TESTIMONY OF THOMAS A. SCULLY ADMINISTRATOR CENTERS FOR MEDICARE & MEDICAID SERVICES ON MEDICARE PAYMENT FOR PHYSICIANS' SERVICES AND PROFESSIONAL LIABILITY INSURANCE
BEFORE THE SENATE APPROPRIATIONS SUBCOMMITTEE ON LABOR, HEALTH AND HUMAN SERVICES, AND EDUCATION

Chairman Specter, Senator Harkin, distinguished Subcommittee members, thank you for inviting me to discuss how Medicare pays for physicians' services. I have worked on Medicare physician payment issues since 1989 when I was one of the primary people in the previous Bush Administration negotiating the creation of the resource based relative value physician payment system, sometimes referred to as RBRVS. I personally think that, over the years, this has been the most stable payment system in Medicare, and historically there has been far less controversy in physician payments than we have witnessed with other health care providers. In fact, the resource based relative value system has worked reasonably well and often is used by private payers. A number of factors have combined to cause the formula, as set in law, to produce negative updates for 2002 and 2003 as well as projected negative updates for future years.

Over the past 15 months, I personally have devoted about a quarter of my time, working with every group imaginable – lawyers, physicians, and the highest levels of the Administration – to fix the fee schedule administratively. The Agency explored every legal option possible to try to fix the negative update, but unfortunately, the statute is too prescriptive to allow an administrative fix. We, however, were able to modify the methodology used to calculate the Medicare Economic Index (MEI), which measures inflation for the cost of providing physicians’ services used in the fee schedule formula. The change we made resulted in a 0.7 percentage point increase in the 2003 MEI update, from 2.3 percent to 3.0 percent, and is expected to increase the MEI in all future years by roughly 0.5 percentage points per year. This change will increase Medicare spending for physicians by about $15 billion over the next 10 years. While the 2003 update is a negative 4.4 percent, without this technical modification, the update for 2003 would have been negative 5.1 percent. This does not resolve the fee schedule update problems, but it is a step in the right direction. We point out that Medicare expenditures for physicians’ services increased by 5 percent in 2002, and are expected to increase by 2 percent in 2003, despite the negative update.

We believe that the physician update formula is broken and can only be corrected by legislation. The physician update has both short-term and long-term problems. The short-term problem is that the negative 4.4 percent update will go into effect on March 1 without a change in law. The long-term problem is that there will be negative updates for future years under current law. No one ever intended that the physician formula have this result. Working with Congress to find a legislative fix is a top legislative priority of the Administration. We believe that fixing the physician update would be a technical correction to a broken system, not a "giveback" that other providers have been seeking. We believe it is important that Congress takes immediate action to prevent the 4.4 percent reduction scheduled to go into effect on March 1. Such an action is the only equitable solution that will stop physicians from experiencing payment reductions that result from this flawed formula.

HISTORY

Let me explain the origins and logistics of the physician fee schedule so that we can all understand how the system works. In 2002, Medicare paid about $44 billion for physician fee schedule services. Between 1997 and 2002, Medicare physician spending increased from 17.6 percent to 19.8 percent of total Medicare fee-for-service spending. Each year, Medicare processes about 600,000,000 physician claims. The fee schedule reflects the relative value of the resources involved in furnishing each of 7,000 different physicians' services. By law, we actually establish three components of relative values -- physician work, practice expenses, and professional liability insurance -- for each of these 7,000 services. The actual fee for a particular service is determined by multiplying the relative values by a dollar-based conversion factor. And the payment for each of the services is adjusted further for geographic cost differences among 89 different payment areas across the nation.

Payment rates for physicians' services are updated annually by a formula specified in law. The annual update is calculated based on inflation in physicians’ costs to provide care (the MEI), then adjusted up or down by how "actual" national Medicare spending totals for physicians’ services compare to a "target " rate of growth. If spending is less than the target, the physician payment update is increased, and if spending exceeds the target, the update is reduced. The system was designed to constrain the rate of growth in Medicare physician spending and link it to growth in the overall economy, as well as to take into account the physician role in the volume and intensity of services. Until 2000, in large part, the formula has been working as designed.

