Reformers have expended considerable effort to change the way Medicare buys health care for millions of Americans, finding a wide range of support across the political spectrum. But Medicare’s current structure may doom these reforms. Some primary-care doctors fear that the skewed incentives in Medicare’s pricing system will get built into any new system, whether it is “accountable,” “affordable,” or “value-based.” This fear is well-founded. If special interests buy into a reform, it’s because they imagine that they will continue to profit as much as they did under the old system. If the system changes, yet the current problems continue, what’s the point of reform?
For two decades, doctors have controlled the process that determines how Medicare pays them, with predictable results. Specialists encourage greater numbers of procedures, devaluing primary care. Primary-care doctors are the medical equivalent of general contractors. They can speak the specialists’ language and weigh the competing priorities. Sometimes you can save a life with aspirin or fix back pain with yoga, but no one gets rich offering that kind of advice. While primary-care doctors earn a better living than does the general population, they are losing ground to specialists with just a few more years of training. As a result, fewer doctors are entering primary care, leaving fewer unbiased experts qualified to tell specialists when to put down the scalpel. Going into primary care is starting to resemble getting an MBA but then joining the Peace Corps—noble, perhaps, but financially ruinous. Policymakers must fix this imbalance before any real Medicare reform can occur.
Medicare Part B covers doctors’ visits and outpatient care costing more than $170 billion. In the early 1990s, Part B had a problem: the historical prices that the program used to pay doctors had lost connection with market realities. Doctors doing procedures, mostly specialists, got paid too much, while cognitive medicine—the problem-solving done in primary care and some of the less glamorous specialties—got paid too little. Procedures once strenuous, risky, and time-consuming—and therefore expensive—had become routine. Cataract surgery was dangerous when it was new, says William Hsiao, a Harvard health economist who developed the replacement for the old Medicare reimbursement system. At first, the procedure required a three-day hospital stay. When Hsiao did his study for Medicare, though, the procedure could be done in about an hour; now it takes even less time. “So your charge does not reflect the amount of work you have to do or even the risk you are taking,” Hsiao said. He pointed out that a physician could charge about $40 for a 20-minute office visit, while an ophthalmologist could charge $2,000 for a cataract operation that took about an hour.
Medicare’s current system for paying physicians dates back to the early 1990s. To fix the historical-prices problem, Congress ordered the creation of an elaborate price-setting formula based on the relative value of procedures (not the value to the patient, but the value of the inputs). Hsiao worked with hundreds of doctors, economists, and other experts to develop the resource-based relative-value scale (RBRVS). Hsiao measured how doctors spent their time—and the intensity of that time. This so-called “work value” accounts for about half of what a doctor is paid by Medicare. The other half covers practice costs. Medicare also pays for malpractice-liability insurance, but this accounts for only a small fraction of fees. Medicare adjusts what it pays doctors based on geography and uses a conversion factor to determine how much to pay doctors for each diagnostic code. Until 2014, there were more than 10,000 codes. With the implementation of ICD-10, an update to the World Health Organization’s medical-classification conventions, that number has exploded to 140,000.
Measuring a doctor’s time and effort takes its own share of time and effort. The George H.W. Bush administration didn’t want the government to play that role, so the American Medical Association assumed it. The AMA created a committee of doctors heavy on surgical specialists. Thirty-one doctors help Medicare set prices for every doctor in the country. Until recently, it was an even smaller group, formally called the Relative-value-scale Update Committee (RUC).
The RUC’s accepted recommendations have significant implications on Medicare spending—and therefore on the federal budget. They also affect people who aren’t on Medicare, because private insurers base their payments on what Medicare pays. “Why? Because there’s no other metric,” says former Medicare administrator Bruce Vladeck—indeed the first administrator of what is now the Centers for Medicare & Medicaid Services (CMS) after RBRVS was implemented. Vladeck remembers seeing an estimate that the new “fee schedule had actually moved several billion dollars from specialists” to cognitive practitioners. Later in the 1990s, after he left CMS, those improvements started to come undone.
