Blog Home


MEDICARE: Can We Fix the SGR to Control Spending?

March 9th, 2007

This post is based on testimony at a hearing of the House Ways & Means Subcommittee on Health, March 6, 2007.

The primary problem with the SGR [sustainable growth rate] is that while it controls total spending by physicians, it does not affect the volume and intensity of spending by individual physicians. In fact, there is some concern that it may actually exacerbate the incentives for individual physicians to increase the volume and intensity of services they provide. The reason is that nothing that they do as individuals is likely to affect the overall spending level for physician services.

This has led to serious questions about what the current Medicare fee schedule is and is not rewarding. While some trade-offs are inevitable because multiple goals may not be entirely compatible, the incentives associated with an SGR, particularly when applied only to one segment of Medicare spending (that is physician spending), have been viewed by many as being perverse.

MedPAC [the Medicare Payment Advisory Commission] has identified two major choices for the Congress. The first is to repeal the SGR and focus on the development and adoption of payment reforms that would improve incentives for physicians to provide high-quality services at lower costs. The second is not to repeal the SGR but to extend spending limits across all of Medicare, perhaps with targets differing across regions to reflect the well-known variations in spending by region.

While I don’t disagree with the basic dichotomy that has been laid out — repeal the SGR or extend the concept to all of Medicare — I would phrase the choices slightly differently. As MedPAC has also noted, Medicare needs to institute policies that improve the value of the program and reward providers for the efficient use of services — as well as creating incentives to improve quality and care coordination — no matter what it does about expenditure targets.

Moving in this direction for physicians will require, among other things, the use of a less disaggregated fee schedule; and the use of a less disaggregated fee schedule should make it possible to do away with the SGR for at least a period of time. Care coordination and the focus on better treatment for chronic conditions, which is where so much of Medicare spending occurs anyway, is unlikely to work well when physicians are being reimbursed on a micro unit basis. Figuring out how to create the right bundles and recognizing the significant power shifts that could result will be difficult and time-consuming. But there are interesting demonstrations that are already started or at least being developed that should provide some assistance.

The coronary artery bypass graft (CABG) demonstration that was started when I was at HCFA [the Health Care Financing Administration, now the Centers for Medicare and Medicaid Services] bundled all Part A and Part B expenditures into a single payment and although not conclusive appeared to result in lower costs and as good or higher quality for the participating groups. A gain-sharing demonstration is starting that would allow physicians and hospitals that are not financially at risk to work together and share savings that result from better care of complex cases and chronic care patients. Other demonstrations are attempting to show the effects of disease management or better care coordination in a fee-for-service system.

MedPAC has recommended elsewhere — and as co-chair of the IOM subcommittee that was responsible for the recent release of the IOM report on pay-for-performance, I concur — the idea of adopting payment strategies that reward institutions and, when we can put in the proper measurement systems, clinicians who provide high-quality, low-cost care and do so in at least a budget-neutral manner. The early results from the Premier demonstration are consistent with conventional wisdom that increasing quality can be associated with lower costs. But this is just one small example. As the IOM report makes clear, however, moving to a system that realigns incentives will require a lot of changes and a lot of difficult decisions, not the least of which is a uniform, national set of performance measures.

I regard pay-for-performance, or results-based payment, as part of the process to begin realigning incentives in Medicare so that the payments that are made are more in line with the objectives of the Medicare program. Many changes will need to be made to restructure the payment system so that it encourages more of what we want produced (i.e., high-quality, efficiently produced, appropriate care) and that recognizes that much of the care needed by an aging population will have to focus on the needs of individuals with multiple chronic conditions.

As important as it is to realign incentives, it is also important to provide both payers and providers with better information on the relative clinical effectiveness of alternative medical procedures and technologies. A number of other countries have been involved with the concept of comparative clinical effectiveness but generally only for new pharmaceuticals and medical devices. Similar information needs to be available for medical procedures as well, since that’s where most of the money is spent.

