Best Care Anywhere

by Phillip Longman

  And so we wind up making choices in health care based on whether a doctor is in the “preferred provider network” and is taking new patients. Or maybe your best friend recommended someone. Or perhaps you selected a doctor who agrees with your diagnosis and refills the Ambien prescriptions you want. We use criteria such as liking a doctor’s bedside manner or the fact that all the rich people in town go to this hospital. According to a recent survey, Americans spend twice as much time researching a car or computer purchase as they do selecting a doctor, which is not surprising given the scarcity of useful information about any provider.20 Those of us who are truly diligent might consult rankings of different hospitals such as those published by U.S. News & World Report—as Robin and I did when we needed to decide on a cancer clinic—or consult similar Web sites. But such surveys rely primarily on surveys of reputation—that is, on word of mouth—and give little weight to objective, statistical measures of quality. If they did, as we’ll see in the next chapter, most of today’s highest-ranking hospitals would be revealed as among the nation’s most dangerous and ineffective.

  Not knowing how to judge the quality of care that doctors provide, we place inordinate value on our ability to switch doctors. If one disappoints us for whatever reason, we can move on to another. “Choice of doctor” has become in most Americans’ minds the single greatest measure of the quality of any health-care plan—a sad irony indeed since the poor quality and fragmentation of the U.S. health-care system are ultimately both causes and consequences of our insistence on choice above all else.

  And thus we see results like what happened in Cleveland during the 1990s. There, a well-publicized initiative sponsored by local businesses, hospitals, and physicians identified several hospitals as having significantly higher-than-expected mortality rates, longer periods in the hospital, and worse patient satisfaction. Yet not one of these hospitals ever lost a contract because of its poor performance.21 To the employers buying health care in the community, and presumably to their employees as well, cost and choice counted for more than quality and safety. Unfortunately, as we’ll see in the next chapter, the cost of this market failure in money, injury, and death can only rise as American medicine adopts more and more expensive, complicated, and often ineffective or dangerous technologies.

  EIGHT

  When Less Is More

  Marjorie Williams died of liver cancer in early 2005, leaving behind a bereaved husband, two young children, and a network of loved ones and admirers who filled Washington’s National Cathedral to the last pew at her memorial service. Her legacies are many, including a brilliant career as a political writer for Vanity Fair and columnist for the Washington Post. Yet she will perhaps be best remembered for her sharp and revealing descriptions, written near the end of her life, of her struggle with cancer, and about what it’s like to be treated for it at some of America’s very “best” hospitals.

  In the first two and a half years of her illness, Marjorie received treatment from thirty-two doctors in six hospitals. She and her husband, Tim Noah, also an accomplished journalist and a friend of mine, were shrewd and well connected. They consulted with many medical luminaries in search of the very best care and treatments. And yet their quest for quality in health care was as disappointing as Robin’s and mine.

  “My most memorable brushes have been with an eminent surgeon,” Marjorie wrote in her next-to-last column for the Washington Post, “whose method is to stride into the examining room two hours late, pat your hand, pronounce your certain death if he can’t perform an operation on you, and then snap at your husband to stop taking notes, he can’t possibly follow the complexity of the doctor’s thinking.”

  In the same column, Marjorie described another memorable moment on her journey.

  During one hospital stay, as I sat in a wheelchair outside Radiology waiting to be pushed back to my room, I began idly flipping through my chart. A young female doctor-in-training I had never seen before stopped in front of me and said, “You know, you really shouldn’t be reading your chart.” I thanked her for her advice and continued reading. She repeated her admonition. I explained that I was 43 and couldn’t possibly read anything worse there than I had already been told by five real doctors. Upon which she actually wrested it from my grasp. (From this I learned always to go to a stall in the ladies’ room when I want to read my chart.)1

  Such anecdotes by themselves prove nothing. But Marjorie’s experience helps us to visualize the reality behind a very odd truth in American medicine. Generally, the more prestigious the hospital you check into, and the more eminent and numerous the physicians who attend you, the more likely you are to receive low-quality, or even dangerous and unnecessary, care.

  America’s Worst Hospitals

  The evidence for this assertion is as overwhelming as it is threatening to the medical establishment. The first hint of its truth came in the 1970s, when researchers John E. Wennberg and Alan Gittelsohn noticed strange patterns across regions in how doctors treated patients. By combing through old medical records, Wennberg and Gittelsohn discovered wide and seemingly inexplicable differences in how often doctors diagnosed people with peptic ulcers, for example, or in how often people received such operations as tonsillectomies.2 In the town where Wennberg’s kids went to school, Waterbury, Vermont, 20 percent of the children had their tonsils out by age fifteen; but in next-door Stowe, 70 percent of the children got tonsillectomies.3

  Differences in socioeconomic status could not explain the contrast. Nor was it plausible to believe that the kids in Stowe were far more in need of tonsillectomies than were kids in Waterbury. The wide variation in practice patterns suggested that something besides scientific rationality was at work in deciding which patients received what care—an idea that at the time was as radical as it was novel. Doctors, after all, were supposed to be professionals who put their patients’ interests before their own and who administered care according to the dictates of science.

