By Hamilton Davis
Vtdigger.org, April 25, 2011
For the second time in 15 years, the state of Vermont is making a concerted effort to rebuild its health care system. The odds were long for the mid-1990s campaign, which failed, and they are equally long now. That is not to say that the drive toward reform initiated by Gov. Peter Shumlin is doomed, only that it is a very long shot.
In some ways the program recently adopted by the Legislature is considerably more far-reaching than the earlier try. The aim now is a single-payer system, which almost certainly can’t be fully achieved, but which does at least pinpoint the two central issues: the need to provide health care coverage to everyone in the state and the need to control costs in the delivery system, which routinely rise at 6 percent to 9 percent above the underlying rate of inflation.
Moreover, Vermont policy makers, most importantly the governor himself, understand that no reform effort can survive unless the costs can be brought under control. Shumlin has been pressing for a single-payer system for several years now, but among the thousands of words he’s expended on the issue, none were more important than his statement that he wouldn’t go forward with single payer unless the system costs can be constrained.
In 1966 when the federal government took responsibility for providing health care to the poor and the elderly, costs in the system amounted to a little more than 6 percent of our total national output. That figure is approaching 17 percent today and is heading steadily toward 20. It’s a cost engine that reaches into virtually policy corner of American life. Some 50 million people in the United States have no health insurance; many millions more are underinsured. In Vermont, 47,000 people are uninsured; another 160,000 are underinsured; the total encompasses just under one-third of the state’s population. Even those who are well insured may pay huge monthly bills.
The damage is not limited to individuals. Medicare (for the elderly) and Medicaid (for the poor and lower income workers) play a major role in both federal and state budget deficits. Those pressures in turn foreclose investments in such critical areas as education and physical infrastructure, like bridges and roads. And medical inflation saps the competitive fiber of American business firms, large and small.
The clear need for reform is an obvious asset for the Shumlin administration going forward. There are some other promising auguries. One is the staff the governor has assembled to manage the program. Anya Rader Wallack, whom Shumlin has retained on a contract, is a national class health policy analyst. Steve Kimbell, the newly-appointed commissioner of BISHCA, is tough and knowledgeable about state government. Shumlin has called them a “dream team,” and that isn’t hyperbole. An added plus: When the final design comes into the legislature next year, it will be managed in the House by Shap Smith, the most adept Speaker since Ralph Wright.
So, given the pressing need and the assets assembled, why are the odds so long?
Health care is nothing like a market
There are many reasons why the health care issue is one of the gnarliest the country has faced in the last half century. One is that physical health is vital to every single person. This fact invests health care with enormous importance: emotionally, socially, financially. Another is that it is technically complex, well beyond the reach of most people. It takes 10 years to train a brain surgeon. Even the most intelligent, well-educated patient has no chance to figure out difficult medical questions on his or her own.
The single most important reason, however, is the strange, almost perverse way the health care system is wired, or more accurately, not wired together. While many policy makers talk about trying to utilize market forces to control costs, health care is nothing like a market.
Consider a patient, call him John. He wakes up one morning and feels badly. He goes to see his primary care physician, Sally Jones. Dr. Jones examines him, makes a diagnosis, and prescribes treatment. This process could include lab tests, imaging like x-rays or scans, and possibly a course of drugs. Possibly this regimen works. If not, Dr. Jones could refer him to a specialist physician, or a local hospital.
The new physicians, or the local hospital, carry out the same procedure, at a more elaborate level of sophistication. Maybe that works. If it doesn’t, the specialist could refer John to a subspecialist with narrower but deeper expertise; he could also send him from a community hospital to a tertiary medical center nearby. The specialists and the staff of the medical center carry out the same diagnosis-treatment regimen. If John can’t be cured at this level he may be referred to very sophisticated physicians out-of-state, to Pittsburgh, say, for a heart transplant, or Boston for high-level cancer treatment, or the Mayo Clinic for a stubborn illness that has frustrated diagnosis.
It is also possible for John to pick and choose his route through this maze. He can pick his own specialist and subspecialist. He can choose one hospital over another, and patients do, all the time.
