Fee‐for‐service payment is not the (main) problem

Abstract

Objective

To understand the effect of physician payment incentives on the allocation of health care resources.

Data Sources/Study Setting

Review and analysis of the literature on physician payment incentives.

Study Design

Analysis of current physician payment incentives and several ways to modify those incentives to encourage increased efficiency.

Principal Findings

Fee‐for‐service payments can be incorporated into systems that encourage efficient pricing – prices that are close to the provider's marginal cost – by giving consumers information on provider‐specific prices and a strong incentive to choose lower cost providers. However, efficient pricing of services ultimately will need to be supplemented by incentives for efficient production of health and functional status.

Conclusions

The problem with current FFS payment is not paying a fee for each service, per se, but the way in which the fees are determined.

1. INTRODUCTION

There has been a growing awareness that high prices (fees), rather than high quantities of services, are the main reason that per capita spending on health care services is higher in the United States than in other developed countries. ¹ , ² , ³ Health policy analysts have argued that fee‐for‐service (FFS) payment creates an incentive for physicians to prescribe more services, including more low‐value services. These are separate but related criticisms. The first is that FFS payments lead to an increase in the overall cost of health care services. Because total spending is equal to price times quantity, high prices increase total spending both directly, holding quantity constant, and indirectly through an induced increase in the supply of services. This criticism often ignores the effect of high prices on the demand for services, particularly by the increasing number of consumers in high deductible health plans. ⁴ The second criticism is that FFS payments also distort the relative prices of different types of health care services. ⁵ , ⁶ , ⁷ , ⁸ Our observation is that neither criticism is a criticism of FFS payments per se, but instead are criticisms of the way in which FFS payments are determined in the United States. The result is distorted prices that lead an inefficient allocation of health care services.

What do we mean by distorted prices? When economists focus on efficiency, the benchmark they use for distorted prices is not FFS payments in other countries or in private versus public health insurance programs in the United States. Instead, prices are said to be distorted to the extent that they diverge from the physician's marginal cost of supplying services. Why might the prices of physician services be distorted and if they are, why has the market not brought prices in line with marginal costs? The economist's answer to both questions is market failure and there is no shortage of candidates for market failure when it comes to health care services.

There are two mechanisms for setting the price of health care services in the United States: administrative and competitive systems. We compare the two approaches and discuss them in the context of the traditional FFS Medicare program. However, all of our observations are generalizable to other payers who often follow Medicare's lead in their physician payment policies.

This paper focuses on ways to achieve efficient pricing of physician services, but the principles we discuss could be, and in many cases already have been, applied to providers of other health care services. In the summary section, we acknowledge that efficient pricing, alone, may not be sufficient to address the affordability problem that is beginning to impact middle‐class Americans.

2. ADMINISTERED PRICES IN THE FFS MEDICARE PROGRAM

The US Medicare program uses an administered pricing system to set physician fees. Berenson and Ginsburg have provided a comprehensive historical summary and international comparisons. ⁹ The Centers for Medicare and Medicaid Services (CMS) realizes that without information on the physician's marginal cost of providing services it cannot set efficient Medicare fees, and CMS has limited information on the physician's true cost of providing services. Physicians, on the other hand, know quite a lot about their true costs. CMS’s approach is to ask physicians to reveal their “true” costs. This request is made to the American Medical Association's Specialty Society Relative Value Update Committee (RUC).

Physicians serving on the RUC face no direct penalty for intentional or unintentional errors in evaluation of their costs—that is, if the errors or misrepresentations even are discovered. Specialty societies send out practice cost and physician work surveys and provide courses on how to complete the surveys, introducing the possibility of bias. ¹⁰ Ideally, administration of the survey should be not be carried out by those who stand to benefit from the responses. The entire process, in short, depends on a great deal of trust in physicians to do the right thing.

The RUC produces a set of relative value units (RVUs) for each procedure, which are converted into dollar amounts (fees) by a multiplier. The fees for new services, even if RVUs are appropriate when generated, typically are higher than they need to be after a number of years, due to the productivity gains after a new service becomes more established in practice. Efforts to address misvalued services under the Affordable Care Act are coordinated under similar processes as the general review; therefore, the same distortions can occur.