The law that sets this formula is extremely prescriptive. It does not give the Centers for Medicare & Medicaid Services (CMS) the administrative flexibility to adjust physicians' payments when the formula produces unexpected payment updates, as we witnessed in 2002 and 2003. The size of the negative update for 2002 was a surprise when it first became apparent in September 2001. As we looked at the actual numbers going into the formula, we explored every issue and every alternative that could have produced a different update, but we concluded that we did not have any flexibility. We made sure that every part of the update was accurate and fully in accord with the law.

Several factors lead to the negative updates. First, there were differences between estimates and after-the-fact actual data for components of the FY 1998 and FY 1999 Sustainable Growth Rates (SGRs). As I explain below, this means that the cumulative target was lower than it should have been. Second, there has been a downturn in the economy in recent years, which affected the SGR because it is tied to the growth in the country's Gross Domestic Product. Third, spending for physicians’ services has grown very rapidly in recent years. The combination of lower targets and higher expenditu res accounts for negative updates. In addition, in 2002, our measure of actual expenditures had to be adjusted to capture spending information on services that were not previously captured in the measurement of actual expenditures. Counting these previously uncounted actual expenses, as required by law, also increased cumulative actual expenditures – driving down the update.

While there are significant differences between the Medicare Volume Performance Standard (or MVPS, the predecessor to the SGR) and SGR, both use the same general concept that expenditures for physicians’ services should grow by a limited percentage amount of allowed expenditures each year. One important feature of both the MVPS and the SGRs for fiscal years (FYs) 1998 and 1999 was that the percentage increase was based on estimates of the four component factors specified in the law, made before the beginning of the year. Under the MVPS and the SGRs for FYs 1998 and 1999, the statute did not permit us to revise the estimates used to set the annual percentage increase. Beginning with the FY 2000 SGR, the statute specifically requires us to use actual, after-the-fact data to revise the estimates used to set the SGR. For some of the component factors of both the MVPS and the SGR, there have been differences between the estimates used to set the annual MVPS and SGR and the actual increase based on actual, after-the-fact data. For instance, under both the MVPS and SGR, we are required to account for increases in Medicare beneficiary fee-for-service enrollment. Because it is difficult before the beginning of the year to predict beneficiary enrollment in Medicare+Choice (or Medicare managed care, as they were known under the MVPS) plans, there have been differences between our estimates of the increase in fee-for-service enrollment and the actual increase. Under the MVPS, we generally estimated higher growth in beneficiary fee-for-service enrollment than actually occurred, although the statute has required us to revise our estimates for subsequent years.

Because the physician fee schedule conversion factor has been affected by a comparison of the actual increase in expenditures to the level of allowed expenditures calculated using the MVPS and the SGRs for FYs 1998-1999, revision of our estimates would have resulted in different conversion factors than those we actually determined. The 4.4 percent reduction to the physician fee schedule conversion factor is occurring, in part, because of a flawed statute that does not allow us to revise estimates for previous years. Physicians argue that these negative payment updates will hinder their ability to care for beneficiaries, and may result in some physicians not accepting new Medicare patients. We take these statements seriously, and are taking steps to monitor beneficiary access to care to ensure that our nation's most vulnerable citizens continue to receive the care they need. As we consider how to improve the Medicare physician payment formula, I think it's important to understand, from a historical perspective, how and why the formula operates the way it does today. It is, in fact, operating precisely as it was designed in 1997 – but we recognize that this has produced some large short-term adjustments.

PHYSICIANS' PAYMENT BEFORE 1997

As the Medicare program has grown and the practice of medicine has changed, Congress and the Administration have worked together in an effort to ensure that Medicare's payments for physicians’ services reflect these changes. As a result, the physician payment system has changed significantly in the past two decades. For many years, Medicare paid for physicians' services according to each doctor's actual or customary charge for a service, or the prevailing charge in the physician's area, whichever was less. From 1970 through the 1980's, spending for physicians' services grew at an unaffordable and unsustainable average annual rate of more than 14 percent. And, because the system was based on historical charges, it produced wide discrepancies in payments among different localities, medical specialties, and services. These payment differences did not necessarily reflect actual differences in the cost of furnishing services. As a result, the system was roundly criticized in the 1980's as overvaluing specialty services and undervaluing primary care services.