Hsiao’s team estimated the amount of time and effort that went into thousands of procedures. They knew from the outset that these estimates would not remain valid forever. The changing level of work involved in medical procedures created the need for a new system. Hsiao said that officials in Bush the Elder’s administration viewed his project as “government interference in the market” and didn’t want to spend any money on it.
Gail Wilensky, a senior fellow at Project HOPE, ran Medicare and advised Bush while Congress implemented RBRVS, over the administration’s objections. She never expected Congress to implement RBRVS for all 10,000 codes. The resource-based approach, she said, tries to “do more rationally and logically what markets do all the time.” RBRVS reminded her of the input-output models used by the Soviet Union: “They were elegant and beautiful, but not functional.” On its website, the AMA says that the RUC “was established in the course of the AMA’s normal activities and as a basis for exercising its First Amendment right to petition the federal government.”
The first loyalty of most doctors, Hsiao said, is to their specialty society or college, because these groups provide certifications, continuing education, and publication in medical journals. The AMA doesn’t have that same relevance for many doctors. Today, only about one in four American doctors belong to the AMA. “So the AMA was trying to find some way to regain its influence and control over all the specialties,” Hsiao said. “What’s a better way to control you than through your pocketbook?” Thus the RUC was born.
“You’re basically asking the people who get paid how much they should get paid,” said Brian Klepper, a co-founder of Replace the RUC!, an advocacy campaign. Klepper is surprised that more policymakers aren’t paying attention to the RUC because it is “the beating heart of the healthcare cost crisis, which in turn is the beating heart of our fiscal crisis.”
Medicare Part A, the hospital-insurance program, pays for about 1,000 different diagnosis-related groups, essentially bundles for treating a patient. Medicare Part B, which uses the RBRVS, has a more a la carte approach, reimbursing doctors for more than 100,000 different codes. Once the Centers for Medicare & Medicaid Services determine the relative value of each procedure for the next year, it multiplies those values by last year’s volume to estimate the total relative-value units for the year. It then divides the pool of money for physician payments by the estimated number of RVUs to determine the dollar value of each unit.
This approach makes increasing the relative value of codes, or adding new ones, into a zero-sum game: if the relative value of one procedure goes up, every other procedure gets paid a little less. Some procedural specialties can make up the difference by increasing volume. For example, one study published in Annals of Internal Medicine reported that, while Medicare reduced the payment for a colonoscopy by 15 percent between 2001 and 2004, the number of colonoscopies increased by 30 percent over the same period. Steven Weinberger, CEO of the American College of Physicians, a medical society for internists, said that primary-care doctors have limited ability to increase volume. “You can’t tell a patient, ‘Now you have to be quicker recounting things to me,’” Weinberger said. “How much time a patient takes, you have no control over that.”
Though it’s intended to control Medicare spending, the zero-sum budget creates the opposite incentive for individuals. “No matter what they do,” Wilensky says, “nothing one physician or physician practice can do is going to influence the aggregate.” This means there is no penalty for an individual who increases volume to make up for lower payments. “The good guys get screwed and the bad guys get more money,” she said. More than two decades after Hsiao and his colleagues created this elaborate system, Congress—or someone—needs to fix Medicare’s prices once again.
“I always assumed specialists make more money because they train longer,” said Robert Clark, a primary-care doctor in Georgia. “The real reason is because they got more folks on the committee.”
Supporters of more government involvement in health care often hold up Medicare’s low administrative costs as a virtue, but Medicare probably should spend more on administration. Because its administrative costs are so low, the program effectively outsources pricing decisions to the AMA. How does the AMA measure physician time and effort to guide billions in taxpayer spending? By surveying at least 75 doctors for procedures done more than 1 million times each year; less frequent procedures, occurring at least 100,000 times a year, require 50 respondents.
The AMA says that the RUC is nothing more than a sophisticated way of exercising its First Amendment rights to influence government. Critics find the relationship cozier than that. In 2011, Georgia family doctor Paul Fischer and some of his colleagues, including Clark, sued Medicare. They argued that the RUC should follow the rules Congress set out for groups that formally advise federal agencies. They lost in two federal courts, where judges ruled that Congress protected Medicare’s fee schedule from legal challenges.
