Escaping from Medicare’s Flawed Physician Payment System

What This Paper Covers
There have been many changes over the years to Medicare’s reimbursement system for physicians and other clinicians. This paper summarizes these changes before and after the enactment of the Physician Fee Schedule, including policies for setting payment rates and controlling aggregate expenditures. It then discusses how current policy balances the goals of maintaining access to physician services for Medicare beneficiaries, containing costs for patients and the Medicare program overall, and minimizing distortions of the health care sector resulting from payment policy.

What We Found
Medicare policies enacted in recent decades have helped to contain costs by slowing the growth of payment rates. However, expenditures have continued to grow with the volume of physician services and with rapidly rising spending on other Part B services. Access to physician care seems to be stable or improving by several metrics, but it is important to maintain this trend in the long term. Physician payment in Medicare continues to rely on a fee-for-service approach that incentivizes quantity of care over quality and administrative pricing that misestimates the value of health care services.

Why It Matters
Balancing adequacy, containment, and accuracy of payments has important consequences. Inadequate payment levels could discourage doctors from participating in Medicare and thereby compromise access to care for seniors. On the other hand, excessive payments directly increase costs borne by patients and their families while also undermining the fiscal sustainability of the Medicare program for taxpayers and future beneficiaries. Administrative pricing and fee-for-service payment increases these risks in both directions because government agencies tend to misestimate the value of services, which are best determined by the aggregate preferences of consumers expressed through market prices.

Policy Suggestions
While maintaining patients’ access to physician services is important, doing so by simply setting payment rates to an external measure of inflation such as the Medicare Economic Index is a flawed approach because it would increase costs and do nothing to ensure the accuracy of payment rates. Policymakers should tie increases to physician payment to other policy changes, such as:

• reducing overpayments on outpatient hospital services, Part B drugs,
  clinical laboratory services, and durable medical equipment in order to
  contain overall Part B spending;
• reforming the Physician Fee Schedule, including by incorporating marketbased pricing and strengthening budget-neutrality requirements;
• eliminating or significantly reforming failed efforts to promote value-based care such as the Merit-Based Incentive Payment System and financial enticements to participate in advanced alternative payment models.

The most realistic prospect for a long-term transformation of Medicare that balances payment adequacy and efficiency while moving away from a government-driven approach to price-setting and quality improvement would be greater growth in Medicare Advantage.

Medicare’s method for reimbursing physicians has gone through many changes over the years. The last major reforms in 2015 were intended to simultaneously avoid massive cuts, control spending, and reform Medicare’s payment methodology to encourage greater value. While these were worthy goals to pursue, success has been mixed. In recent years, Medicare participation by doctors has been high, but stagnating pay could reduce this in the long run and thereby compromise access to care for seniors. On the other hand, overpaying for physician services would inflate costs of care and worsen the program’s already-strained finances, necessitating higher taxes or benefit cuts and potentially contributing to a fiscal crisis. Medicare’s current government-driven payment policies make under- or overpayments more likely because they are not based on the economic value of services as dictated by consumers in a free market. This policy brief outlines the history of physician payment in Medicare, its current shortcomings, and potential options for reforming it.

The Evolution of Physician Payment Policy
Physicians and clinicians have consistently been the second highest source of Medicare expenditures behind hospitals, accounting for about 25 percent, or $222 billion, in 2021.¹ Given this scale, physician payment policy in Medicare Part B has attracted much attention— and many changes—over the years. At the start of Medicare, doctors had a great deal of leverage over pay. Reimbursement was based on “customary, prevailing, and reasonable” rates—similar to “usual, customary, and reasonable” private insurance rates at the time. Medicare paid for a service at the lowest of the physician’s billed charges, the service’s customary or median charge, or the service’s “prevailing charge” in a location.² Providers could also charge more than the Medicare payment rate and balance bill patients for the remainder.³