To address these criticisms, Congress directed the Physician Payment Review Commission, an advisory body established by Congress and one of the predecessor organizations of the Medicare Payment Advisory Commission (MedPAC), to examine different ways of paying physicians while protecting beneficiary access to care, as well as slowing the rate of growth in Medicare physician spending. On a bipartisan basis, and with the support of the first Bush administration, Congress accepted these recommendations and passed these and other reforms in the Omnibus Budget Reconciliation Act (OBRA) of 1989, and the new fee schedule was implemented beginning January 1, 1992. The resource-based work component of the fee schedule was phased in between 1992 and 1996.

Specifically, in its 1989 Annual Report, the Commission recommended a number of ways to change how Medicare pays physicians. The Commission first recommended instituting a fee schedule for physicians' payments based on the resources involved with furnishing each physician’s service, rather than on historical charges. The Commission also recommended that the relative value of three separate components of each service – physician work, practice expense and professional liability insurance – be calculated, as discussed above.

Under the Commission's recommendations, once the relative values were established, they were adjusted for cost differences, such as in staff wages and supply costs, based on the area of the country where the service was performed. Then the actual fee for a particular service for a year was determined by multiplying the relative value units by a dollar-based conversion factor. The Relative-Value Update Committee (RUC), a multi-specialty panel of physicians that is supported by the American Medical Association (AMA), plays an important role in making recommendations so that the relative values we assign reflect the resources involved with both new and existing services. We generally accept more than 90 percent of the RUC’s recommendations, and our relationship is cooperative and extremely productive. The Physician Payment Review Commission also recommended that HHS provide financial protection to beneficiaries by limiting the amount that a physician could charge beneficiaries for each service.

The Commission’s third major recommendation was to establish a target rate of growth for Medicare physician expenditures, called the Medicare Volume Performance Standard (MVPS). The MVPS target growth rate was based on physicians' fees, beneficiary enrollment in Medicare, legal and regulatory changes, and historical measures of the volume and intensity of the services that physicians performed. The MVPS was set by combining these factors and reducing that figure by 2 percentage points, in order to control to growth rate for physicians’ services. OBRA '93 later changed this to minus 4 percentage points. Actual Medicare spending was compared to

the MVPS target, which led to an adjustment, up or down, to the MEI to determine the update for a future year. The law provided for a maximum reduction of 3 percentage points, which OBRA '93 lowered to 5 percentage points.

PHYSICIANS' PAYMENT SINCE 1997

The Balanced Budget Act of 1997 (BBA) further changed the physician payment system based on subsequent Commission recommendations. In BBA, the SGR replaced the MVPS. Like the MVPS, the SGR is calculated based on factors including changes in physicians' fees, beneficiary enrollment, and legal and regulatory changes. However, the BBA did away with using the historical volume and intensity of physicians’ services as the factor in the target for growth in the volume and intensity of services. Instead, the real per capita Gross Domestic Product, which measures real economic growth in the overall economy per capita, was instituted as a replacement.

One other important difference between the old and the new growth targets is that the old method compared target and actual expenditures in a single year while the SGR added a cumulative comparison. Under the MVPS, if expenditures exceeded the target in the previous year, the update was adjusted for the amount of the excess in the current year, but there was no recoupment of excess expenditures from the previous year. Under the new SGR, the base period for the growth target was locked in at the 12 months ending March 31, 1997. This is the base period and remains set for all future years. Annual target expenditures for each following year equal the base period expenditures increased by a percentage amount that reflects a formula specified in the law, and they are added to base period expenditures to determine the cumulative target. This process continues year after year, adding a new year of expenditures to the cumulative target. A comparison of actual cumulative expenditures is made to cumulative allowed or target expenditures. Under a cumulative system, if expenditures in a prior year exceed the target, the current year update is adjusted to make annual and target expenditures equal in the current year and to recoup excess expenditures from a prior year. While the BBRA made some further technical changes to allow these adjustments to occur over multiple years, that is the general way the formula was established in law. The SGR is working the way it was designed.

BBA also increased the amount that the update could be reduced in any year if expenditures exceeded the target. The maximum reduction was increased by 2 percentage points, to 7 percentage points. Thus, for example, inflation updates in the range of 2 percent, reduced by the 7 percentage point maximum reduction, would yield a negative update in the range of 5 percent. BBA also established a limit of 3 percentage points on how much the annual inflation update could be increased if spending was less than the target. For example, an inflation update of 2 percent increased by the 3 percentage point maximum increase would yield an update of 5 percent.