Even if incentives are appropriately aligned, we can hardly expect to “spend smarter” if clinicians and payers (and patients) don’t know “what works when, for whom, under what circumstances.” Getting such information will require a significant investment and take several years to develop. But in a sector that is now spending $2 trillion, it is hard to explain why that type of investment would not be appropriate.

The bottom line: “We need to know more and pay for it better.”
If that doesn’t work to slow down Medicare expenditures, we had better be prepared to introduce expenditure targets across the board in Medicare, but recognize it will be hard not to exacerbate problems with medical silos in a world that really needs better coordination across medical boundaries.

Email This Post Email This Post Print This Post Print This Post

 to the #1 source of health policy research.

3 Trackbacks for “MEDICARE: Can We Fix the SGR to Control Spending?”

  1. Health Care BS » Blog Archive » Medicare: Can We Please Try the Market? Please?
    April 18th, 2007 at 9:15 pm
  2. Health Care BS » Blog Archive » SGR: An Exercise in Futility
    March 13th, 2007 at 9:07 pm
  3. HealthBlawg
    March 12th, 2007 at 9:19 am

5 Responses to “MEDICARE: Can We Fix the SGR to Control Spending?”

  1. acavale Says:

    I am glad that Ms. Willensky finally replied to this blog. Unfortunately as a physician, I am totally unsurprised by her response, which is typical coming from someone who obviously has no cluse as to what actually occurs at the time and place of delivery of health care. Her argument is primarily based on artifactual values and takes no account of how health care is actually provided to Medicare beneficiaries. I am very surprised that she is unaware of the ACP’s PCH paper.

    One would like to know what she means by “Fees have been going up very modestly while spending in part B has been increasing rapidly — 12 to 15% per year”. As far as I know, physician reimbursements from Medicare have remained at 2005 levels. So one would like to know where this 12-15% per year spending increase is going – certainly not to reimburse physicians for medical care.

    I would also like a simpler explanation for this statement – “Getting to the right type of reimbursement will be difficult as long as physicians practice in small, single specialty groups, since any payment strategy that focuses on paying for narrowly defined, specific interventions is inconsistent with paying for efficiently-provided, high quality care”. So if I get it right, Ms. Willensky will even try to tell physicians how they should set up their practices. We could practice in Russia or Cuba in the same manner, not in the U.S., the bastion of free enterprise capitalism; or is this thinly veiled Socialism?

    Finally, why is the concept of fostering a trusting relationship between providers and consumers of healthcare so alien to Ms. Willensky? We don’t trust the same people who take of our health to make decisions on health care? Or do we trust the bean counters more? I am appalled by this statement by Ms. Willensky. Maybe looking at healthcare from a political background is not the best way to approach it.

    Finally, how can the physician community speak with one voice, when Medicare and other carriers have so successfully implemented the “divide and rule” policy over the past 15 years or more? I as an individual physician have come up with specific ideas, which Ms Willinsky conveniently avoids addressing or pooh-poohs, rather ineffectively.

  2. Gail R. Wilensky Says:

    I’m not familiar with the American College of Physicians plan described by Bob Doherty, However, I agree with him on the need for a Medicare physician payment system that emphasizes coordination of care. Such a system would indeed offer physicians prospective, bundled, and risk-adjusted payments for treating patients with chronic diseases. It would also allow physicians to share in the savings to Medicare Part A from avoiding expensive procedures such as CABG surgeries, and – just as importantly – it could penalize physicians whose patients end up inappropriately needing many such expensive interventions.

    Doherty’s right that simply grafting P4P onto the existing system itself isn’t like to affect change—certainly not at the levels of performance-based payments currently under consideration. I would also emphasize that, like P4P, payments for coordination of care, whether under Doherty’s proposal or any other, can’t simply be grafted onto the current system. We need to transform the payment system, not just add new payments to the old ones.