This process encouraged doctors to raise their fees to obtain higher reimbursement. As a result, annual growth in clinician spending averaged 16 percent in the 1970s.⁴ In response to this unsustainable trend, Congress froze Medicare fees from 1984 to 1986 and disallowed balance billing by participating providers.⁵However, physician spending still increased during this time due to increases in the volume of services and changing practice patterns by doctors.⁶

Over time, commercial insurers began to set their rates in fee schedules that established relative values for each service based on prevailing charges.⁷Congress in turn instructed the Department of Health and Human Services (HHS) to develop a fee schedule based on a resource-based relative value scale (RB-RVS). The RB-RVS weighs each service by the estimated resources needed for physician work, practice expenses, and liability insurance, with each adjusted by estimated geographic differences in costs such as wages, rent, and other input costs.⁸ In 1989, Congress enacted the Physician Fee Schedule (PFS) based on these components. It also restricted balance billing of Medicare enrollees by nonparticipating practitioners to 15 percent of Medicare rates.⁹

The PFS sets payment rates for individual services by multiplying a dollar-value conversion factor by the services’ relative values. For example, with the 2023 conversion factor of $33.8872, a particular service in a given location with a relative value that is 1 percent higher than the overall average receives a base payment rate (before other modifiers) of $33.8872 * 1.0100 = $34.2261. The Centers for Medicare and Medicaid Services (CMS) updates the PFS’s components in annual rulemaking. CMS receives—and often accepts—recommendations from the American Medical Association’s (AMA’s) RB-RVS Update Committee (RUC) for updating the RB-RVS. Regulatory changes to the PFS must typically be budget neutral.¹⁰ Figure 1 below summarizes the PFS’s payment methodology.

Fiscal Constraints
Despite changing from a charge-based to a cost-based price-setting methodology with the PFS, Medicare’s payments to doctors continue to be on a fee-for-service (FFS) basis. Doctors earn more when they provide more services, even if doing so is unhelpful or even harmful for patients. Therefore, policymakers have experimented with mechanisms to control overall expenditures.

In 1975, HHS created the Medicare Economic Index (MEI) at Congress’s behest to track increases in physician practice costs and earning levels, using it to limit the growth of prevailing charges.¹¹ Still, expenditures continued to grow rapidly, and the initial conversion factor set under the PFS incorporated overpayments from the previous system.¹² Therefore, when Congress enacted the PFS in 1989, it also created the Volume Performance Standard (VPS). The VPS was a target growth rate for physician spending calculated from projected and historical changes to payment rates and the volume of services. Under the VPS, PFS payment updates were based on (1) the MEI and (2) expenditure growth relative to the VPS target. For example, if Medicare physician spending increased by 11.1 percent in Year 1 compared to a 9.1 percent target for that year, then PFS rates in Year 3 would update by the percentage change in the MEI minus the 2 percent excess growth from Year 1 relative to the VPS, with a cap on downward adjustments. If actual growth was lower than the target rate, the difference would be added to the MEI.¹³

In 1997, Congress replaced the VPS with the Sustainable Growth Rate (SGR) formula due to the instability of annual PFS updates under the VPS, which ranged from 0.6 percent to 7.5 percent.¹⁴ The SGR functioned similarly to the VPS, except that it also accounted for economic growth. Within a few years, continued growth in physician spending coupled with a recession led the SGR to mandate a 4.4 percent cut for 2003. The SGR restricted annual adjustments to be between -7 percent and +3 percent, but this reduction was unpopular, particularly as the cuts applied across the board regardless of how cost effective individual physicians or practices were. As the SGR required more cuts, Congress began to override them with “doc fixes.”¹⁵ These doc fixes did not necessarily mean pay raises for physicians, though they did sometimes change the SGR itself. For example, the Tax Relief and Health Care Act of 2006 bumped the 2007 update from -5.0 percent to +0.0 percent and required CMS to update future PFS rates based on SGR adjustments despite Congress not letting them go into effect. This meant the SGR’s impact accumulated over time. Although Medicare’s spending increases eventually fell below the SGR’s targets, the SGR formula, which accounted for years of excess cost growth, called for a 20 percent payment cut in 2015.¹⁶ Figure 2 below shows increases to PFS rates over time, including the deviations between the updates required under the SGR and those enacted by Congress starting in 2003.