The critical point here is that these nodes in the system are normally financially and professionally separate from one another. Whoever is paying for John’s health insurance just has to keep paying the individual bills. If there is duplication of effort—repeated imaging or lab work, for example, or unnecessary care—well, that still has to be paid for. No one is responsible for John’s total care; there is no single price. The driver of this process is that most doctors and hospitals get paid on a fee-for-service basis—the more things they do, the more they can bill for. That reality is an engine that drives cost.
And it has powerful purely medical effects also. American doctors always take responsibility for their own patients. But they hate taking responsibility for what other doctors do. So if John gets unnecessary care or poor quality care no one is responsible for repairing the situation. Anyone who doubts this should read the Institute of Medicine report issued in 1999 showed that hospital errors in the United States kill as many as 98,000 people per year, three jumbo jet crashes worth per day.
That does not mean that all, or even most care, in the United States or in Vermont, is bad. Much of it is excellent—indeed some of it approaches the miraculous. But there are no financial incentives for brilliant care or even fully-up-to-standard care. The incentives, in fact, run in the other direction. If you go to the hospital to have your right leg cut off and they cut off the left, your payer will end up getting charged for cutting off two, since you still have to cut off the left. A few years ago the Henry Ford health care system in Detroit decided not to bill for two procedures in such a case, a departure so striking that it made the front page of the Wall Street Journal.
Cutting off legs is pretty dramatic, of course, but the same dynamic works across the system, from messed up appointments to misdiagnosis, to all manner of errors and unnecessary care. Mistakes in the system get paid for by the public, not by the health care providers. Markets don’t generally work like that; if a system does work like that, it isn’t a market at all.
The key piece of evidence that the above scenario is valid is the technique, developed in Vermont by Dr. John Wennberg, called small area variations. In the early 1970s, Wennberg, then at the University of Vermont College of Medicine, began collecting and then publishing data showing that medical use rates varied dramatically between communities n the state. At that time, for example, 60 percent of the kids in Stowe had had their tonsils removed; in nearby Waterbury, the fi
gure was just 20 percent. The rates of gallbladder surgery varied between hospital services areas by three times, and hysterectomy varied by four.
The health status of one community can differ from another, of course, but the main indicator of a significant difference is a difference in age and Wennberg adjusted for age. No one was ever able to make a straight-face case that those kind of variations were driven by anything other than doctors dealing with similar problems in radically different ways. The small area variations analysis was particular striking in Vermont, which is quite homogeneous in its population—mostly white, with a few small urban areas and the remainder suburbs and rural settlements.
Over time, Wennberg, who soon moved to Dartmouth Medical School, steadily expanded his data analysis to other states; today, these data are regularly published in what is now known as the Dartmouth Atlas of Health Care. This document is one of the foundation sources for any assessment of health care in the United States today. In his book Complications, Atul Gawande, a surgeon at Brigham and Women’s Hospital in Boston, cited Wennberg’s work as evidence for the baleful influence of variation in medical practice in the United States.
“What he has found is a stubborn, overwhelming, and embarrassing degree of inconsistency in what we do,” Gawande wrote. “His research has shown, for example, the likelihood of a doctor sending you for a gallbladder-removal operation varies 270 percent depending on what city you live in; for a hip replacement, 450 percent; for care in an intensive care unit during that last six months of your life, 880 percent.”
Given that the lower utilization rates tend strongly to show up in areas with the most sophisticated and high quality care, it is a reasonable inference that a significant volume of the care in the higher use-rate areas is unnecessary. And a major contributor to the health care cost problem.
Generally speaking, Vermont has tracked the national trends in the rapid growth of health care expenditures. Comparing Vermont to other states and to national averages is complex. In the 1970s and 1980s Vermont’s costs tended to be comparatively low, but over the last two decades Vermont has grown more rapidly than the country as a whole. Irrespective of how such analyses play out, it seems clear that the Vermont cost inflation is not sustainable.
Per capita health care spending increased by 8 percent between 2004 and 2008, at a time when the underlying rate of inflation in the economy ran around 2 percent. The difference of about 6 percent could be called the unsustainability rate. There is no way Vermont state government, or any government, can keep raising major revenues at that rate for long. And that performance has been going on for 40 years.
