Beyond The Silver Bullet

A new Congress wants health care reform. But for lower-cost, higher-quality care, Americans first must shed the myth of the one-shot solution.
By Kerry Tremain

Four years ago, a 55-year-old catholic priest named Father John Corapi set off an FBI investigation that brought down Redding Medical Center’s chief cardiologist and cost Tenet Healthcare, the hospital’s corporate parent, hundreds of millions of dollars in legal settlements. The cardiologist, Dr. Chae Hyun Moon, had told Father Corapi that he needed an emergency triple-bypass operation. But as Corapi was being wheeled into the operating room, another surgeon examined Corapi’s x-rays and could find nothing wrong. His heart was healthy.

Stunned, Corapi called law enforcement officials, who launched an investigation. Government lawyers ultimately accused Moon of operating on scores of healthy hearts, and of fraudulently billing Medicare for the surgeries. At the time, Moon’s cardiology unit was one of Tenet’s biggest profit centers.

The Redding case drew local and national media attention, including a segment on 60 Minutes. But a more significant if quieter case had unfolded a year earlier when San Jose cardiologist Eleanor Levin astonished a convention of American Heart Association doctors with results from a comprehensive heart care program she’d spearheaded at Northern California Kaiser. Levin’s study showed that plan members were 30 percent less likely to have a heart attack than other Californians. Remarkably, Levin’s program involved no dazzling new drugs or therapies. Instead, Kaiser employed computer records to identify people at risk and ensure they received treatments known to be effective—mostly ordinary drugs such as aspirin, beta-blockers, and statins.

The two cases represent opposite poles in American health care, and one obvious moral would be for all doctors to be more like Levin and all health care plans more like Kaiser. But health care, as Hillary Clinton learned at great cost, turns out to be too complex and varied—rural and urban, rich and poor, primary and specialist care, emergencies and chronic illnesses—for single-shot solutions.

For years, Americans believed their health care system to be the best in the world. We rightfully claim leadership in new research-based diagnostic tools and new therapies that have increased our life spans and quality of life. Future U.S.-led innovations in stem cell research and bio-engineered drugs hold transformative promise. True, costs have soared, but so has our success in treating diseases that were once crippling or fatal.

In delivering health care, though, dozens of studies paint a different picture. One RAND study concluded that nearly a third of American health care is the wrong care. Only half the hospitals in the U.S., for example, reliably provide five simple follow-ups proven to reduce deaths among heart surgery patients; one is prescribing aspirin. Procedures are performed that shouldn’t be, as Dr. Moon was accused of repeatedly doing. Misadministered drugs and other errors account for up to 98,000 hospital patients dying unnecessarily every year. The extent of wrong care goes a long way toward explaining why America spends far more per capita on health care than any other advanced country—more than twice as much as Sweden, Italy, Japan, or Britain—yet ranks dead last on critical measures such as life expectancy, infant mortality, and equity.

To address these shortcomings, we’ve experienced waves of reform in the past three decades. Looking at the rising costs and the rising number of uninsured people, senators in the incoming Congress have vowed to put health care reform back on the national agenda. California lawmakers anticipate a new wave, too. Most reforms have been based on a single, big idea, such as managed care. That’s also the appeal of single-payer plans similar to those in Canada or the U.K. And although it’s not fair to say all have failed, they have all fallen short. As a result, several experts, including leaders at Berkeley’s School of Public Health, now advocate a more modular approach, combining the better elements of past reforms and adapting them to different kinds of payment and delivery systems, and to different patients and diseases. It’s a reasonable idea. Will it work?

The Redding Case, although extreme, illustrates the shortcomings of the still-dominant fee-for-service system, in which doctors make more money by doing more stuff. In multiyear studies of Medicare treatment patterns, Jack Wennenberg and his colleagues at Dartmouth Medical School have richly documented one result: The care you get depends on where you live. In fact, Dartmouth’s data registered the statistical spike in Redding heart surgeries, which rose steadily in the late ’90s, well before 40 FBI agents raided the hospital to seize patient records. By 2000, you were more than six times more likely to receive bypass surgery in Redding than in Oakland.