Earlier in his career, when he was an academic, Fischer published a study on the impact of tobacco advertisements on children. After showing children knew Old Joe, the mascot for Camel cigarettes, as well as Mickey Mouse, he started a family-medicine practice near Augusta, Georgia, that has grown to 29 doctors at seven locations. “I spend much of my day talking people out of having invasive procedures done to themselves that they don’t need,” Fischer said. “The same way they were talked into smoking Camel cigarettes, they’re now getting talked into getting unnecessary procedures. We had all these cardiologists because we had a lot of cardiac disease when people were smoking and had high cholesterol, and now they sit around not doing anything. So that’s why we have to advertise, ‘Get to the ER when you have chest pain.’”
He said there are many causes of chest pain—reflux, infection, pain from coughing, anxiety—and most of them are not life-threatening. “That’s the marketing message. Where does it come from,” he asked. “Every billboard has a picture of a big heart and a local hospital name.” He believes that it’s wrong for the federal government to delegate price-setting for billions in spending to an advocacy group. He makes enough money, but he fears that there won’t be enough doctors in primary care if medical students see an easy opportunity to earn more in another specialty.
Doctors don’t have to sit around counting money, says Clark, a partner physician in Fischer’s practice and co-plaintiff. “The system is designed to do it for them. He can just follow the guidelines. The doctor who is not following the guidelines all the time is the one who’s probably saving the most money long-term.”
What is primary care? Fischer insisted that I see for myself. One Saturday, I followed along as he treated more than 30 patients. Fischer did a well-child visit for a six-month-old, saw a woman in her seventies after a fall, and vouched for the health of a young missionary preparing to leave the country. He treated a man recently discharged from a mental hospital and another, in his thirties, struggling with heart failure.
One patient, a woman who recently went to the hospital with bronchitis, represented what Fischer says is wrong with medicine. She has five stents but has never experienced chest pain. She said that her cardiologist found an artery 75 percent blocked.
“The vast majority of people getting stents don’t need them,” Fischer said. “The cardiologists are like plumbers. They see a blockage, they want to open it, despite the fact that the research is very clear that doing that does nothing to benefit the patient.” He said this patient was never in specific danger, and that stents actually increase the risk of a heart attack originating in adjacent areas. “If I had a 75 percent blockage,” he said, “I would take statins, aspirin and exercise.” Fischer gets paid to give patients that advice, but a cardiologist gets paid much more to insert a stent.
In reaction to media attention and perhaps Fischer’s lawsuit, the AMA has made some changes to the RUC. Its vote tallies are now public, though the votes of individual members are not revealed. The AMA has added two seats to represent geriatrics and primary care.
The CMS has also recognized the need for change. Robert Berenson, an internist and former member of the Medicare Payment Advisory Committee, says that the CMS previously accepted 90 percent of RUC recommendations, but only 60 percent after a recent update. “They have a pretty good process but the results are distorted prices, so something’s wrong,” said Berenson, now an Institute Fellow at the Urban Institute.
The CMS has contracted with the Urban Institute to collect better data on how long procedures take. Berenson said that this is necessary because, for example, the time estimates for doing colonoscopies are inflated by 300 percent, interpreting an echocardiogram by 500 percent, and reading an EKG by 1,000 percent, or more. “The times lack face validity,” he said. “They are inflated hugely.”
Take reading an electrocardiogram or EKG. “What has changed is automation,” Berenson said. Doctors used to get a tracing and had to measure the interval between the waves. “Now it’s just a printout.” The computer even suggests an interpretation of the results. The current estimate for the amount of time it takes to read an EKG is about five minutes. “That’s wildly exaggerated,” Berenson said. “We’re talking five seconds versus five minutes. There shouldn’t even be a payment for interpreting an electrocardiogram. It’s a trivial activity.” Cardiologists are having nurse practitioners handle office visits while the doctor “is reading off an automated report.” The cardiologist is getting paid more for the value of his time and for an inflated amount of time. “He shouldn’t get a double payment.” The exaggerated estimate for colonoscopies is making the procedure so lucrative that some gastroenterologists don’t want to see sick patients anymore, he says. “That’s not good for patient care.”
