Web feature

Can America reduce medical costs while expanding the availability of health care?

Policy expert Stephen Shortell focuses on the disconnect between costs and outcomes, and calls for a robust comparison of treatment options

| 25 June 2009

The current U.S. health care system — with its ever-rising price tag and its gaping discrepancies in care — is a "ticking time bomb" for the federal budget, President Obama warned members of the American Medical Association. Without corrective action, he said on June 15, "America may go the way of GM — paying more, getting less, and going broke."

Stephen Shortell — an expert in health policy, management, and organization behavior and dean of UC Berkeley's School of Public Health — currently serves as an adviser to the Obama administration on health care reform. The NewsCenter spoke to him recently about the feasibility of reducing costs while expanding the availability of health care insurance.

Stephen ShortellDean of Public Health Stephen Shortell (Jeffery Kahn photo)

Q. In talking about an essential overhaul of our health care delivery system, you've noted that one symptom of its dysfunction is the huge variation in care across the country — rates of hospitalization, lengths of hospital stays, number of physician visits — and ultimately of health care costs. Would you talk about those discrepancies and why they matter?

A. Sure; let me give an example. There is a three-fold difference between southern and northern California in the number of physician visits and intensive-care unit days in the last six months of life. When you look back at the kind of care patients received in their last six months (risk adjusted) of life, they had three times as much intensive care in southern California as in northern California.

These patients all died; it's not that some improved because of the care they were given. In fact, there's data that has shown a negative correlation between cost and quality: that in high-cost "hotspots" — like parts of southern California, Florida, where data suggest that people are actually getting a lower quality of care.

So this is very troubling, and is part of the impetus for reform. We spend a lot on health care, — over $8,000 per capita, more than any other nation — but we're not getting the value for our dollars. The U.S. is ranked 38th in terms of health-status statistics.

Q. What's behind these wide variations in health care costs?

A. Mostly it's that the hotspots have more doctors; it's supplier-induced demand. If you have more doctors, more hospital beds, more high-tech testing and treatment options, you're going to use them. Why? A lot of medicine is a matter of judgment. If you have insurance coverage for it, why not put a person in the hospital, and why not visit him or her every day? The incentives reinforce practices that lead to high costs.

Q. Besides high-cost regions, aren't there also high-cost conditions, for which there many options for prevention and treatment — of varying costs and effectiveness?

A. The big, high-cost diseases tend to be the chronic ones: diabetes, congestive heart failure, asthma. Depression is high cost. With chronic illnesses like asthma or diabetes, depression is often an accompanying condition. No surprise. When you have some of these conditions, you get depressed, because you can't do the things you used to do. There have been studies comparing different kinds of anti-psychotic drugs — generics vs. the brand-name drugs. As you know, generics cost less, yet they are based on the same basic formula as the brand names, the one you see advertised on TV by drug companies. The studies have found no difference at all in outcomes — yet we would have saved a lot of money if people had taken the generics.

Stephen Shortell: How schools of public health can contribute to health care reform

Probably the biggest lever for containing health care costs over time is public-health interventions. The Association of Schools of Public Health has a position statement to that effect; we have weighed in.

But here's the thing to understand: not all prevention practices save money. ...

Another example: hypertension leading to coronary artery disease. A recent study of anti-hypertensive drugs made the same finding: no clinical difference between the generic and the brand-name drug. We could save hundreds of millions of dollars, probably, if patients and doctors would use the generic brand.

Prostate cancer in men is a high-cost condition that is very prevalent today. There are three or four different treatment options. One is called "watchful waiting" or "active surveillance." (For a fair number of cases, that's sufficient; the condition is benign.) So there's that option, or surgery, or various medications, or a new treatment called proton-beam therapy, which is pretty expensive; we spent $150 million a year across the country on proton-beam therapy. Recent comparative-effectiveness studies of these four different options have shown that none of them has any clear advantage. They cause different combinations of side effects, but there's no clear winner in terms of curing patients earlier or improving overall functional health status.