In contrast to the Redding case, the success of Dr. Levin’s heart-care program highlights why reformers long advocated managed care systems like Kaiser, a group plan designed in WWII for Bay Area shipworkers. (Levin’s study helped guide my personal decision to switch to Kaiser.) And as one policy expert told me, “Kaiser is finally starting to live up to its promise.”

He meant that as a health maintenance organization, or HMO, Kaiser’s primary care physicians, specialists, and other caregivers can share information and best practices (now with electronic means), more closely coordinating their efforts on behalf of a patient. Significantly, Kaiser is paid a set amount for their patients each year, giving them a strong incentive to keep their patients healthy and avoid expensive interventions—cholesterol-lowering drugs rather than heart surgery. Less is more.

But reducing what doctors do runs up against a common belief that more care is better care. Doctors, particularly specialists such as surgeons, while trained to “first, do no harm,” often share with patients the bias to do something. This more-is-better belief helped drive the public backlash against HMOs in the 1990s, which was fanned by media stories of patients denied treatments by HMO execs trying to save a buck. In the emblematic 2001 movie John Q, Denzel Washington plays a man whose son needs a heart transplant but whose insurance won’t pay for it. The desperate father seizes the hospital at gunpoint to force the surgeon’s hand.

While brandishing guns at doctors may be hyperbolic Hollywood fare, John Q’s critique was not entirely wrong. Americans are not getting the care they need, particularly the 48 million who have no health insurance. Another RAND study reported that adults receive only 55 percent of recommended care for their illnesses. And HMOs do have some incentive to restrict care (because they get paid the same whether you see the doctor 2 or 20 times a year), and to limit their risk by recruiting younger, healthier people rather than older, sicker ones.

Are americans then caught between getting too much “care,” like those Redding heart patients, or too little, like Denzel’s fictional son? As it turns out, the Kaiser and Redding cases do share one trait: Doctors determine the care someone receives, and doctors are paid directly by health plans. Sometimes the patient doesn’t even see the bill. After the HMO backlash, some reformers advanced a new idea called “consumer-driven” care. If we gave consumers more choice in where they spent their money, the reasoning went, their choices would drive down prices and drive up efficiency as relentlessly as they do in markets for DVD players and blue jeans. If the problem is that health care is overly “supply-driven” (regions with more surgeons do more surgeries), then perhaps the solution is a more “demand-driven” system. If consumers have the illusion that health care is a free or unlimited resource paid for by others, then give them more control over their money and they will spend more wisely and responsibly. They’ll think twice before requesting a test or treatment of dubious value.

Value is the key word. Advocates of consumer- driven care believe that individuals will naturally choose to pay more only for services of higher quality. You might choose Wal-Mart care or Nordstrom care, but you’d know what you’re getting, which is far from the case now. Again, the Dartmouth data are revealing. The studies show that Medicare payments vary widely, even wildly, with little relation to quality. A 2005 study, for example, reported that Medicare payments to doctors and hospitals in California for chronically ill patients during the last two years of life ranged, depending on where you live, from $35,000 to $106,000 per patient. (End-of-life care is studied because the outcome, i.e., death, is the same.) The average for inpatient hospital care in Los Angeles was the highest in the state, and 67 percent more costly than Sacramento, the lowest.

Since Medicare is the closest thing the U.S. has to a single-payer system, these studies also suggest that simply having the government insure everyone—although clearly better than having so many people without coverage—is not a silver bullet for our health care problems. But these data do have a silver lining. They bolster the conclusion that by making health care better, we might also make it less expensive. Wennenberg estimates that over a five-year period improved hospital efficiency could have saved Medicare $1.7 billion in Los Angeles alone.

