definitely have one that's different.
ELLWOOD: I can remember at the beginning, when we started pushing this system, nobody believed it. Absolutely no one of consequence thought that this could conceivably happen. And today we do have price competition and we're moving toward quality competition.
MC: What happens when prices start going up again, which some experts expect?
ELLWOOD: That's another argument why we have to know more about what works so that we can be more selective about what we do and justify it. I'm interested in this comment about how come none of the things that I predict happen. I think I'm the best forecaster in the health business. My problem is that I'm too impatient. We talk about things happening so far in advance of their happening and the mistake I think people like me make is that we tend to predict that something will happen that we want to have happen, without a lot of evidence that it is actually happening. You were talking to me a couple of years ago. I went through a very pessimistic period about the quality and accountability stuff. I was really discouraged.
MC: Did your pessimism include impatience with the National Committee for Quality Assurance?
ELLWOOD: That's one of the things that made us suggest FAcct. I started out focusing my efforts to get quality accountability going on — Kaiser was probably the first target, and health plans in general. I didn't get very far there. Then I turned to the Joint Commission on Accreditation of Healthcare Organizations and the AMA. And then along came NCQA. But they all seemed to want to rely on the old methods for doing it — you know, process measures and one doctor looking over another doctor's shoulder, peer review. Our pushing FAcct did in part stem from the idea that the NCQA seemed to be relying on methods that really weren't going to affect to any great degree the emphasis plans put on quality. It wasn't going to change the way they operated, I thought. But NCQA and FAcct are working very closely now. NCQA has just agreed to have FAcct be the one that develops pediatric measures.
MC: So you think NCQA has sort of got religion?
ELLWOOD: I think that this week. [Laughs.] It has also got a lot of competition, though.
MC: Do we need all these different organizations measuring quality?
ELLWOOD: I didn't think so at first. I also didn't want to see a lot of nongroup health organizations. But you take what you can get. I really think information technology will enable us to overcome the subtle differences among these various systems. Maybe measuring quality isn't like measuring dollars in financial accounting, and we have to go through a stage of letting a thousand flowers bloom.
MC: Yet in the New York Times Magazine last December, you were said to advocate NCQA's being made into an agency with teeth, like the FDA. Do you still favor that?
ELLWOOD: It wasn't NCQA necessarily, but some Securities and Exchange Commission-like entity in health care. But in the latest congressional go-around this year, there were some quality proposals on the table, and those proposals got nowhere. I am at this point persuaded that our government is incapable of reaching a consensus on how to manage quality. At the same time, working relationships among HCFA, the Federal Employees Health Benefits Program and the Defense Department in the area of quality are getting better. Maybe it's something that has to be demonstrated in practice before we're ready to legislate it. You know, I made the same mistake with HMOs. I thought we were going to reform the health system with Medicare. Now we're going to reform Medicare with the health system.
MC: Starting with Truman, administration after administration has impaled itself on the goal of national health insurance or some form of universal coverage. And it still doesn't seem to get anywhere, even though many other countries have it. What's the lesson of this experience?
ELLWOOD: The conclusion I've reached about getting everybody covered is that we've moved a long way toward unlegislated managed competition, and in the course of it have managed to get more people covered. The children's health insurance provision contained in the recent budget agreement is the first significant expansion of health insurance coverage since 1965, since Medicaid. A huge thing. And I think the reason it was possible to spend $24 billion that way was because managed care had been so successful in alleviating the wage-benefit spiral. The real behind-the-scenes debate over managed competition had to do with whether or not demand would be elastic for health care, allowing money to be saved through managed care. Well, money was saved through managed care, and it's coming back to the government in the form of taxes. This whole budget agreement was really possible not because of Medicare reforms, but because of private-sector reforms leading to greater revenue generation for the government. Health care has been one of the keys to the success of the whole economy. Health care costs and benefit costs in general were out of control until managed care came along. I was reading the other day about Federal Reserve Chairman Alan Greenspan. Apparently, he has his most profound ideas when he's in the bathtub. And he has decided that one of the reasons for the sustained growth of the economy is that the Labor Department wasn't fully measuring increases in productivity. I hope one of these days when he's sitting in the bathtub he looks at the health care figures and thinks, "Wow! Here's something that's really different about labor costs and productivity."
MC: Do you expect medicine still to be a satisfying profession for the personality type that wants autonomy, when all is said and done?
ELLWOOD: No. I think people who really need to do things all by themselves don't belong in medicine. But all the changes haven't proved to be a deterrent to applicants to medical school. The number of applicants is way up. My dad was a doctor, and I remember what it was like. My mother used to complain, "Why don't you stay home with me and the kids?" He used to think he had to spend an hour with every patient. We would stop by at a patient's home on our way to my grandmother's house for Thanksgiving dinner, and we knew we'd be waiting in that car for an hour or more.
MC: What prompted you, back in 1968, as a physician and the son of a physician, to start looking at how health care worked in society as a whole?
ELLWOOD: I think that's a peculiar personality characteristic some people have. They try to generalize from their experiences, and look at the greater significance of the particular. Most inventors and visionaries are kind of discontented with the way things work right now and are trying to figure out some better way. I think it's in the genes. I've talked to other people who are inventors or reformers, and I don't think inventors or artists or reformers are very different from each other. You're just lucky if you're working on something that is perceived as being in the public interest. Whether you're tinkering with computer chips or tinkering with social systems, it's not that different.
MC: Finally, may I ask you what kind of health care coverage you have personally?
ELLWOOD: Well, we don't have managed care here in Jackson Hole. I was in a managed care plan until I moved here, and I've worked on getting HMOs started in Wyoming, because I'd join one in a minute if there were one here. I think Medicare is a crappy insurance policy.
MC: Thank you, Dr. Ellwood.