A fourth example is low-back pain. In the U.S. we spend at least $50 billion annually on low-back pain, and it's the most common cause of job-related disability. There are a variety of treatment options for low-back pain, from surgery to various non-invasive treatments.

Q. How can we get a handle on which treatments are most effective and most worth investing in?

A. I was on an Institute of Medicine committee that called for the development of a Center for Comparative Effectiveness, whose purpose would be to compare the effectiveness of treatment alternatives in a rigorous way. It would not compare a treatment or drug or medical device or procedure against nothing, which is what the Food and Drug Administration does in its clinic trials: compares the drug or treatment against a placebo and then says "Well, it's better than nothing." In contrast, this center would compare current treatments, devices, drugs, or other interventions, to see which has the best outcomes and which might be worth an added cost. And then it would feed that information to doctors, patients, and health-insurance plans. Most other countries have an entity like this. We're about the only western nation that doesn't.

There is interest in Washington in this idea. In fact a committee appointed by the Institute of Medicine has been asked to come up with the first 50 procedures or medical devices that should be evaluated for comparative effectiveness. Some of us with a public-health perspective have suggested that they ought to broaden the criteria — to include not just clinical procedures, but interventions that make us well and keep us well. Rather than just react downstream, after people are sick, intervene upstream with prevention programs and health promotion. So, for example, how do we prevent obesity in children? There are various interventions, such as healthy diet, or healthy diet plus physical exercise. Why not evaluate these for comparative effectiveness?

Q. Some may object that collecting and disseminating comparative-effectiveness data would lead to the government dictating what treatments doctors should be providing or insurance companies should cover.

A. We're just saying "Let's get valid information out there, and then people can decide." No one would be dictating to Blue Cross or United Healthcare or WellPoint, saying "Here is the data, and by the way, you shouldn't cover this treatment for people."

And yet there's going to be enormous push-back from the pharmaceutical and medical-device companies. The argument being "it stifles innovation." In my view this is not about stifling innovation; just innovate on different criteria. There are technologies, devices, and drugs that would improve people's quality of life and overall health and would cost less. (Or maybe they are higher cost for a while, but as they're used the unit costs drop.) So target some of your R&D towards these lower-cost treatments, and not just the blockbuster drug that takes 10 years to develop.

Stephen Shortell on health care payment reform — two proposals

"Under the current payment system, hospitals make more money if a patient is readmitted. This past year, for example, 18 percent of Medicare patients (Americans 65 and older) who were discharged from a hospital had to be readmitted essentially for the same problem or issue. And it's estimated that half didn't need to go back into a hospital — if they'd had better follow-up medical care. And that alone would have saved the taxpayers $12 billion.

"With a 'bundled' or 'lump-sum' payment system, which has been proposed, you say, for example, 'Here is $70,000 for doing a risk-adjusted coronary bypass graph surgery. Share this amount among the hospitals and doctors.' So that they have an economic incentive to work together on behalf of the patient.

"Another payment reform that's been proposed is called 'episode-of-care.' If you have diabetes, say, or asthma, your provider gets so much money up front — risk-adjusted for the severity of patient illness (maybe $2,500) — to care for you throughout the whole year — no more. So your provider has every incentive to keep you well, to make sure you're on your medications, you're following your diet, you're exercising — because he or she gets to keep some of the savings on that $2,500. But to guard against 'under treatment,' various quality-of-care standards and measures would also have to be met."

The latter may be an outmoded model for pharmaceuticals. Yet the industry is locked into a mindset: "This is the way we've always done it. It costs a billion dollars and ten years to bring a drug to market — so we deserve that high mark-up, because it feeds back into R&D." (And their profit, too, of course.) So they're going to fight this proposal.

Q. Don't insurance companies already refuse to pay for expensive treatments, saying they work no better than a cheaper alternative?

A. Yes, informally and on a small scale, many of the big insurance companies are doing just that, and patients understandably respond: "You say the treatment I want is not as effective; but it's in your self interest not to pay for that treatment." It would be in patients' interest to have an independent body, such as a Center for Comparative Effectiveness, conducting rigorous studies and making its findings public.

