In the battle over bending the cost curve in Medicare, a recent article in Health Affairs should set off alarms. In it, Francis Lukas and colleagues describe the proliferation of new cardiac surgery programs—300 in 10 years–at exactly the same time that the number of cardiac bypass grafts fell. Moreover, the new programs generally did not appear in rural areas, where they might have increased access for underserved populations: instead, they appear in markets that are already well resourced.
It would be nice if the new programs generated competition, particularly, given our urgent cost problems, price competition. However the researchers found no evidence for this, and there is no reason to expect it: price competition is problematic in medicine, where because of varying degrees of market power, everyone outside Medicare pays a different price, and the uninsured pay the most. More likely there is competition on perceived quality, where each program touts its new, expensive equipment and pricey, up-to-date facilities: the so-called arms race.
Based on the findings of the Dartmouth Atlas, that more care is not necessarily better care and that excessive care is wasteful, current health policy is to pay for “value instead of volume”. The appearance of these new specialty surgical programs—and there are surely new programs in other profitable medical specialties, such as interventional cardiology and orthopedics—will seriously complicate this approach.
In particular, as the authors of the Health Affairs study recognize, Dr. John Wennberg demonstrated early on that in medicine, supply can create demand. More physicians to do procedures mean more procedures are likely to be done. To a man with a hammer, every problem looks like a nail.
Moreover, once significant investments have been made (in human as well as capital resources), there is a financial imperative to fully employ them, and significant investment loss if the new programs are forced to downsize, or relocate. As capacity increases beyond what is necessary to perform all clearly indicated procedures (say those that carry IA and 2A recommendations), practitioners will move on to cases where the indication is weaker. Academic physicians, supported by equipment and pharmaceutical manufacturers, will develop evidence for new indications.
Once new treatments are available, if CMS tries to limit them to patients for whom they are cost-effective — CMS would need an Act of Congress to do this — there will be howls of “rationing,” particularly from Republicans, but also doctors and patients. There are very few procedures that don’t benefit anyone. And if private insurers fail to cover the “inefficient” procedures, or restrict patients to hospitals that are cheap instead of large, nationally known teaching institutions (think the Mass General, Brigham and Women’s), patients will complain bitterly however many quality measures are satisfied. This “market” solution will please economists and Republicans: it will not mollify fearful, angry patients. Just such tight restrictions on procedures, drugs and providers were what did in managed care in the 1990’s.
Reducing costs at the same time that ever more expensive medical services proliferate has tied this nation in a Gordian knot. Republicans have threatened to bring down the American financial system to introduce “market discipline” to Medicare. Democrats modestly suggest reorganizing the entire U.S. health delivery system according to a model that has never been demonstrated at scale, and requires significant upfront costs of its own.
Medicine: A Dysfunctional Market
The data that Lukas and colleagues have pulled together show us that, in many places, medicine is a highly dysfunctional market. There are many and sufficient reasons for this: getting your heart fixed is not the same as getting your car fixed. You can always take your car to a different mechanic, and, at worst, you can trade your failing car for a new one. A patient with 3-vessel coronary disease has fewer options.
The study also suggests why this particular market works so poorly: “cardiac services contribute 25 percent to 40 percent of a hospital’s net revenues.” Even dysfunctional markets provide useful information. We pay too much for cardiac services, so we get too many of them. (And indeed, we pay nothing for coordination of services — that is supposed to be covered by $75 for a fifteen minute office visit — so we get too little of that.)
In a perfect market — where no supplier or customer has market power and everyone has perfect information — excess profits bring in competitors, who drive down prices until they equal marginal costs. But that doesn’t happen often in medicine. Providers prefer to compete on perceived quality, rather than on cost.
One solution the authors suggest is resurrecting Certificate of Need (CON) requirements, a clumsy tool but better than nothing.
A Better Way Forward: Improve Price Regulation
I would suggest a different solution for Medicare, a solution which has been demonstrated at scale, and does not require an Act of Congress: improve price regulation.
Price control, particularly control of relative prices, is one reason the Europeans have succeeded in covering all their patients, and getting better outcomes than we do, for a fraction of what we pay. Even Dr. Robert Berenson, an early and forceful advocate of managed care, recently suggested that $650,000/year surgeons could afford to take a small hit.
CMS already regulates physician fees under a statute passed in the 1980’s. The system the agency currently uses, the Resource-based Relative Value System (RBRVS) was intended to improve the compensation of evaluation and management services relative to procedures. However, partly because of the disproportionate influence exerted on the process by an AMA committee, the Relative Value Update Committee (RUC), the difference in reimbursement has instead grown. CMS can develop and deploy a modified system, using administrative rulemaking, without going back to Congress.
The problem with procedures is not what we pay the doctors: their fees are the least of it. The problem is everything else. If a cardiologist places an intracardiac defibrillator, he is paid, say, $700. But the system also has to pay for the defibrillator (up to $20,000), the catheter used for its insertion, the cath lab with its top-of-the-line imaging equipment, and the technicians who staff it. There is little price competition for any of the equipment—little incentive to insert a less expensive defibrillator, or to buy a used camera after a new one has come out. The whole operation costs the system, conservatively, north of $30,000, many times what the doctor is paid.
In a recent article in JAMA, Dr. Sana al-Katib and colleagues found that more than 20 percent of the debrillators inserted last year in the U.S. may not have been medically indicated—exactly the result one would predict from the Health Affairs study.
The Key: Changing Incentives For Medical Students Choosing Specialties
The reason to change relative reimbursement for physicians is not to save money today, although that is useful. The key to change incentives tomorrow, when a medical student selects a specialty. Now, three or four more years of training will double, sometimes triple, his or her future annual income. If the increment were only, say, 20-30 percent as it is overseas, maybe more would choose cognitive specialties, where demand is high and arguably the value to patients and society is greater. Innovators would focus on techniques for improving care rather than ever more expensive gadgets for proceduralists. And in the long run, fewer procedural specialists would likely mean fewer procedures, especially where the indication is uncertain.
At the same time, CMS needs to pay for critical services physicians currently are expected to provide for free, such as coordinating care, especially during transitions across care settings, and counseling families. For example, it should pay for conversations between the primary care physician and the hospitalist, the Visiting Nurse, the family. Pay twice the going rate for the first post-discharge visit.
Today, everything must be documented, so there is no special opportunity for fraud, as perhaps there was in the 1950’s. There is no need for everyone to work for an ACO, or to abandon fee-for-service. We just need the right fees for the right services, to get a better balanced workforce.
Obviously it would be unfair to specialists in the prime of their careers to change the rules for them today. But CMS can change the rules prospectively, just as industry and local governments have, so that new entrants are on notice that their compensation will be lower. In this way, CMS can bend the cost curve without cutting services to beneficiaries, and without even the appearance of rationing: new expensive interventions of marginal value will simply not be developed.
Although no one has said so, all the health reforms of the last twenty-five years have been designed to reduce costs to payers while maintaining the profit margins of providers. Patients simply get less care, whether by “choice” (constrained by financial pressure) or “rationing”. Yes, paying for “value instead of volume” makes a good sound bite, but even enthusiasts are finding that identifying value is hard. Every procedure today is of value to someone.
Compared to that, reducing price gouging is easy. Economists now understand that prices are not just market-clearing devices, they are signals. To bend the cost curve, we need to send better signals to providers as well as patients.
Winston Churchill is said to have said: “Americans will do the right thing—after they have tried everything else.” For Medicare, the time is now.