Frustration among lawmakers with the political pressure to mitigate SGR cuts and develop offsets led to the passage of the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA). MACRA repealed the SGR, created a new Quality Payment Program with financial incentives for value-based care, and set a fixed schedule of PFS updates.¹⁷ Annual payment updates ranged between -3.3 and +0.3 percent during 2016 to 2023, including with congressional intervention to prevent bigger cuts schedule for 2021 through 2024.¹⁸

Medicare payment policy continues to be a source of contention, with competing concerns about the adequacy of PFS updates on one hand and Medicare’s excessive costs on the other hand prompting lawmakers to consider further reforms to MACRA. Policy changes should safeguard both access to care and program finances, ideally while minimizing government distortions of the health care sector.

Payment Adequacy
Although Congress did not perfectly adhere to the SGR or MACRA statutory updates, they have arguably succeeded in containing physician costs in Medicare. The PFS conversion factor declined roughly 8 percent between 1998 and 2023 in nominal terms, and expenditures fell from 48 percent to 32 percent of FFS Part B spending, indicating that other services grew at a faster rate.¹⁹ Although PFS per capita spending rose 128 percent in that time—versus 66 percent for gross domestic product (GDP), 88 percent for overall inflation, and 126 percent for medical inflation—this was due to growth in volume and intensity of those services.²⁰

Physician groups have argued that a continued decline in fees will compromise Medicare beneficiaries’ access to care, as lower pay may attract fewer doctors to participate. Medicare’s trustees and the Medicare Payment Advisory Commission (MedPAC) have cited these as long-term concerns as well, although MedPAC reports have often found that access to such services in Medicare is roughly equal to or better than the private insurance market, which offers higher rates.²¹ Figures 3 and 4 also demonstrate this based on alternative metrics.

Budgetary Concerns
Containing per unit physician costs has been helpful but insufficient to put Medicare on a sustainable trajectory: expenditures are expected to rise by nearly $1 trillion between 2022 and 2033, from 3.0 percent to 4.5 percent of GDP.²² Part B is the fastest growing part of the program.²³ Part B premium levels are based on overall program spending, and coinsurance is typically applied as a fixed percentage of costs, so Medicare cost growth directly raises beneficiary expenses. Because general revenues mostly fund it, Part B will also gradually crowd out other federal priorities or increase deficits, potentially contributing to a debt crisis with dire economic consequences.²⁴ As the number of Americans over age 65 grows, Medicare needs to be even more efficient to be sustainable. Policy changes that simply increase physician payments will leave the program in a more tenuous position.

One proposal would tie PFS updates to the percentage growth of the MEI. The Medicare trustees found that this and other payment changes, such as extending bonuses from MACRA, would increase spending as a share of GDP by 0.4 percentage points over ten years relative to current law ($97 billion) or 2.9 percentage points over 20 years (roughly $824 billion).²⁵ MedPAC found that updating PFS rates by half of the MEI would increase spending by up to $10 billion over five years (see Figure 5). If PFS conversion factor updates between 1998 and 2023 had been based on MEI, all else being equal, spending would have been over $500 billion higher during that time, as Figure 6 shows.

Payment Distortions
Reforming physician payment also offers an opportunity to remedy two major distortions in the Medicare program: FFS reimbursement and administrative price-setting. Medicare FFS incentivizes the provision of more health care services regardless of their necessity or effectiveness, leading to significant spending on wasteful or even harmful items and services.²⁶ Furthermore, the incentives faced by doctors uniquely impact overall health care costs with their prescriptions, referrals, and other recommendations. The “physician’s pen” affects 80-85 cents of every health care dollar spent.²⁷ MACRA tried to move Medicare away from FFS toward “value-based care” by using a portion of physician payments to reward performance on quality metrics in the Merit-Based Incentive Payment System (MIPS) or participation in advanced alternative payment models (APMs).²⁸ This effort has generally failed to achieve its goals, and significant reforms are needed to align value with consumer preferences rather than the priorities of federal bureaucracies.²⁹