Additionally, BBA created a single conversion factor (previously there were three separate ones for different types of services). BBA also required that the practice expense component of the relative value calculation, which reflects a physician's overhead costs, be based on the relative resources involved with performing the service, rather than the physicians' historical charges. This change made the practice expense component of the calculation similar to the physician work component, and reflected actual resources. The change was phased in over four years, and was fully implemented in 2002. BBA further required that the professional liability insurance expense component of the relative value calculation also be resource-based. The law required that the resource-based practice expense and professional liability relative value systems be implemented in a budget-neutral manner. The BBA provisions affecting physicians accounted for about 3 percent of total BBA 10-year Medicare savings. Because physician payment accounts for about 17.6 percent of program payments in 1997, the physician savings in the BBA represented by these changes were relatively modest.

The Balanced Budget Refinement Act of 1999 (BBRA) made further revisions to the SGR in an attempt to help smooth out annual changes to physician payments such as blending cumulative and annual comparisons of target and actual spending. Beginning with the 2000 SGR, the law required us to revise previous SGR estimates based on actual data that became available after the previous estimates. BBRA also required us to make an annual estimate of the expected physician payment update for the succeeding year available to MedPAC and the public. This estimate is due on March 1 of each year, and is very difficult to make, because none of the claims used to determine actual spending are available by the time we are required to make the estimate. In 2001, we estimated that the 2002 update would be around negative 0.1 percent. However, when we determined the actual update, which was published 7 months later – on November 1, 2001, revised figures lowered the Gross Domestic Product figures for 2000 and predicted a slower growing economy for 2001 than was previously estimated. Further, 2001 physician spending was higher than our March estimate.

Additionally, in making updates to the list of codes for specific procedures that are included in the SGR, we discovered that a number of codes for new procedures were inadvertently not included in the measurement of actual expenditures beginning in 1998. Therefore, the previous measurements of actual expenditures for 1998, 1999, and 2000 were lower than they should have been. As a result, the physician fee schedule updates for 2000 and 2001 (5.5 percent and 5.0 percent, respectively) were higher than they should have been had those codes been included. The downward adjustment to the 2002 physician fee schedule was due, in part, to recoup the higher than intended expenditures in 2000 and 2001. The combination of these factors led to the large negative update for 2002.

As you can see, the process for calculating payments for physicians' services is highly complex. It is the result of years of efforts by Congress, previous Administrations, the Physician Payment Review Commission, and MedPAC to ensure that Medicare pays physicians as appropriately as possible. Today, while the underlying fee schedule and relative value system have been successful, we recognize that the update calculation has produced large short-term adjustments and instability in year-to-year updates. I know that you, Mr. Chairman, and others on this Subcommittee and elsewhere in Congress, are involved with legislative efforts to improve the formula. I want to work with you and the physician community to smooth out the yearly adjustments to the fee schedule in a way that is budget-neutral across all providers while ensuring that Medicare beneficiaries continue to get the care they need.

PROFESSIONAL LIABILITY COSTS

We are very disappointed that, despite our best efforts to find an administrative fix, physicians will face a negative 4.4 percent update this year, particularly because we understand the financial pressures that doctors face and want to ensure that Medicare beneficiaries continue to receive the care they need. One contributing factor is the cost of professional liability insurance. CMS’ Office of the Actuary (OACT) considers this cost as it develops and maintains input price indexes, or market baskets for the major medical sectors (hospitals, physicians, nursing homes, home health agencies, and the like) that are used to annually update Medicare payments. These market baskets reflect what it would cost in a future period to purchase the same mix of goods and services purchased in a base period. The mix of goods and services that providers purchase include labor (wages and benefits), utilities, pharmaceuticals, food, equipment, capital, and the lik e. One of these inputs is professional liability insurance, which we want to appropriately reflect in our market baskets.