    I’m troubled by aspects of the other two comments that don’t seem to see the need for this kind of transformation. David Catron seems to be saying that everything would be fine if we would just let the market work. As a market-oriented economist, I have a lot of sympathy with trying to rely on markets, but that’s hard to do in a Medicare system that in some areas dominates the market. It’s also hard to do in health care generally, where 1) people are using someone else’s money because third party payment as a strategy makes sense at least for low probability-high cost events; 2) spending in very concentrated; and 3) many seniors don’t have sufficient income to pay for even the significant amounts of routine care they use. Nonetheless, some strategies are better at trying to mimic market incentives than others. The Medicare fee schedule coupled with a spending limit is particularly perverse in the incentives it provides, but an unconstrained fee schedule would be even more perverse. If Catron has a whole new world he is proposing in place of Medicare, that’s different, but he needs to tell us how that world would work.

    Arvind Cavale sets up a straw man when he says Medicare spending is bound to keep rising because of the aging population. No one is suggesting controlling spending in the sense of preventing it from increasing at all — rather the attention is on moderating spending increases, which Medicare has not been doing very well — especially in Part B. Fees have been going up very modestly while spending in part B has been increasing rapidly — 12 to 15% per year. That kind of spending growth, particularly given what we know about variations in mix and severity of use coupled with mediocre outcomes, is a real problem and will only get worse if not confronted directly.

    In a recent IOM report for which I was a member of the panel, we emphasized the need to build performance measures that cross all of the institutional and individual clinicians who treat a patient with chronic disease for a particular episode of illness or a particular period of time. P4P based on those sorts of measure is not a panacea for all of our spending problems, but it is one important step among a variety that are needed. Getting to the right type of reimbursement will be difficult as long as physicians practice in small, single specialty groups, since any payment strategy that focuses on paying for narrowly defined, specific interventions is inconsistent with paying for efficiently-provided, high quality care.

    Cavale pleads for trust in the medical community, but in a world where patients only receive, on average, 55% of the care appropriate, and where spending goes up at a rate of 12 to 15% while fees go up by 1.5% or not at all, it’s hard to imagine just turning over the decision-making to the medical community and “hoping for the best” — even if patients would find this attractive, which if they knew more, they probably would not. Maybe Cavale means Medicare needs to take steps to engender trust within the medical community as a partner. If that’s what he means, a first step might be for the medical community and its leadership to come forward and suggest credible ways to reconfigure payment that reward the provision of high quality/efficient care.

  3. David Catron Says:

    This piece is informed and thorough, but Wilensky’s approach to the subject is symptomatic of an ailment that afflicts many designers of health care policy: a tendency to disregard the market as a viable solution for any problem. Here’s a revealing selection:

    “Medicare needs to institute policies that … reward providers for the efficient use of services … Figuring out how to create the right bundles and recognizing the significant power shifts that could result will be difficult and time-consuming.”

    The market will reward providers for “efficient use of services” if the policy makers will allow it to do so. The current central-planning methodology is “difficult and time-consuming” because no group of experts, no matter how sophisticated, can possess all of the information that the market instills in a price.

    Indeed, any government-imposed “sustainable growth rate,” whether it is based on the CPI, per capita GDP, or the effusions of the Delphic Oracle, is doomed to failure. Such efforts succeed only in creating perverse incentives that CMS must then attempt to fix by even more regulatory interference in the market.

    I wish Milton Friedman were still alive and in charge of CMS.

  4. Arvind Cavale Says:

    While agreeing with Bob Doherty in his comments, I would like to point out the fallacy in the whole idea behind the notion of “controlling spending”. If the approach is directed towards controlling spending, it is bound to fail or succeed in such a way whereby a large percentage of Medicare recipients will lose access to quality care altogether. It is unreal to expect anything but increases in Medicare spending in the coming decades just based on the demographic nature of the population. SGR fails as a policy on all counts.

    The smart policymaker will look into how these Medicare dollars are best utilised. This is where the ACP’s concept of Patient-Centered Medical Home will fit well. All these years Medicare dollars have been spent largely in covering catastrophic events and procedures. For example, a physician will get paid better to treat a patient who has a heart attack than if he spent most of his/her energy in educating and managing chronic diseases and prevent a heart attack.