The PFS also provides multiple avenues for government and special interests to directly influence the health care sector. In most markets, prices are based on economic value as determined by interactions between consumers and producers. Medicare payment systems, which influence or are directly adopted by the private insurance market, instead rely on calculation methodologies that give a veneer of technocratic sophistication. But no central planner can replicate the organic order that emerges from decentralized decisions of numerous individuals and firms.

There are also clear examples of purely political considerations influencing this process, especially in the PFS. Congress’s frequent interventions over the past few decades—although motivated by a well-intentioned desire to balance access to care and fiscal responsibility— are a key example, as they do not account for the value of physician services to beneficiaries.

The predominant role of the AMA in setting RB-RVUs in the PFS is another instance of this, as federal officials usually defer to the RUC, allowing practitioners to decide specialties’ pay relative to others. However, the RUC’s estimated payment weights often do not comport with reality.³⁰The Government Accountability Office has suggested that the RUC’s results are flawed and has recommended that CMS better document its process of establishing relative values, including validating RUC recommendations and identifying potentially misvalued services.³¹

The accuracy of Medicare payments is not simply an academic question. Administrative pricesetting by the government significantly risks shortages by underpaying for certain services or waste by overpaying for them. Analysts have argued, for example, that the PFS routinely overvalues specialist pay at the expense of primary care services, which has distorted the supply of practitioners.³²

Medicare payment policy should ensure access to high-quality physician services, contain costs for beneficiaries and taxpayers, and move away from FFS administrative pricing. The sections below discuss how potential policy options would measure up against these goals.

Incrementalism in the PFS
The simplest approach suggested by some policy experts has been to leave the PFS largely the same and replace MACRA’s statutory payment updates with a measure of inflation such as the MEI, similar to how other Medicare payment systems use a wage or market basket index.³³ However, this would increase government spending and patient costs while continuing to rely on FFS administrative pricing. If policymakers overreact to the potential future risk of reduced provider participation in Medicare, the harms created by these additional costs may outweigh the benefits.

Some have suggested other incremental reforms to the PFS, such as providing more oversight of the RUC, rebalancing pay by specialty or geography, calculating provider costs similarly to other payment systems, or incorporating more empirical data to determine relative weights.³⁴ While certain individual policies to address identifiable shortcomings with the PFS are sensible, tweaking the PFS would retain a flawed approach that leads to mismeasurement of the value of services. Instead, lawmakers should couple any changes to physician payment that increase spending with more substantive reforms to the program and should not increase overall Medicare costs.

Reducing Part B Spending
One approach would be to offset new physician spending with changes elsewhere in
Part B.

Part B overpayments are common for outpatient hospital services in particular. A major reason for this is site-of-service differentials, where Medicare pays hospitals twice as much on average for routine services that could be performed just as safely and effectively in a physician’s office or ambulatory surgical center. Hospitals have a financial incentive to acquire
independent physician practices and convert them to hospital outpatient departments in order to command higher outpatient hospital payment rates, which have grown faster than physician fees. Pursuing site-neutral payments for such services – that is, paying the same rate for the same service regardless of the setting – would be a straightforward way to reduce overpayments.³⁵

Outpatient drug spending has also grown rapidly. Part B’s price controls, which reimburse most drugs at 106 percent of their average sales prices, are a major driver of this overspending. Reducing these baked-in overpayments or enacting more comprehensive reforms to incorporate market-based pricing would be an improvement. For example, policymakers could move coverage of Part B drug benefits to Part D, where spending growth has been far below expectations because of competition among private plans.³⁶ Addressing federal programs that further distort the drug market, such as the 340B program, is also an option. Under 340B, hospitals acquire drugs at a sizable discount without having to pass savings along to patients. Medicare purchases these drugs at the same rate as nondiscounted drugs. Lawmakers could authorize or require CMS to adjust payment for 340B drugs.³⁷

