ARN: There’s a general feeling that the current Medicare system does not employ payment policies to try to improve the quality of care. While I don’t necessarily agree with the statement, the committee members said that fee-for-service is "toxic" to quality. I think that reflects a sense that Medicare payment policy has not done what it should do to encourage value in purchasing or improvement of quality. Congress directed the IOM to identify and prioritize options for aligning performance with payment in the Medicare program. The mandate identified three topics for the study to address: the performance measure set to be used and how that set should be updated; the payment policy that should be used to reward performance; and the key implementation issues involved, such as data and information technology requirements.

ARN: The report set forth three fundamental premises: the quality of medical care in the U.S. is not as high as it can and should be; it is part of the professional obligation of physicians to work to improve the quality of care they provide; and high quality care should be valued and rewarded more than poor quality care. That is, why should a primary care physician who pays no attention to preventive services be paid the same as the conscientious one who fulfills the primary care obligation? Payors are pushing pay-for-performance because they want to buy value, not just save money. There are private sector programs that are now covering millions of people, being met with variable responses from physicians and other health care providers. I believe the private sector experience can’t necessarily be transferred to Medicare, and we have to be cautious what we recommend to Congress for the Medicare program, which covers almost 40 million beneficiaries. What works in the private sector, where you have a controlled data system, a controlled patient population and to some degree controlled panels of physicians, may not work in the Medicare program – with its size, scope and beneficiary population, which is often more vulnerable and has more complex illnesses. There is an assumption that, since pay-for-performance is working to some degree in the private sector, that it’s a slam dunk for Medicare. I think that’s a hazardous leap.

ARN: First, the committee wanted to define what quality was, and what was to be measured. The IOM had six quality aims that it described in the Quality Chasm report half a dozen years ago: quality care is care that is safe, effective, efficient, timely, equitable and patient-centered. The IOM committee collapsed those to three – clinical quality, efficiency, and patient-centeredness – and recommended that all of those be measured, reported and eventually rewarded. The scope of measured activity is going to go well beyond just the process measures – like HbA1c for diabetics, or providing ACE inhibitors for patients with congestive heart failure – that form the basis of the private sector programs, and beyond current measures of patient-centeredness as patient satisfaction. In the future, patient-centeredness will be measured by surveying Medicare patients about the care they receive: whether they were treated with respect, whether they got an appointment when they wanted one, whether their health care provider answered their questions.

A good deal of thought has gone into the process by which clinical quality measures are developed and approved. An IOM report a year ago identified about 140 measures that were "ready for prime time," and additional measures are being developed. The AMA has convened, for well over a year, the Physician Consortium for Performance Improvement, which aims to have about 170 measures by the end of 2006. Those measures are available in the public domain. One of the recommendations in the earlier IOM report was to form a national quality coordinating board that would be quasi-independent, operating like the Federal Reserve to approve measures and contract to have other measures developed.

An important private sector component of that process has been the Hospital Quality Alliance and what was called the Ambulatory Quality Alliance – it’s now called the AQA because its mission has broadened to incorporate all areas of physician practice. The AQA consists of not only the AMA and major specialties, but also America’s Health Insurance Plans – the trade association for health insurance plans – and the Agency for Healthcare Research and Quality. They have taken measures that were developed by the Physician Consortium and reached consensus on those that are ready for implementation. A final vetting process involves the National Quality Forum. By the time a measure is put into a measure set that will be considered by Medicare, it has gone through a deliberative process to make sure that it is evidence-based, measurable, and does not pose an insurmountable administrative burden on collecting it. All of this has been put in place in anticipation of some action on pay-for-performance.

There are a whole host of secondary considerations addressed in the report that have to be kept in mind: What is the administrative burden? What are the unintended consequences? Where should the money come from? Will there be new money to provide the rewards? Should it be voluntary or mandatory? If mandatory, is it tied to conditions of participation – that is, if physicians don’t report quality data they won’t be eligible to receive any payments from Medicare. All of these are important questions involved in the implementation of the program.

ARN: The first recommendation was that the Secretary of Health and Human Services should implement pay-for-performance in Medicare using a phased approach as a stimulus to foster comprehensive and system-wide improvements in the quality of health care. The initial funding, over the next three to five years, should come largely from existing funds – but the IOM left open the possibility that additional resources will need to be put into it, particularly in areas like assisting physicians to implement electronic health records in their offices, and things of that sort. While there is a strong demand for budget neutrality in the Medicare program, there was a recognition that, in order to get this off the ground and running, some new money may be necessary. To the degree possible, pay-for-performance should be implemented with existing resources, which includes having some level of withhold – MedPAC said two percent – that would be redistributed in pay-for-performance rewards. The report questioned whether or not physician behavior could be changed with such a small withhold and said that work needs to be done to determine what the appropriate level of rewards should be. Conventional wisdom is that, for hospitals, a half-percent made a difference: voluntary reporting went above 90 percent with a half-percent reward. For physician services, because there’s so much less money on the table, I think most private sector programs are looking at something like 10 percent in order to make it worthwhile for physicians to set up the data system reporting capability. There is experience in the United Kingdom that we ought to be looking at. They’ve had a pay-for-performance program for over a year, and they’ve put up substantial new money to infuse it. The National Health Service has been startled with what happened because they’ve been forced to pay out far more money in rewards than they expected to, or budgeted for. The general practitioners over there really stepped up to the plate.