    Unfortunately P4P as currently envisaged is a very poor method to asses efficient and effective care. The ideal method to asses overall outcomes of medical interventions involves the patient, patient’s caregivers and physicians. How does one prove objectively that an intervention actually saved an Emergency Room visit or hospitalisation? There is no good method to do this. The best alternative is to change the reimbursement concept for chronic disease management from an “episodic” to a “continuous” care model. Here one can clearly document (with an EMR for example) how ongoing interventions (such as regularly reviewing blood glucose logs for diabetics) potentially prevented the patient from seeking Emergency Care in a particular time period. In fact such interventions can clearly be shown to reduce even repeated office visits. However, without appropriate reimbursement for such active interventions, it is unlikely that physicians will adopt such methods on a large scale. What this also entails is adoption of appropriate technology so that there is rapid and effective transmission of critical information from the patient to the physician’ s office, thereby interventions can be made in a timely manner, so that unnecessary visits to the E.D. can be minimised. Such technologies are currently available, but their use is not encouraged by the payers, so severely under-utilised. We have a beautifully documented records of such activities (wholly unreimbursed but achieving great results).

    Eventually, it comes down to how we as a society assign value to medical service and whether Medicare is willing to encourage an enviroment of trust in the medical community. If the average Medicare consumers had a say in this, I bet they would chose this model of care. We can only hope that the policymakers realise this soon.

  5. Bob Doherty Says:

    I agree with Dr. Wilensky’s observation that the SGR does not effectively influence the volume and intensity of services ordered by individual decisions. MedPAC’s option of regional targets would suffer from the same problem as a national SGR: it increase or reduce payments to efficient and high quality physicians by the same amount as the least efficient and lowest quality physician, a problem that MedPAC acknowledges in its report. Pay for performance is also not likely to substantially improve quality or achieve savings if it is grafted onto a system that pays doctors soley based on the volume and intensity of services procedures. (The new Medicare Physicians Quality Reporting Initiative would allow physicians to achieve a 1.5% bonus for reporting on as few as 3 clinical measures, hardly a compelling incentive to consider efficiency whan the overall payment system rewards volume. Plus, physicians get the same “bonus” regardless of the impact of the measure being reported).

    I also agree with Dr. Wilensky that Medicare should begin moving in the direction of paying physicians for care coordination and acquiring the systems needed to coordinate care, especially for patients with multtiple chronic diseases. I co-authored a paper for the American College of Physicians, my employer, which proposes that Medicare pay physicians who demonstrate the capability to coordinate care through a patient centered medical home under a hybrid system that would: combine bundled, prospective and risk adjusted payments for care coordination (for the physician work outside of a visit and the practice level systems needed, such as patient registries and evidence-based clinical decision support), a fee for service componet for face to face encounters, and a performance based component. More information can be found at: This model would require, though, a willingess by Medicare and payers to recognize the value of care coordination and to pay primary care and principal care physicians at a high enough level that recognizes the value, work and expense involved. It also would require a willingness to break down Medicare’s payment silos, so that physicians who are able to reduce avoidable hospitalizations for patients with chronic diseases (such as by helping a diabetic patient avoid complications leading to amputation) can share in part of the savings in Medicare Part A. The Medicare Medical Home demo, enacted by the 109th Congress, is a start, since it would pay physicians who demonstrate the capabilty to coordinate care for patients with chronic illness under a different payment structure:, consisting of a care coordination fee and shared saviings from reducing Part A exxpenses.

    One thing should be evident: none of these or other other promising appproaches to redesigning Medicare payment policies will succeed if physicians are facing an SGR cut of almost 10% next year and cumulative cuts over more than 40% over the next several years. The SGR simply has to go if we are to be able to move forward on re-aligning Medicare incentives to reward physicians for managing and coordinating care, for following evidence based guidelines at the point of care, and for acquiring the HIT and systems needed to do this effectively. Congress shouldn’t wait until all the demos are completed to recognize this reality and act accordingly.

Leave a Reply

Comment moderation is in use. Please do not submit your comment twice -- it will appear shortly.

Authors: Click here to submit a post.