In fact, Medicare physician payments for malpractice are determined partly by relative value units and partly by other elements of Medicare’s physician fee schedule. Payments for each of over 7,000 services under the fee schedule are based upon three factors:

• Relative value units (RVUs) for each service, reflecting the relative amount of physician work effort, practice expenses, and malpractice insurance expenses involved with furnishing each service;
• A dollar conversion factor that translates these RVUs into monetary payment amounts, and;
• Geographic practice cost indexes (GPCIs) for physician work, practice expenses, and malpractice insurance expenses to reflect differences in physician practice costs among geographic areas.

All three of these factors affect the total payment amount for a service. There is a malpractice element in each of these factors.

The first way that malpractice is reflected in the Medicare physician payment system is through a specific component of RVUs for each service for malpractice expenses. Since January 1, 2000, the statute has required that the RVUs for malpractice be based on the resources physicians actually expend to acquire professional liability insurance. We established resource-based malpractice RVUs through notice an d comment using a methodology that incorporates actual malpractice premium data and weighting by specialty and frequency of each service. The law requires that we revise the malpractice expense RVUs no less than every 5 years. If malpractice insurance premiums rise for some specialties more than for others, the higher expenses would be reflected in the periodic revisions we make to the malpractice expense RVUs. Subsequently, malpractice expense RVUs for services typically performed by these specialties would increase. We most recently revised these relative values for 2001. By law, we must make these revisions to relative values in a budget neutral manner, which means that if we increase the payment for a service then we are required to reduce the payments for other services. Because malpractice represents 3.2 percent of the average physician fee (physician work is 54.5 percent and practice expenses are 42.3 percent), even relatively large changes in premiums for some specialties would have relatively modest effects on Medicare payments.

The second way that malpractice is reflected in the physician payment system is through the annual update to the fee schedule conversion factor. The statute specifies a formula for the update based on the MEI adjusted by performance under the SGR system. The MEI is an inflation index measuring year-to-year changes in physician practice costs. One component of the MEI is specifically designed to recognize national year-to-year changes in the costs of malpractice insurance. If malpractice insurance premiums increase nationally, then these higher costs are incorporated into a higher MEI, which subsequently raises all payments under the fee schedule. These calculations are made each year by the Office of the Actuary based on data collected from major insurers. For 2003, the malpractice expense component of the MEI, which represents about 3.2 percent of the total index, was increased by 11.3 percent to reflect the recent liability insurance premium increases that have occurred nationwide. Thus, increases in malpractice insurance costs are reflected in Medicare physician payments relatively quickly.

The third way that malpractice is reflected in the physician payment system is through the geographic adjustment among areas. The statute requires a separate geographic adjustment for the malpractice RVU component to reflect the relative costs of malpractice expenses in different fee schedule geographic areas compared to the national average of such costs. There are 89 physician fee schedule geographic areas. Thirty-four of these areas represent one state apiece, while the other areas represent portions of the remaining 16 states. If malpractice insurance premiums increase more in some geographic areas than others, these higher geographic-level expenses are reflected in the periodic revisions that are made to the geographic practice costs indexes for malpractice expenses. These revisions are made in a budget neutral manner, no less than every three years, as required by law. The next revision will take effect in 2004; these changes will be proposed in the notice of proposed rulemaking for the annual update, which we plan to publish in Spring 2003.

To make these changes, we must gather data from the Insurance Commissioners in each State, which is an involved undertaking that requires the cooperation of 50 State offices. To make adjustments for 89 geographic areas covering the entire country requires much more detailed data than the data used to measure year-to-year national changes in costs for the MEI calculation. I am personally following up to ensure we have the best and most recent data possible for this revision. We have examined whether we should make these updates more frequently in the future. Our experience has been that the changes in the malpractice geographic practice cost indexes from revision to revision have been relatively small, and the effects on physician payments, given only 3.2 percent of these payments are affected by these index values, are also small. For example, a 10 percent increase in malpractice costs in a geographic area would result in only a 0.32 percent increase in Medicare payments in that area, and very small reductions in payments in all other areas. Collecting this data is not trivial and involves time and resources for both Federal and State governments. On balance, given that the annual changes in the Medicare Economic Index capture overall increases in malpractice costs, we do not believe that making revisions in the malpractice geographic practice cost index more often than every 3 years would be a efficient use of resources.