Continuing to implement other market-based policies in Part B for clinical laboratories and durable medical equipment (DME) would reinforce previous reforms. Congress created a DME competitive bidding program starting 2008 and has conducted or re-conducted several rounds of bids since then. This program has driven down the cost of certain products, but CMS significantly narrowed the latest round of bids in 2021. Lack of guidance on the 2024 round has created uncertainty. Policymakers should require more timely guidance for future rounds of the competitive bidding program and consider expanding it, including to more DME items. In 2018, Medicare updated its clinical laboratory fee schedule so that payment rates would be based on private payer data. Industry groups have raised concerns about the accuracy of these rates, and Congress has prevented them from going into effect since the start of the COVID-19 pandemic. Finding a sound way to implement these policies—even if that requires updating the methodology for calculating rates—would improve the accuracy of Medicare payments for laboratory services.

Overhauling the PFS
Reducing other Part B spending would help overall program finances, but it would not control physician spending. Lawmakers could mitigate this risk by limiting pay raises, given that physician participation in Medicare is currently strong, and coupling them with other meaningful changes. Adding guardrails to the PFS is one possibility. For example, CMS routinely creates new procedure codes through the rulemaking process and must predict the impact on future expenditures. Despite budget-neutrality requirements, utilization growth for new items and services can nonetheless drive spending up over time. Improving budget-neutrality requirements—including by adding retrospective adjustments of prior years’ utilization estimates or requiring oversight of new codes—could help control spending.³⁸

Another option is to incorporate more pricing information from the private market, such as rates paid by Medicare Advantage (MA) plans. Medicare payments can be set at these rates, or CMS can use MA rate data as an input for PFS rate calculations. Using private payer data for the RB-RVS could eliminate the need for an RUC and move the PFS away from administrative price-setting. MA rates tend to resemble traditional FFS Medicare rates more closely than commercial insurance rates, but MA plans tend to attain similar discounts as commercial plans do in areas where FFS overpays, such as clinical laboratory services and DME.³⁹

In the near term, using private payment data may have a limited impact because private payers frequently negotiate based on percentages of Medicare rates. Over time, however, the market could adapt and incorporate a more accurate valuation of physician services, balancing cost reduction and access to care. CMS finalized a similar approach for inpatient hospital services in 2020 but reversed course in 2021 after the transition from the Trump administration to the Biden administration.⁴⁰

Quality Payment Program Reform
Lawmakers should also address MACRA’s Quality Payment Program as part of broader payment reforms. The performance incentives under MIPS have been ineffective and burdensome for participating physicians. Many stakeholders, such as physician groups and accountable care organizations, may prefer that Medicare provide bonuses for participating in advanced APMs rather than requiring pay-for-performance in MIPS, but there is no evidence that the bonuses increase APM participation.⁴¹ Extending bonuses at just 1.75 percent is estimated to boost Medicare spending by about $680 million per year.⁴² APMs themselves have also had a mixed record; models developed and managed by the Center for Medicare and Medicaid Innovation (also called the CMS Innovation Center) have been found to cost $5.4 billion on net from 2010 to 2019 rather than saving $2.8 billion as expected, and the Congressional Budget Office now projects that new models will likely increase spending on average.⁴³ Only three out of nine models with advanced components have lowered costs, of which two had non-advanced tracks that did not have separate evaluations, as Table 1 above shows. Furthermore, studies have cast doubt on the official results of the Medicare Shared Savings Program.⁴⁴ It makes little sense to continue to reward participation in models that often do not work.⁴⁵

Policymakers should ideally eliminate MIPS and advanced APM bonuses, including the differential payment updates between their participants. Instead of MIPS, CMS should encourage private payers and other entities to provide quality metrics that enable patients to shop between providers on the basis of quality. To the extent CMS does measure quality, it should focus its efforts on penalizing poor performers based on misdiagnosis, mistreatment, or appropriateness metrics.