Unfortunately, if Medicare is unwilling to validate RUC estimates with better data, there is little else constraining the overvaluation of services. Private payers use RVU multipliers to adjust fees, but it may be difficult to address individual service prices on a granular level when they contract with many providers. With one exception, private payers do not participate in the effort to call out misvalued services by bringing those to the RUC: In February 2018, Anthem nominated seven services for re‐assessment by CMS: hip and knee arthroplasty, colonoscopy (with lesion removal), two computed tomography services, esophagogastroduodenoscopy (EGD), and a transthoracic echocardiogram. ¹¹ Ideally, all payers would aggressively and independently identify misvalued services and correct for errors of valuation.

To address the pricing problem, new payment reform models, such as Accountable Care Organizations (ACOs) and bundled payment initiatives, have been introduced. These are good steps toward more efficient pricing, but ACOs and bundled payment initiatives primarily create incentives to constrain the volume of services provided to beneficiaries. They do nothing to reform the fee‐setting process or address misvalued services, and are unlikely to bend the cost curve. Another approach is to adopt a different method of setting fees – one that provides a stronger incentive for correspondence between prices and marginal costs.

3. COMPETITIVE PRICING

In the economist's ideal and perfectly competitive market, prices are the result of market‐wide supply and demand and individual producers (physicians) treat prices (fees) as given, for example, beyond their control. As a result of competition, prices are driven down toward the producer's marginal costs.

The market for physician services is not perfectly competitive, however, for at least three reasons. First, patients are shielded from the true price differences among physicians by a myriad of tax subsidies, direct price subsidies unrelated to income, and the structure of health insurance itself. Second, patients have poor information about the relative value of the services prescribed by different physicians for patients in the same health condition and prices charged by different physicians for the same service. Third, consolidation in markets for physician services through large physician networks and physician‐hospital integrated delivery systems has resulted in significant market pricing power that creates quasimonopolies in geographic areas from market concentration. ¹² Dunn and Shaprio estimate that concentration of physician markets over the past 20 years is responsible for an 8 percent increase in fees on average, with substantially higher increases in more concentrated markets. ¹³

Competitive pricing approaches can focus on incentives for physicians, consumers, or both. The threat of exclusion from the set of preferred physicians or from coverage entirely, in the case of some narrow network health plans and Center of Excellence initiatives, ¹³ gives providers an incentive to agree to prices closer to their marginal costs.

Competition to be included in a health plan's network is a good start, but under most health insurance benefit designs, higher in‐network physician prices do not result in higher consumer out‐of‐pocket costs. The copayment for an office visit is the same regardless of the cost to the health plan. When a physician's higher price translates directly into a higher price for the consumer, an informed consumer has a stronger incentive to choose lower cost physicians. ¹⁴

4. REFERENCE PRICING AND TIERED COST SHARING

Two pricing approaches, reference pricing and tiered cost sharing, combine improved consumer information and strong financial incentives within an FFS payment system. Instead of being paid administratively determined fees, reference pricing would allow physicians to set their own fees either for specific services or for the RVU multiplier. Physicians’ fees would be reviewed by the health plan, for example, Medicare, and the health plan would set its payment for the procedure at a point in the distribution of prices –the reference price. If a patient preferred to seek care from a physician whose price exceeds the reference price, the patient would pay the difference out of her own pocket. As result, at least over the longer term, physicians would learn about their prices relative to their competitors and realize that prices above the reference price will be translated directly into higher out‐of‐pocket costs for the patient.

The California Public Employees' Retirement System (CalPERS) has used reference pricing with some success. ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ To date, reference pricing in the CalPERS applies to hospital and not physician services. However, the principles are the same: Information on prices flow from the provider to the payor, and consumers who choose higher priced providers face higher out‐of‐pocket costs. This encourages California hospitals to submit bids that are close to their marginal costs.

Tiered cost‐sharing has been applied to pharmaceuticals, hospitals, and physician services. ¹⁹ , ²⁰ Under tiered cost‐sharing, physicians are placed in groups based on their risk‐adjusted costs. Consumers who choose higher cost providers face higher deductibles and copayments.

Unlike high deductible health plans per se, reference pricing and tiered cost‐sharing give consumers both the information they need to choose more efficient cost providers and a strong financial incentive to do so. Unlike the typical ACO or shared savings model, high‐priced providers do not risk losing merely a portion of their revenue from each patient, but instead risk losing all the revenue from a patient who switches to lower cost provider.

Both approaches have limitations that require a thoughtful application, not indiscriminate adoption. Not all services are “shoppable.” ²¹ Reference pricing is best applied to specific elective services like total knee replacement, where ongoing care coordination among multiple physicians is less important. Tiered cost‐sharing for hospital and physician services must take into account existing organizational relationships, for example, in integrated delivery systems.