In short, Medicare revises physician payments to reflect changes in malpractice premiums on an annual basis. We do this through changes in the Medicare Economic Index, and thus the update to the physician fee schedule. Changes in malpractice relative value units, which reflect any changes that may have arisen across physician specialties, have been accomplished relatively recently. The malpractice geographic practice cost index in the next update will incorporate the latest available data. While I believe we need to address increases in malpractice premiums forthrightly, I also believe the Medicare physician fee schedule already does a reasonable job in this respect.

HOSPITALS’ PAYMENT

I also know that this Subcommittee is concerned about how we update hospital payments. Medicare paid approximately $100 billion in FY 2003 for Medicare inpatient hospital services. Medicare’s approximately 6,000 inpatient hospitals are paid under a prospective payment system (PPS), which is updated annually. This update factor is set in law and determined primarily by the projected increase in the hospital market basket index, which measures the costs of goods and services purchased by hospitals.

CMS has forecasted the FY 2003 market basket to be 3.5 percent. However, the Medicare, Medicaid, and SCHIP Benefits Improvement and Protection Act of 2000 (BIPA) established an update of market basket minus 0.55 percentage points for FY 2003. Therefore, the actual FY 2003 hospital market basket update is 3.5 percent minus 0.55, or 2.95 percent. Since the inception of inpatient PPS, hospitals have only once received a full market basket update (FY 2001), and on average, the actual update has been approximately 40 percent lower than the market basket increase.

Despite the difference between the payment update and the market basket increase, inpatient hospital margins have remained very high. In fact, since the early 1990's, there has been a significant drop in the number of hospitals with negative inpatient margins. For example, 61 percent of hospitals had negative inpatient margins in 1991 compared to approximately 25 percent in 1999. Moving forward, we need to continue to ensure that hospitals are paid adequately. In the meantime, we will continue to work hard to do what we can to help physicians and other providers in a variety of ways.

HELPING PROVIDERS OUTSIDE OF PAYMENTS

I worked in the hospital industry for years, and I know how frustrating it can be for physicians and providers to work with Medicare. We know that in order to ensure beneficiaries continue to receive the highest quality care, we must streamline Medicare's requirements, bring openness and responsiveness into the regulatory process, and make certain that regulatory and paperwork changes are sensible and predictable. This effort is a priority for me personally, as well as for Secretary Thompson and President Bush. And we have a lot of activities underway to make Medicare a more physician- and provider-friendly program.

To promote improved responsiveness, we created eleven "Open Door Policy Forums" to interact directly with physicians, as well as beneficiary groups, plans, providers, and suppliers, to strengthen communication and information sharing between stakeholders and the Agency. All of these Open Door Policy Forums facilitate information sharing and enhance communication between the Agency and its partners and beneficiaries. My goal is to make CMS an open agency – one that explains its policies to the beneficiaries and providers who rely on us.

Furthermore, our Physicians' Regulatory Issues Team (PRIT), chaired by an emergency room physician, Dr. Bill Rogers, integrates practicing physicians into our decision making process, allowing us to develop policies that will better serve beneficiaries and physicians. Specifically, PRIT members work within the Agency to serve as catalysts and advisors to policy staff as changes and decisions are discussed. These outreach efforts allow us to hear from physicians and all other Americans who deal with our programs. We are listening and we are learning. I am committed to making lots of common-sense changes and ensuring that the regulations governing our program not only make sense, but also are in plain and understandable language. This will go a long way in alleviating physicians' fears and reducing the amount of paperwork that, in the past, has all too often been an unnecessary burden on physicians.

CONCLUSION

I took this job because I know how important Medicare, Medicaid, and SCHIP are to Americans, and because I want to make a difference in improving our health care system. I am just as frustrated as you and all of the physicians that you hear from when it comes to how confusing and complex these programs are, and I am working hard to improve them. I also am working hard to monitor beneficiary access to care, and ensuring that America's elderly and disabled can receive the high quality care they need and deserve.

The Administration understands that the physician payment system is complex and will continue to work with Congress to smooth out the physician update system. I completely support legislative efforts to fix the short-term and long-term problems with the physician update. We owe it to America’s physicians to fix the system so that they can continue to provide Medicare beneficiaries with the vital care they need. Thank you for the opportunity to discuss this important topic with you today. I hope that I have helped to explain the issues, and I look forward to answering your questions.