  Gradually, Wennberg and other researchers, most of them on the faculty of the Dartmouth Medical School, found clever ways to tease out what was going on, and the emerging truth was grim. For most Americans, the two biggest determinants of what kind of diagnoses and treatments they receive are how many doctors and specialists hang a shingle in their community and which one of them they happen to see. The more doctors and specialists around, the more likely they are to be “upcoded” as being sicker than they are, and more tests and procedures will be performed.4 And the results of all these extra tests and procedures? Lots more medical bills, exposure to medical errors, and a loss of life expectancy.

  This last conclusion was truly shocking, but it became unavoidable when Wennberg and others broadened their studies. They found it’s not just that renowned hospitals and their specialists tend to engage in massive overtreatment. They also tend to be poor at providing critical but routine care. For example, Dartmouth researcher Elliott S. Fisher has found that among Medicare patients who share the same age, socioeconomic standing, and health status, their chance of dying in the next five years is greater if they go to a high-spending hospital in their community than to a low-spending hospital.

  One reason is that patients in high-spending hospitals with lots of specialists and high technology are also less likely to receive many proven routine treatments. For example, standard, evidence-based medicine has identified aspirin as a highly effective treatment for heart attack victims. Yet, in the highest-spending hospitals, only 74.8 percent of heart attack victims receive aspirin upon discharge from the hospital, as opposed to 83.5 percent in the lowest-spending hospitals. This may be one reason why survival rates for heart attack victims are actually higher in low-spending hospitals than in high-spending hospitals.

  Patients in high-spending hospitals are also far less likely to receive flu vaccines (48.1 percent versus 60.3 percent) as well as such routine preventive measures as pneumonia vaccines, Pap smears, and mammograms. This general lack of attention to prevention and follow-up care
in high-spending hospitals helps to explain why not only heart attack victims but also patients suffering from colon cancer and hip fractures also stand a better chance of living another five years if they stay away from “elite” hospitals and choose a lower-cost competitor. By doing so, they not only gain a better chance of receiving effective preventive and follow-up care, but they also gain a better chance of avoiding unnecessary and often dangerous surgery and tests. Given this unexpected reality, it is perhaps not surprising that patient satisfaction also declines as a hospital’s spending per patient rises.5

  This isn’t a statistical fluke. Sure, even among equally ill patients, those who are more aware of the risks of their illnesses may move closer to more prestigious and expensive hospitals. But, while that may be a small factor, the relationship between more money spent per patient and mortality exists among teaching hospitals themselves, even within the same region. And it also applies equally to patients who moved recently and to those who did not.

  From evidence like this, Fisher and other doctors estimate that if medical practice in the highest-spending hospitals could be brought in line with medical practice only in the lowest-spending hospitals, financial savings of up to 30 percent could be achieved in Medicare, thereby preserving the solvency of its trust fund indefinitely. And, not only would we have that happy result, but Medicare patients would receive less dangerous and higher-quality care as well.6

  Results like these have been repeatedly confirmed by a cascade of similar studies.7 Tellingly, doctors themselves seem to know instinctively the truth behind them. In 2006, when Time magazine had the brilliant idea of asking doctors what scared them most about being a patient, three frequent answers were fear of medical errors, fear of unnecessary surgery, and fear of contracting a staph infection in teaching hospitals.8

  Insurance Against Overtreatment

  Once one becomes sensitized to the degree of mistreatment and overtreatment in U.S. healthcare, the perversity of our long efforts to expand access to our broken system comes into harsh focus. It turns out, for example, that the quality of care received by the uninsured is, by some important measures, better than that of insured Americans.

  Yes, that’s right. A landmark RAND study, published in the New England Journal of Medicine, has found for example that uninsured patients receive only 53.7 percent of the care experts believe they should get, which is a tragedy. But according to the same study, patients with private, fee-for-service insurance are often even less likely to receive appropriate, evidence-based treatment. Indeed, among Americans receiving acute care, those who lack insurance stand a slightly better chance of receiving proper treatment than patients covered by Medicaid, Medicare, or any form of private insurance.9

  How can this be, you might ask? It turns out that being uninsured paradoxically provides insurance against some of the leading causes of death and injury in the United States. Overtreatment, for example. Once patients who are unable to pay are in a hospital’s door, they cost it money until its doctors make them well enough to leave. There is no incentive whatsoever to give them tests or treatments they don’t need. Since about 30 percent of all health-care spending in the United States goes for overtreatment—much of it dangerous—this is no small advantage.