With consumer-driven care, reformers argued, individuals could use their purchasing power to drive out that waste. Interest in consumer-driven care also dovetailed with the trend toward entrusting and empowering people with more of their own care. Increasingly, patients are showing up at their doctors’ offices with information downloaded from the Internet—a good thing. Consumers have become far more interested in preventive care and over-the-counter devices such as those for monitoring blood pressure or sugar. Lifestyle changes—quitting cigarettes, eating better, exercising more—and self-monitoring for warning signs are ultimately more important than doctor visits in managing chronic illnesses such as diabetes, arthritis, and heart disease. The goal with these kinds of illnesses is to keep them from tipping into something urgent and costly, such as being hospitalized for a raging infection. Currently, the chronically ill make up most of the 20 percent of patients demanding 80 percent of American health care costs, but given the right information, tools, and support, they can greatly assist in their own care. (Some futurist tools are in the works, too. One is a “smart wall” designed into your home that reads out your vital signs or reminds you to take your medicine. Implantable devices—a medical chip in your arm—will be able to tell your blood pressure or cholesterol levels, conveying instant feedback through a read-out device such as a cell phone.)

To encourage consumer-driven care, employers and health plans designed new insurance products, including Health Savings Accounts, that give employees a set amount for health care and broad discretion in how they spend it. Employers like such plans because they make their business expense for health care more predictable, and limit their risk. Young people tend to like it because they don’t often get sick and can usually roll over amounts they don’t use. But that’s a problem. HSAs shift the risk, and the costs, to the older and sicker employees.

There are other drawbacks, too. When you buy a toaster or an airline ticket, it’s easy to comparison shop on the Internet. But try going online to compare gastroenterologists. True, there is far more quality information available on doctors and hospitals than even a few years ago, but it remains sparse and hard to translate into meaningful decisions. Besides, it’s a lot easier to order a book from Boston than to fly there for a mammogram. Health care is still largely local. Without informed and easily available choices, consumer-driven care becomes not a panacea but a chimera.

Once again, Americans seem caught between bad options—either a doctor-knows-best, demand-driven system and its attendant wasteful expense, or an inadequately supported consumer-driven system that exposes us to unacceptable personal risks.

Stephen Shortell believes there must be better choices. Shortell, the dean of Berkeley’s School of Public Health, has the wiry frame of a runner, topped with a puff of gray hair. His placid demeanor and flat tone of voice suggest someone who has chaired scores of committee meetings. Indeed, Shortell is well known in health care reform circles. He’s spearheading a California Health Strategy Summit, part of a longer-range plan to improve health care quality and access in the state.

“There is no single model that is going to improve the delivery system in California or across the country,” he says. “But there are ingredients everyone can adapt and use.” One is Pay for Performance. A few years ago, Shortell helped broker an arrangement with California’s big insurance companies to boost pay to doctors and hospitals that demonstrate, according to agreed measures, that they are providing superior care. At Kaiser, about a quarter of doctors’ pay depends on meeting quality performance measures. Doctors paid by Blue Cross also receive such incentives. Pay for Performance helps cut through the most difficult issue in health care—the problem highlighted by the Redding case—of aligning incentives with what we all want: better quality care.

I once heard a speaker colorfully compare our traditional system to a Skinner Box: The “pellet” drops only when the doctor is “in the box” with a patient, i.e., when there’s an office visit. But often office visits aren’t that useful or necessary, never mind the irritating waits. Although many people would prefer to email or phone their doctors with questions or concerns, few insurers will pay for that. A few years ago, I visited Greenfield Health, a clinic in Portland, which aspires to create a model clinic from the ground up. There, Dr. Chuck Kilo, an indefatigable advocate for quality improvement, told me that nearly 80 percent of the interactions between doctors and patients at Greenfield occur by email or phone.