There was consideration of whether or not each of the provider groups should remain in "silos" for Medicare payment purposes. That is, should hospital care be paid out of Part A and physician care paid out of Part B – when you understand that physicians providing more effective medical management in the outpatient setting may actually save hospital costs, and should not physicians then be able to receive some benefit from that? The committee said that initially, it’s not practical to blow down all the silos, but that the ultimate goal should be to have the payment pool for Medicare converted into one large pool, so that the savings in one area can be utilized more effectively in another.

Any implementation of pay-for-performance for Medicare should be phased in, in a "learning environment," so that unintended consequences can be identified and corrected. Acknowledging the administrative burden is important, particularly in small practices, there’s a reality that has to be considered and confronted, if all of a sudden physicians have to dig out and report a lot of data from their paper records, when they’re already up to their neck in administrative burdens and just want to be left alone to take care of sick people. Because of the difficulty in redesigning clinical practice in small groups and solo practices, pay-for-performance in Medicare should be voluntary for the first three years, and then the Secretary should assess the progress that’s being made in quality improvement and make a decision on whether or not it should be mandatory. I think that’s a very important finding because, in my view, it would be disastrous if Medicare pushed into this in a way that would deprive Medicare patients access to care. Physicians have a tipping point in which their good natures can’t be relied on further, and if you combine a huge administrative burden in digging out and reporting quality measures with proposed Medicare fee cuts from the sustainable growth rate formula, I worry about serious access to care problems by Medicare patients who need to find new doctors. Even if we’re just talking about pay-for-reporting – the administrative burden raises the potential for adverse selection by doctors avoiding the person with multiple chronic illnesses, or by doctors not taking any more Medicare patients. The access issue also includes increasing racial and ethnic disparities. This committee gave a lot of thought to, and included a substantial amount of text in saying that this is a new field that Medicare is plowing. Let’s make sure that we do it in a way that we can recognize mistakes early on, and correct them.

ARN: The stickiest issue of all is measuring and rewarding efficiency, and yet the drive of efficiency is the thing that’s driving pay-for-performance in the private sector. Purchasers want to purchase quality – they’re willing to pay more to have quality, but they want to make sure that they have value. Efficiency, in the IOM’s earlier definition, is simply reducing or avoiding waste. There are more sophisticated definitions now, but they all boil down to that, and it comes from the fact that a patient can have a service provided in an academic medical center in one part of the country and it costs twice of what the same procedure costs in another part of the country, after adjusting for age, sex and other severity indicators, with no improvement in outcomes. It’s clear that there is some level of waste in the system that needs to be confronted. Some researchers have said that as high as a third of all Medicare dollars could not be spent, and still have the same outcomes.

Efficiency is one of the most difficult things to implement in a pay-for-performance system because the average Medicare patient is receiving care from five or more physicians at the same time, on average. A recommendation for a patient to get an MRI, for example, may not only be shared among multiple physicians, but is impossible to accurately assign to any one. And yet, if we are rewarding the more cost-conscious physician who achieves the same outcome with the use of substantially less resources, then it’s necessary to find some way to assign or attribute resource use to the individual physician. Private sector programs have used software that is able to group services into a certain episode of care around what is presumed to be the major diagnosis. Other software seeks to assign to a particular physician the resources that are used. This is working with greater or lesser effectiveness, depending on who you talk to in the private sector, but it isn’t foolproof. If a physician is blamed for running up the bill, when they weren’t the one who was really responsible for it – and if it is laid out through public reporting that they are "inefficient," and also results in lower payments under pay-for-performance – that is a certain recipe for outrage. The IOM report said that rewarding efficiency has to be a goal, that information technology needs to be perfected to assist in that, that we can’t let perfect be the enemy of the good, and that sooner or later we must achieve greater efficiencies and wring out some of that waste.

ARN: Since prices are essentially controlled, the differences in "efficiency" are a direct function of volume. Efforts in pay-for-performance to control volume, for example, would be in publishing the names of the high-cost or low-cost physicians. The big-growth categories are in imaging, minor procedures and office visits. You have to make a distinction between good volume and bad volume. Some of the volume increases that we are seeing are a direct result of encouraging cholesterol screening, diabetes management, high blood pressure care – the kind of volume we should be encouraging. The volume that is worrisome is that which doesn’t contribute to a better outcome, and pay-for-performance will struggle to try and get a handle on that, but it is not easy. Systems of care differ in their efficiency. What the IOM visualized is payment policy being one part of a necessary restructuring, so that efficiency is improved.

ARN: I think that the direction that the country is going, in terms of the need for information technology, is going to be a powerful disincentive to solo and small group practices. It isn’t anything that I particularly favor, but I have to be a realist and assume that the restructuring that is taking place in the health care system because of information technology – and because practically nobody is paying for care out of their pocket anymore – is resulting in a conversion of medicine from a cottage industry to an industrial model. Still, half of practitioners in America today are in groups of five or less, but I think that payment policies, the demands of large purchasers of care, and the direction that government is moving are all going to make it harder and harder to be an island.