Q. You've been advising members of Congress and the Obama administration on health care reform. What changes did you and your colleagues recommend?

A. The thing that has struck me, that I'm encouraged by, is that during one recent meeting in Washington, for example, all of those present — largely academics and health-policy analysts — were converging on the same ideas. It wasn't everyone having a pet theory or pet prescription. We were all pretty much agreeing on key elements for reform, in some cases based on considerable research and in other cases based on best judgment.

One point of agreement was what we've just been discussing: the need to collect and disseminate data on comparative effectiveness of various treatments. And on the performance of doctors and hospitals, in terms of patient outcomes, as well.

Another: we've got to reform the payment system, begin to get rid of fee-for-service and focus more on bundled payments. Under the current system, doctors charge on a fee schedule. The more they do, the more money they make; there's no incentive for cost containment. We need to move away from that systematically over time, and instead begin to reward providers for better outcomes, and pay them in bundled payments. For a given procedure, doctors and hospitals would get a lump-sum payment (rather than being paid separately, as is done now) — thus encouraging them to work together to care for the patient. 

So we were all in agreement that payment reforms are essential, and that we need to spend more on primary care and less on highly specialized care.

Q. What about electronic health records?

A. We agreed on the need to modernize health records using information technology. The government has already allocated $19.6 billion, in the federal stimulus package, for electronic health care records, so that physicians and hospitals can better coordinate care.

My chief concern there is that too much of that money will be spent on hardware and software, and not enough on the technical assistance that doctors need to really reorganize care and use these electronic systems. The School of Public Health has done evaluations of physician practices as they switch over to electronic health records. We've found that this change is very disruptive initially; medical offices probably lose six months of productivity when these systems are first installed. In the long run, it begins to pay off — but in the short run, if you're a primary care doctor not making a lot of money, it hurts. I'm hopeful that some of our IT investment will be used to assist doctors and hospitals in coming up to speed with these new technologies.

Q. What are the stakes for the Obama administration as it seeks to make good on his promise to bring health care reform?

A. The stakes are very high. There are too many important people giving too much time and energy to this for something not to happen. Lawmakers cannot go back to the voters in their districts with a total failure. So I think we will achieve some degree of health care reform. Perhaps not by August, which they're racing to do; but I think there will be a bill. Given the energy and the commitment that's gone into this, no one wants to have total egg on their face.

On the other hand, I think the reform will be far from what is needed. And the question then becomes whether it's anything that can be built on in Obama's second, third, and fourth year, or if he has a second four-year term.

Q. Hasn't this been referred to as the chance of a generation for health care reform?

A. That was what was said, too, in 1994 and in 1976. But yes, this is the chance of a new generation. It's interesting: when I was a graduate student in the 1970s, health care accounted for about 6 to 7 percent of the U.S. gross domestic product. I can remember people saying "Oh, we've got to do something about these costs! They can't go to 10 percent." And then we got to 10 percent. And it was "Oh, my God, we passed 10 percent! It can't go any higher." Then it went to 12 percent — and now it's 16 percent.

So people are now saying "It can't go to 20 or 25 percent, can it?" It's certainly going to do that unless we do something really radical. At the current rate of spending, the Medicare Trust Fund is going to run out of money; it's five or six years off. If we don't do anything, in four or five years the cuts to Medicare will have to be totally draconian. Medicare will cut payments to doctors and hospitals by maybe 50 percent, rather than just 5 or 6 percent. Some doctors will change careers; they'll leave medicine. And the insurance companies will be in turmoil. That's why the insurance companies and the medical-specialist societies, at least for now, are saying "We'll cooperate." The devil is in the details, though; they're going to oppose many proposed aspects of reform. 

So this is very momentous, and the Obama administration is putting a lot on it. I think something will get done — though not what many of us think needs to be done. But it's a start. Let's take encouragement if it happens.