An even more important innovation is the electronic medical record, or EMR. When I visit my Kaiser doctor, he pulls up my record on a computer monitor in the exam room. The EMR lists my tests and treatments, and we discuss them together. If I’ve received a treatment in some other part of the hospital—such as the physical therapy I recently underwent for neck pain—it’s in the record for both of us to see. The EMR is another modular reform that Shortell is eager to see adopted. Here again the U.S. is far behind comparable countries. A study released in October showed that only 28 percent of office-based American doctors use EMRs, compared to 98 percent in the Netherlands and 89 percent in the United Kingdom. Only Canada had a lower adoption rate, at 23 percent. The study also showed the two countries lagging in using computers for decision support, such as alerts about harmful drug interactions, or to remind patients to get follow-up care.

The modular reforms Shortell advocates all require better use of information technology. Using IT to spread information on treatments with proven efficacy—so-called evidence-based medicine—would make good health care less dependent on where you live. Its relative success in Europe offers hope for its success here. And as Shortell points out, EMRs are critical not only for individual patient care but also to enable us to collect data that help doctors improve care. Imagine the healing power of a national database of information about treatments and outcomes— data now lost every day because they are buried in paper records in clinics. That’s exactly the type of information-gathering Eleanor Levin used to establish the Kaiser heart program.

“We think we know some of the ingredients of a more cost-effective delivery,” Shortell says. “How to get there politically, policy-wise, is the challenge.”

School of Public Health policy expert James Robinson describes himself as a “glass-half-full” guy, but his conversational tone is “yeah, yeah, heard it all before.” Robinson is particularly weary of hearing that our health care system is in crisis. “If that’s the case, it’s been in crisis for 50 years,” he says. He is more likely to point out the things we have today, such as better arthritis drugs, that weren’t available 20 years ago. The problem, he says, is that no one is willing to give up anything—to pay more to get a knee replacement, or agree to higher taxes to insure everyone.

Like Shortell, Robinson is skeptical of one-size-fits-all solutions, but his modular solutions have a different character. Not only do we need modules, such as EMRs, that can be used across systems, we also need different ways of paying for different types of care. “Health care is not one thing,” he says. “Getting drugs at Walgreen’s and lying in a coma in an ICU aren’t the same.”

Robinson distinguishes care that is sensitive to either doctor or patient incentives, or both, or neither. Insurance payments to both can therefore be tailored for maximum efficiency. An appendectomy is an example of neither. If your appendix ruptures, there’s only one choice: take it out or die. Health insurance should fully cover that. Although it’s the kind of thing that is perhaps most familiar, from ER and Grey’s Anatomy, in fact, “very little medical care is like that,” he says.

When it comes to treatments such as generic or brand drugs, “Doctors should not care,” Robinson says. “But patients usually think the one they saw on TV is the good one, and want it.” That’s fine, he says, but then they should pay more. Certain procedures such as implanted knee replacements show huge regional variation among doctors. Physicians also overuse MRIs, sometimes to make the payments on their very expensive machine, but patients also love them. Slap on higher copayments and deductibles, and wasteful use will go down, says Robinson.

By combining aspects of managed competition (i.e., larger integrated care systems like Kaiser competing with one another) and consumer-driven care, Robinson believes delivery systems that serve different purposes will evolve, driving up the value of care. To some extent, that’s already occurred. Some outfits do only a single procedure, such as laser eye surgery or even heart transplants, and can do them quickly, with more consistently positive results. Higher quality, lower cost.

Shortell thinks that adapting reforms like Pay for Performance and EMRs across the broad spectrum of American practice will also drive higher quality at lower cost—and possibly make it easier to offer more health care to more people. Robinson agrees, but with an asterisk. If there were a lot more money to be made by squeezing out waste, it would have been done, he contends.

Still, the Dartmouth data and the world health statistics suggest that we are not yet spending our existing health dollars optimally. These studies should be required reading for state and congressional health care reformers. As the population ages, Medicare could become an even bigger driver of change by encouraging the reforms that Shortell, Robinson, and others advocate. In the meantime, the wise consumer is the one who knows not only what a health plan offers but also how it pays its doctors, and how its treatment patterns compare to others. Knowing just those facts could have spared those patients in Redding a lot of heartache.

Filed under: Law + Policy
Share this article:
Google+ Reddit

Add new comment