Medical students are reacting to these incentives when choosing a specialty, and fewer doctors are going into primary care. More than 80 percent of internal-medicine residents go on to sub-specialize, rather than remaining in primary care. Berenson coauthored a 2007 study showing that the gap between primary care and specialty income had grown from a little more than $80,000 in 1995 to more than $135,000 in 2005, meaning specialists earned nearly twice as much as generalists. The same study, which appeared in the Annals of Internal Medicine, cited a study showing the share of medical school graduates going into family medicine fell from 14 percent to 8 percent between 2000 and 2005. “Although incomes of primary care physicians are far higher than the earnings of most persons in the United States, and the public has little sympathy for physicians who cry poor, the lower income of primary care physicians is a major factor leading U.S. medical students to reject primary care careers,” Berenson and his coauthors wrote.
Even incremental fixes to the existing system could lead to dramatic improvements. Medicare needs to devote more time and resources to planning its annual spending; right now, it makes these decisions over the equivalent of seven days. This leads to inefficiencies and leaves the CMS dependent on the AMA for pricing advice. In a 2006 report, MedPAC, a congressional advisory group on Medicare policy, recommended changes to the process of updating relative work values. “CMS should play a lead role in identifying misvalued services so overvalued ones are not ignored,” the report said. “CMS could gain the requisite expertise by establishing its own group of experts, separate from the RUC, to help the agency conduct these and other related activities.”
That would be a good start. Here are some other reforms that the CMS should seriously consider.
Acknowledge that conflicts of interest can never be completely eliminated, and increase transparency to compensate. Last year, the National Commission on Physician Payment Reform, a group convened by the Society of General Internal Medicine, made 12 recommendations to improve the way doctors get paid. The RUC “should make decision-making more transparent and diversify its membership so that it is more representative of the medical profession as a whole,” the commission said.
Make incremental improvements to change incentives now so that more significant reforms have a chance later. Fischer believes that the Affordable Care Act is hurting the country’s fiscal situation because it allows doctors to do “unnecessary back surgery and unnecessary stents on 30 million new people.” He doesn’t have much hope for accountable-care organizations (ACOs), either, because they’re part of the same flawed system. ACOs will need to hire doctors or buy out their practices, and those prices will be based on what they can make doing fee-for-service work. The Urban Institute’s Berenson made a similar point: reforms to the existing Medicare fee schedule will make specialists more likely to participate in a primary-care dominated system like ACOs “because they won’t have such a great stake in the status quo.”
Prioritize patients over doctors’ procedures. Medicare pays primarily by number of procedures and secondarily by number of patients. By shifting the payment priority toward patients, policymakers can reduce the incentive for unnecessary procedures. Fischer suggested another way to pay cardiologists as an example: $500 for caring for a patient in heart failure and $200 for doing a heart catheterization—putting the financial focus on the patient instead of the procedure.
Revise current prices, with the goal of shrinking the pay disparity between primary care and specialties. Set in motion changes that will make it easier for new doctors to choose primary care. Over time, specialists should see their incomes shrink relative to primary-care doctors. Specialists can spend twice as much time in residency, but that doesn’t mean that they trained for twice as long. Considering all levels of school, from kindergarten through college, medical school, and residency, another three years adds up to about 10 percent more training. Other factors beyond just training time should go into determining specialty-physician pay, but a two-to-one pay advantage over primary care is not sustainable. Some primary-care doctors are abandoning Medicare—if not medicine altogether—and setting up concierge practices.
The Trump administration and congressional Republicans had high hopes for a successor to the Affordable Care Act but have failed twice so far to reach consensus on a viable plan. Any hope of dramatic reform could depend on underlying incremental changes. Bundled payments, site-neutral reimbursements, or even dropping government price-fixing altogether will have limited impact if a new system perpetuates specialists’ financial advantages. American health care has come to rely on idealism to supply primary-care doctors. The costs of this historical accident are growing. At some point, it will simply cost less to pay more.
Photo by Joe Raedle/Getty Images