No one in the United States has figured out how to solve this dilemma. And it isn’t for lack of trying. The federal government and the governments of the various states have been trying to put the brakes on health care costs since the early 1970s, once it became obvious that Medicare and Medicaid had increased demand at a much more rapid rate than anyone had anticipated. There were certificate-of-need programs to limit growth in capacity and for system costs,rate setting in many states, and a raft of alphabet soup programs, PROs, and PPROs, and DRGs — the landscape is littered with the smoldering hulks of these programs. Nothing worked: fee for service medicine defeated them all.
It isn’t theoretically impossible for a Vermont single-payer system to solve this problem on the way to controlling costs, but it won’t be easy—at all events, nobody has figured it out yet. The core difficulty even if you had a pure single payer—state government—it isn’t clear how you could divide up a shrinking pile of money among the various unconnected nodes of the system. How much of the pie goes to our patient John’s primary care doc, and how much to the array of specialists, sub specialists, community hospitals, local medical centers, like Fletcher Allen and Dartmouth-Hitchcock, and how much to the out-of-state centers for really tricky stuff? One could argue that the doctors and hospitals could figure that out for themselves—which is actually a very good idea, but there’s no way to do that in the absence of hard-wired connections between the various elements in the delivery system.
And a much tougher question than that: who decides which of the elements of the care in the whole skein is good, very good, kind or poor, unnecessary, incompetent?
In short, fully rationalizing a balkanized system even in a tiny state like Vermont is as tough a problem as anyone could find. Yet, if the state takes the responsibility for taking care of everyone without have a firm grip on the costs then Vermont would end up with a financial problem like it has never seen before. Doesn’t that argue for abandoning the whole idea?
Well, no. Because the other side of the coin is equally valid. Failing to get a grip on the costs and quality in the system will be damaging to our whole society in the ways that we’ve laid out above. It would probably be better if the problem were to be tackled at the national level, but it doesn’t look like that’s going to happen any time soon. If Vermont is willing to step up, we might find our way to higher ground.
The hurdles for a single-payer system
Since the Legislature is close to final passage of a bill authorizing the program development process, it’s worth asking what other hurdles are out there. VTDigger.org will examine each of these in detail over the next several weeks, but for now here are just some of them:
* The exchange and the insurance issue. In a true single payer, private insurance isn’t an issue. One payer pays all the bills. In the original legislation, the Shumlin administration proposed to begin backing insurance companies out of the state before the full system has been designed. This effort not surprisingly upset the private insurers, but it has also brought the business community into the fray. Business firms are concerned that competition in the insurance will disappear before it’s clear that their costs won’t climb even faster than they have been.
Many small and intermediate firms now manage their health care costs by scrambling yearly to switch their coverage from one company to the next, trying to chip a little off the bill by taking advantage of the insurance company competition. It doesn’t really work—the cost increases are still brutal–but it’s all they have. The Shumlin team said that federal law obliged them to establish the exchange now. They did not oppose the Senate amendments increasing the number of insurance firms that can operate in the state.
In fact, the business community probably had a case to make in the short term, but they might be wise to keep in mind the long term. No insurance company can manage a fragmented health care system—Vermont business needs successful reform as much or more than anyone else. Still, the Shumlin team will have to manage a fractious business community over the next few years and it probably won’t be easy.
* The big ERISA companies. Federal law guarantees that companies that insure their own employees can’t be interfered with by state governments, the so-called ERISA (don’t ask) law. It may be possible for the administration to look for some kind of waiver, but that doesn’t mean it would be smart. The elephant in the middle of the room in this issue is IBM, but there are others that are critical also. IDX corporation, now a part of General Electric, is one. Lockheed Martin is another. But IBM is the key employer in the state. It has been the catalyst that brought the Vermont economy into the modern era, and it would be devastating to the state if IBM left.
The company itself has denied considering this, but Shumlin can’t ignore
the possibility. Moreover, the health care bill can’t be looked at in isolation. Shumlin led the drive to shut down Vermont Yankee, which concerned IBM because its Essex Junction plant is such a big electric power user. There was a lot of whispering in the Chittenden County business community that Shumlin would reverse himself on Yankee, but any possibility of that went glimmering when the tsunami in Japan revealed the catastrophic risks posed by shaky nukes.