If lawmakers do not eliminate MIPS, or if they otherwise choose to extend APM bonuses, they should at least significantly modify the bonuses. For example, they could require clinicians to take on more downside risk or encourage more participation in all-payer APMs rather than Medicare APMs in order to encourage private payers, including MA plans, to develop more successful and innovative payment models. Other design changes could fine-tune these financial incentives even more and encourage greater parity in their availability to providers, such as proportionally basing bonuses on the share of patients that participate in qualifying models rather than providing or withholding the full bonus based on a specific threshold and limiting the amount of time clinicians can receive bonuses.

Beyond Fee-for-Service
In the absence of performance-based quality programs, substantial changes across traditional Medicare’s payment systems would be necessary to move it beyond FFS. Shifting more financial risk from taxpayers to health care providers could incentivize quality improvement and cost savings. Two examples of this would be to implement episode- or population-based payments in order to encourage more efficient health care delivery while reducing the government’s need to set prices for individual services. Enacting such policies directly in Medicare payment systems would be more effective than relying on APMs, which have overlapping impacts that diminish their effectiveness, incentive payments that nullify potential cost savings, limited downside risk, and limited scope (i.e., they are temporary and typically targeted toward specific provider types, health care procedures, or geographical areas).

Episode-based payments, such as bundling, would provide a single payment for all the services for a single episode of care or condition. CMS would set a target price for a primary procedure and any other ancillary items and services that are typically related to it, and a provider who delivers care at a lower cost would retain a profit, while those who exceed it would incur losses. Population-based reimbursement, such as capitation, offer providers fixed, regularly occurring payments per patient. Providers again retain any surplus or deficit they incur from delivering care. A more incremental approach is shared savings, where a provider receives a target price and retains a share of the difference between it and the actual costs as either a profit or a loss.

Congress could fully or partially replace the PFS with episode-based payments or population-based reimbursement or combine the two to mitigate the risks of fully relying on any single approach. However, such payment policies would be most effective when implemented across payment systems, as a single patient or episode of care might require services from multiple providers, requiring policymakers to substantially restructure Medicare. There would still be some risk of CMS calculating the wrong payment rates or target prices, which could reduce the ability to estimate savings in practice (and has been a flaw of accountable care organization-based models that are based on administrative benchmarks). There would also be a limited ability to assess performance by individual physicians. Still, episode- or population-based payments offer an improvement on these fronts compared to the status quo in Medicare, and policymakers could mitigate the former risk by calculating target prices and capitation rates based on private payer data rather than simply using PFS rates as a basis.

Another long-term approach to Medicare reform is building on MA’s success. MA plans have more flexibility to design their contracts, negotiate payments with providers and manage utilization, which encourages them to secure discounts and deliver care more efficiently than traditional Medicare. They also receive risk-adjusted capitated payments from CMS, incentivizing them to maximize the value of care for their enrollees, including those with higher health care costs, by bearing risk for the total cost of their care. Finally, MA plans are required to pass along savings from the bidding process onto enrollees in the form of extra benefits or lower cost-sharing and compete with other plans that MA enrollees can choose from. Many of these flexibilities and payment mechanisms echo the features of advanced APMs, with a key difference being that MA plans have an economic incentive to adapt to market conditions in order to deliver core benefits more efficiently, unlike CMS.

MA has already grown rapidly to about half of all Medicare beneficiaries. Policies that enable it to compete on equal footing with traditional Medicare will help deliver better physician (and other) services to seniors on a more efficient basis.

Years of tinkering with Medicare payment policy for physicians have not produced a suitable long-term approach. Enacting yet another government-driven approach to payment would only replicate the same problems and fail to balance access to care with fiscal sustainability. The dissatisfaction with the status quo presents an opportunity for policymakers to take steps to move the PFS—and Medicare as a whole—away from FFS administrative pricing toward a long-term vision that bases payment on the value of health care services to patients themselves.

Footnotes