  Also, the less time you spend in a hospital, the less likely you are to contract a staph infection, or to be killed or maimed by medical errors. Similarly, not having health insurance tends to reduce your exposure to ionizing radiation from imaging devices. With the possible exception of mammography, these scans provide no proven benefits yet are a growing cause of cancer, particularly among persons who have received repeated radiation exposures—typically well-insured elderly patients.10 As a recent commentary in the New England Journal of Medicine observed, “Exposure to even moderate degrees of medical radiation presents an important yet potentially avoidable public health threat”—one that the uninsured by and large do not have to worry about.11

  You might think that at least insured Americans get better preventive care, what with their $10 co-pays to see a physician. Also not true. According to the RAND study, patients with private fee-for-service insurance receive only 53.3 percent of the preventive care they need, while those without insurance receive 54 percent.12 Even HMOs, though they do slightly better on this score, face the reality that their patients are constantly churning from one plan to the next and so, as we’ve seen, don’t have a business case for investment in true prevention. Meanwhile, fee-for-service providers will happily charge you for a high-radiation, full-body, three-dimensional, sixty-four-slice (cancer-causing) scan of no proven diagnostic value, and call it prevention.

  Harvard biochemist Lawrence J. Henderson, one of the Progressive Era’s most renowned men of science, once observed that it wasn’t until somewhere around 1911 that the progress of medicine at last made it possible to say that “a random patient with a random disease consulting a physician at random stood better than a 50-50 chance of benefiting from the encounter.”13 Sadly, the RAND data and a growing pile of other studies suggest that this statement is essentially still true today, for those with or without insurance.

  Now perhaps it becomes more evident why the VA’s health system keeps coming out on top in measures of health-care quality. Although the VA is exceptional in its use of evidence-based medicine, information technology, disease management, and root-cause approach to patient safety, its superior performance is also a measure of how fragmented and ineffective its competition is. America’s medical elites are very good at attracting money and prestige, and they have a huge technology arsenal with which they presume to attack death and disease. But they have no measurable positive medical results to show for it in the aggregate; in fact, many signs indicate that they are providing a lower quality of care than the much-maligned HMOs and assorted “St. Elsewheres.”

  Roemer’s Law

  How can we possibly explain such strange findings? The beginning of all wisdom, when it comes to understanding the business of health care, is to understand that, in this realm, supply often creates its own demand. This of course seems counterintuitive and inconsistent with our experience. After you’ve waited two weeks to see a specialist or to have a PET scan, it is hard to imagine that there is anything but an acute shortage of medical professionals and money to equip them.

  But the real reason you have to wait to see a specialist is not that there are too few of them but that there are too many. Specialists who move into a community induce demand for their services even if their treatments provide little or no benefits, or indeed are harmful to most patients. For a similar reason, hospital beds are almost always full because their supply is so great. Add another hospital bed, and one way or another, the local health-care system will find a way to fill it. The phenomenon is known as Roemer’s law, after the late health-care economist, Milton I. Roemer, who first described it in the late 1950s and early 1960s.14

  There are two basic explanations for why Roemer’s law arises. The first involves a short circuit in the normal way supply and demand adjust to one another. In most realms of the economy, demand eventually meets supply through changes in prices. If GM produces more cars than people need or want to buy, it winds up cutting prices and offering rebates until the cars are sold. Eventually, if GM can’t cover its costs, it will make fewer and fewer cars, until it eventually goes out of business—or, as it happens, gets a bailout.

  By contrast, health-care prices don’t drop when there is excess supply. One reason is because Medicare effectively controls health-care prices. Whatever Medicare will pay for a procedure becomes the benchmark for what other insurers will pay as well. This might not be a problem if market forces determined Medicare’s reimbursement rates for different procedures. But instead, it is an obscure bureaucratic and political process, dominated by the American Medical Association that for years has wound up setting rates and thus ensuring that whatever costs the health-care system accrues collectively are covered.15

  So the more capa
city the health-care system adds in the form of hospital beds, specialists, and high-tech equipment, the more money comes into the system. Some parts may wind up being more amply rewarded than others according to their success in influencing Medicare’s rate schedules, but for the system as a whole, there is little or no check on excess supply. Instead, excess supply is absorbed through overtreatment and inefficiency. In order for a hospital to go broke and shut down under these conditions, it takes extreme mismanagement or a decline in the surrounding community that creates large volumes of uncompensated care. If Medicare reimbursement rates prove insufficient to cover the cost of one kind of treatment a hospital offers—say, managing a diabetic’s care—then a hospital can divert its resources to treatments for which reimbursement rates are more lucrative, such as heart surgery or chemotherapy.

  The second reason Roemer’s law applies is the stunning lack of scientific knowledge about which treatments and procedures actually work. Doctors are highly trained professionals, and most are committed enough to their calling that they would never knowingly subject patients to treatments and tests that are straightforwardly and unambiguously excessive. The problem, however, is that medical textbooks are silent about what constitutes appropriate care for patients with many different illnesses, particularly for those nearing the end of life. For example, medical textbooks offer no evidence-based clinical guidelines for how often doctors should schedule such patients for return visits, when they should be hospitalized or admitted to intensive care, or what palliative care they should receive. Nor do medical textbooks offer clear guidelines, grounded in science, about when a doctor should refer a patient suffering from a specific condition to a specialist, much less when it is appropriate to order a diagnostic or imaging test.