The reality is that Shumlin might have to simply carve the ERISA companies out of the plan, if they insist on going. That would definitely cause a problem for Anya Rader Wallack and the designers of the system—they would have to replace to revenue flows from the IBMs and GEs. However, the only stronger commitment than single payer that Shumlin has made is to job growth and preservation and losing an ERISA employer wouldn’t do much for that.
On the other side of the coin is the prospect that if Shumlin and the dream team can build a cost-sustainable system, then the ERISA companies would be crazy not to jump into it. The big U.S. employers like IBM, GE, John Deere, Citi Bank and others, formed the so-called Leapfrog Group years ago to figure out how to get health care costs under control, and there is no evidence to date they have done so.
Some top officials from IBM met with Shumlin early this month to press their case that they should be able to continue to manage their Vermont health costs as part of their national system. They said they had learned to control their costs and the governor was quoted as saying he agreed with that. There is probably less to all that than meets the eye. It’s far more likely that what IBM controls is its health care contributions rather than the actual total health care costs for their employees. An IBM employees group seemed to suggest that when they held a pro-reform press conference several weeks after the Shumlin meeting with top IBM officials.
For the last several years, major employers in the U.S. have been moving toward replacing a defined benefit—we pay for employee’s health care—to a “defined contribution”—we pay part of the cost and the employee pays the rest. IBM might be more generous than some other ERISA firms, but it’s clear that many middle class employees of large organizations who have gold-plated health plans still pay a sizeable chunk of their paycheck for health care.
We’ll return to this issue in future articles, but for now, a single example from real life: A college professor in Vermont with an income of $90,000 pays $10,140 for health care, between contributions to the college plan and out-of-pocket expenses. Part of that is tax deductible, but the after-tax hit is still more than $6,000 per year. The employee, in other words, is picking up a major chunk of the inflation rate in health care costs.
If Vermont could get its costs under control, it would be a major boon to the big state employers as well as the rest of the employer community.
* Doctor attitudes. No stakeholder group in the single payer issue is as important as doctors. At the end of the day, they run the health care system and any reform that they perceive to come between them and their patients cannot stand. Anyone who doubts that need only look at what happened to the health maintenance organization or HMO movement in the mid-1990s, which was full of promise in the early part of the decade but by 1997 was in full retreat—by that time, movie audiences around the country would boo if an HMO was mentioned in film.
The recent kerfuffle over a survey of doctor attitudes about single payer (read the story by Jon Margolis) looked bad but probably won’t have a lasting effect. Still, effective reform will almost certainly put heavy pressure on physician incomes as well as requiring them to begin taking responsibility for the performance of the whole system, which will mean shifting the whole medical culture. The design team will have to watch this issue closely going forward.
* Medical system connectivity. In Vermont, there has historically been relatively little cross-community medical management. Most local medical communities have been fiercely independent. They will refer patients back and forth, but anyone who thinks, for example, that the physicians at Rutland Hospital think they need help from Fletcher Allen in managing their medical affairs simply hasn’t been paying attention. This situation is changing somewhat, probably under the lash of accelerating pressure for structural reform. Fletcher Allen and Central Vermont Medical Center, for example, have recently adopted a joint structure that looks like a merger. There have also been discussions between Fletcher Allen and Southwest Vermont Medical Center in Bennington. There has been no major work, however, on a full rationalization effort among the state’s 14 hospitals that would eliminate excess capacity and force the state’s physician to take direct responsibility for both costs and quality in the whole system.
A group of health policy people gathered together in the 1980s by the Windham Foundation put forward a rationalization plan that called for two umbrella organizations: the one in west would have been led by Fletcher Allen and would have included community hospitals in St. Albans, Middlebury, Rutland and Bennington, Montpelier and Morrisville. The second would be led by Dartmouth Hitchcock and would subsume Newport, St. Johnsbury, Windsor, Brattleboro, Springfield and Grace Cottage. The reaction of local hospital CEOs to this idea could be fairly described as virulently hostile. The reaction from Rutland’s CEO was volcanic. The idea died a rapid death.