Reference pricing and tiered cost‐sharing require competitive market conditions, but not all states and localities have robust hospital or physician competition. Where there is a lack of competition, consumers could be given rebates or other incentives to access lower cost care outside of those areas. In competitive bidding for durable medical equipment services, CMS uses its own administered prices as the upper limit for what it will pay.22

Successful reference and tiered pricing depend on various supporting policy changes. For example, consumers are more sensitive to out‐of‐pocket cost differentials when they are paid with fully taxed dollars. If consumers pay these differences with tax‐deductible dollars, for example, from health saving or flexible savings accounts, the incentive to select lower cost physicians will be reduced (though not eliminated). ²³ If Medicare Part B adopted these pricing reforms, Medigap coverage would need to be limited. Such limits have a precedent: Under the Medicare Access and CHIP Reauthorization Act (MACRA), Congress limited the ability of Medigap policies to cover Medicare deductibles for new enrollees starting in 2020. ²⁴ The Medicare program could reward beneficiaries who choose lower priced physicians with refunds for their deductibles, and all beneficiaries could be rewarded with Part B premium rebates. Those approaches would be another form of “shared savings,” except the savings to the Medicare program would be shared with beneficiaries rather than with providers.

Potential adoption of reference pricing and tiered cost‐sharing by Medicare and private insurers points to the flexibility of these approaches. These tools are applicable under different health systems, ranging from (so‐called) “free market” health care systems to public and single‐payer models. Different systems likely would address equity issues across sicker, older, or low‐income populations.

Because reference pricing and tiered cost‐sharing have been applied more frequently in the hospital and pharmaceutical sectors, uncertainty exists as to whether application to physician services would produce the same level of savings. Surgeons, for example, may not be willing to accept lower fees, and their marginal costs may not vary as much as those of hospitals. But there is substantial variation in fees. In Minnesota's all payer database, the cost of a comprehensive eye examination, including diagnosis and treatment, ranges from $95 to $335. An MRI of the lumbar spine with and without contrast ranges from $781 to $4429. Vaginal delivery with routine care ranges from $2259 to $7118. ²⁵

Applying these changes in the Medicare program would face significant political hurdles. To increase political feasibility, the application of reference pricing or tiered cost‐sharing within the Medicare program would need to be based on financial rewards for beneficiaries who choose lower priced physicians rather than penalties for choosing higher priced physicians. Again, the point is to share Medicare's savings with consumers, rather than providers.

Consumer‐oriented incentives like reference pricing and tiered cost‐sharing are far less draconian from the physician's perspective than excluding high‐priced physicians from the program as CMS currently does for durable medical equipment, prosthetics, orthotics, and supplies. ²⁶

5. SUMMARY

The process of setting physician fees in the United States is flawed, but the primary flaw is not FFS payment per se, but the way in which prices are set. Administrative pricing provides no feedback loop from the physician's fees to the consumer's choice of physician. This flaw is not unique to administrative pricing. The same is true of all health insurance benefit designs, for example, uniform copayments and deductibles, that fail to pass the natural variation in physician fees onto consumers, thereby combining reliable price information with a strong financial incentive to choose lower priced physicians.

While we are not opposed to initiatives like bundled payments, moving to bundled payments would be unlikely to produce a significant improvement over current payment systems if the payment levels were set through a RUC‐like process. However, giving beneficiaries’ information on physicians who submit lower bids on the bundles, along with a strong incentive to choose lower priced physicians, could bring Medicare's prices closer to the physicians’ marginal costs.

Health care costs now are high enough to cause problems for middle‐class Americans. ²⁷ Bringing prices in line with physicians' marginal costs is a worthy goal, but efficient pricing alone may not be enough to solve the affordability problem. In addition to efficient pricing, a new set of incentives may be required that encourage efficient production of health. If up to thirty percent of health care services are wasteful, ²⁸ , ²⁹ , ³⁰ , ³¹ then the central problem is not how to price wasteful care close to marginal cost. The central problem is how to curtail wasteful care and how to incentivize the health care technology industry to produce innovations that are cost savings, not merely cost‐effective.

Currently, physicians who produce and maintain the individual's health at lower cost are financially penalized. In the future, they will need to be rewarded with greater patient volume, taken from their less efficient competitors.

Supporting information

Author Matrix

Dowd BE, Laugesen MJ. Fee‐for‐service payment is not the (main) problem. Health Serv Res. 2020;55:491–495. 10.1111/1475-6773.13316