One of the most intriguing ideas in the light of the current reform atmosphere is whether it would be possible to go beyond even the Windham Plan through an alliance of some sort between Fletcher Allen and Dartmouth Hitchcock. Both are high quality, but small tertiary medical centers. An alliance between them would accomplish two important results. One would be to give them a greater mass than they now possess and they could do some things they can’t now manage. Neither, for example, have a large enough patient pool to enable them to train doctors in specialty pediatrics. Combining their operations might give them a big enough population base. More importantly, a joint Fletcher Allen/Dartmouth-Hitchcock structure might exercise the kind of domination over the whole state that could enable the kind of cost and quality controls that will be necessary in a reformed system. Significant ties between the two big centers that serve Vermont may be a bridge too far, but still it would be interesting.
* Geographic complexity. Any single payer system finally adopted by Vermont could only cover the state of Vermont, at least in the fully legal sense. But there is bound to be tremendous leakage in and out of the system from neighboring states. Fletcher Allen, for example, gets a significant chunk of its business from northeastern New York. Much if not most of the tertiary care delivered to eastern Vermont goes to Dartmouth Hitchcock, which is in New Hampshire. Patients from Bennington County often get specialized care in Albany, N.Y. If the single payer design gets built on the principle of capitation, then management of this structure will get very complex.
The above are just some of the components of the design challenge facing the Shumlin team. There are some areas where they don’t need to waste much time. Call them the myths of the health care debate.
* The cost problem is the fault of the insurance companies. It isn’t. The principal reason the health insurance industry looks so bad is the rate at which the costs rise in the system. The insurance companies have been expected to control costs in the system, but they have no clue how to do that. What they can do is try to slow down the cost engine by building in bureaucratic require
ments, which simply adds expense.
* Low cost equals bad care. Wrong. Bad care is way more expensive than high quality care. This is a complex issue, but for now, doubters might look at Olmstead County, Minn., where all the care is delivered by one organization — the Mayo Clinic, arguably the finest standalone medical operation in the world. Per capita costs there are consistently among the very lowest in the United States.
* Costs rise as rapidly as they do because the population is aging. In fact, the population is aging, but it’s worth asking whether the rate at which it is aging is anywhere close to the rate at which medical costs go.
The answer to controlling costs is to shift from disease management to “wellness”—to changing people’s lifestyle away from such self-destructive habits as smoking, overeating and too little exercise. Refashioning people’s lifestyles would definitely take costs out of the system, but don’t hold your breath waiting for it to happen. It would take years, if not decades, to effect significant change, and in the meantime the obesity rate in the United States is increasing and the cost engine roars on. For the foreseeable future, doctors and hospitals will be treating disease, not focusing on wellness.
* The answer is Health Savings Accounts, a favorite of conservative ideologues. Health Savings Accounts can be a useful piece of an employer’s insurance plan in the current environment, but they have no chance of solving the cost containment problem. The idea that a sick patient is going to go shopping for cheaper treatment like a used washing machine is absurd.
* More resources means competition and thus lower prices. In a market, yes. In health care, no. The extensive research on this question goes under the heading “Supplier induced demand.” It is why historically costs in Boston with a wealth of medical resources, are much higher than in New Haven, Conn., where much of the care is concentrated at Yale New Haven system.
Here’s an early example from the Wennberg data on Vermont: In the early 1970s two local hospitals in St. Albans, which competed aggressively, merged into the Northwestern Medical Center. Competition gone. In one year, utilization rates for that hospital service area dropped in half, and stayed there.
* Government can’t run health care, single payer is socialized medicine. Well, it’s true that government can’t run health care—the problem is that no one else can either. Anyone who thinks that insurance companies can manage health care isn’t paying attention. The Shumlin Plan at this point contemplates how to pay for the care. How the money gets spent is for the dream team to figure out.
It’s a really big challenge.
END NOTE: This overview of the health reform issue has attempted to touch on the major issues and hurdles facing the reform effort. Over the course of the development process, VTDigger.org will elaborate on several of them, discussing the analysis in more depth and laying out the research and data that undergird them.
Hamilton Davis is a veteran Vermont journalist. He was the managing editor of the Burlington Free Press in the 1970s. He is also the author of Mocking Justice: Vermont’s Biggest Drug Scandal, and two chapters of the book, Howard Dean: A Citizen’s Guide to the Man Who Would Be President.