Almost two decades ago, Gerard Anderson and colleagues published the seminal paper ‘It’s the Prices, Stupid’, identifying why health care spending in the US was so much higher than other developed countries. Spoiler alert: it’s the prices. Twenty years on, this conclusion has continued to be affirmed over time and in greater detail.
Yet, the many differences between national health care systems make it difficult to conclude that transplanting other countries’ prices to the US would be ‘right’ for the US. Nevertheless, the observation that prices are a significant determinant of high US health care spending motivates examination of American health care prices.
The price issue is most salient in the commercial sector, where prices set using market mechanisms are considerably higher than in the public sector. A Congressional Budget Office review suggests commercial prices for inpatient care are 182 percent of Medicare prices, commercial outpatient prices are 240 percent of Medicare prices, and commercial physician fees are 129 percent of Medicare fees. Price growth has also been rapid in the commercial sector, far beyond public sector growth and beyond what is likely attributable to increases in input costs or quality improvement.
The wide differential between commercial and Medicare prices has led some to claim that Medicare prices are too low, rather than commercial prices being too high. On one hand, Medicare payments are higher than public payments in several other countries and often sit at about the 20th percentile of commercial prices, suggesting Medicare prices are not completely out of line with the market. On the other hand, although evidence suggesting that lower public prices cause higher commercial ones is inconclusive at best, high commercial prices may improve access to care and quality for individuals insured through public programs. If commercial hospital prices were set at Medicare levels, hospital revenue would drop about 35 percent, undoubtedly causing some institutions to reduce quality-enhancing activities and other institutions to close. Moreover, price trajectories in Medicare are set to rise at a rate below inflation, potentially further reducing the adequacy of public fees over time. Regardless of whether one thinks Medicare price are too low or not, however, it is clear that the core problem in public programs in the US is not that prices are too high, and thus we focus on commercial prices.
The reason for high commercial prices is clear: It’s the market power, stupid. Well-functioning markets should not have the amount of price variation observed within and between commercial markets in the US, particularly when it comes to services with minimal quality differences. Studies of mergers, both within sectors (e.g. hospitals) and between sectors (e.g. hospitals acquiring physician practices) demonstrate the connection between market power and prices: consolidation leads to higher prices and little improvement in quality. Importantly, while market power is often treated as synonymous with market concentration, factors beyond concentration also generate market power. For example, providers in less concentrated markets may wield market power due to insensitivity of patients to the price of care and hesitancy of employers to steer patients to lower price/higher quality providers.
Widespread acknowledgement of market failures has created growing interest in examining how commercial health care prices are established in the US. Much energy has been devoted to measuring prices, assessing the impact of consolidation, and proposing policy solutions, including price transparency, increased antitrust enforcement, and price caps. These issues continue to be important, but there are several under-explored burning questions in the price debate that we think deserve attention. Here are our top seven:
1. Do Poorly Set Public Prices Distort Commercial Prices?
Our current systems for setting prices in public programs are flawed. For example, Medicare pays different amounts for the same service delivered in different settings. Providers are often paid more if they prescribe more expensive drugs. Relative value units for physician services are often inaccurate. If prices are set too high, or not adjusted when they should be, incentives are created to overuse care. Underpricing, though less common, discourages service use. While these issues clearly affect Medicare, understanding the extent to which they spillover to commercial prices is important. Although there is some evidence that higher Medicare prices lead to higher commercial prices (the opposite of cost shifting), more evidence on the extent to which relative prices in the commercial sector follow relative prices in Medicare is needed.
2. How Should Services Be Defined?
Prices are integrally related to how we define services. Many of our payment systems rely on very granular service definitions. For example, there are now ten Current Procedural Terminology (CPT) codes for office visits, varying based on visit intensity and whether a patient is new or established. Detailed service definitions can help ensure minimal variation in delivery costs within any given service category. However, having many codes creates opportunities for providers to choose more lucrative codes. The sheer number of codes and continual updating of code definitions generates administrative costs (including training and program integrity activities as well as the need to employ specialized coding staff) to ensure codes are used appropriately. This is all exacerbated by the fact that not all payers use the same coding systems.
Broader service categories may be desirable, including those used in payment methodologies such as partial capitation. However, broader payment systems may also encourage stinting and selection of patients with fewer health care needs, though evidence of these problems is scant. Our sense is that our system has erred on the side of too many service codes and too little standardization, but more attention to this point is needed.
3. How Does Quality Respond To Changes In Pricing?
The relationship between price and quality is central to the policy debate. What will we give up if we adopt policies that lower prices? Some evidence suggests that more expensive providers offer better care, indicating higher prices may mean higher quality. However, cross-national evidence suggests countries paying lower prices do not suffer significantly worse quality of care. These cross-sectional examinations do not imply causality; yet, studies of the relationship between mergers and prices suggest antitrust activities may lower prices but not degrade quality, supporting the position that policies intended to lower heath care prices do not necessarily impact quality adversely.
Addressing this issue is challenging for several reasons. First, the relationship between price and quality may not be linear, and reductions in the price of high-priced providers may have a different effect than reductions in the price of low-price providers. Second, quality is multidimensional and different people may weigh different dimensions differently, making broad conclusions elusive. Finding natural experiments that shed light on this issue is important, but so is conducting longer-run studies because quality impacts may play out over time. Investigating how infrastructure investments, innovation, and specialty choice respond to financial rewards may provide useful long-run insights about how price changes affect health.
4. How Should We Price New Digital Services?
Health care in the US is experiencing a digital revolution. A wide array of new digital tools and services and new ways to digitally communicate have been recently introduced, including portal messages between patients and their care team, artificial intelligence-enabled algorithms to support diagnosis and treatment, remote patient monitoring, web-based care support tools, and digital therapeutics. Questions of how these services should be paid for have not been resolved. Given the fee-for-service chassis of the US health system, the instinct is often to create codes for these services then assign prices, but that is problematic. For many interventions, there is limited evidence about their appropriate use. Low unit costs and limited barriers to access for digital and virtual services are appealing but raise the potential for overuse. Bundling digital tools into broader service packages might be valuable, but more work must be devoted to assessing how that might be accomplished.
Moreover, expansion of these services may have spillover effects on the availability of traditional services, requiring broader consideration of prices. Understanding the impacts these tools have on the broader system is a first-order challenge that would benefit from the best available evidence.
5. How Much Spending Is Flowing Outside Of The Claims System?
Most pricing research is based on claims data, a valuable but flawed resource. Increasingly, funds are flowing from payers to providers outside of the claims system via infrastructure or other fixed payments, quality bonuses, or shared savings from alternative payment models. These payments are often not a traditional ‘price’ as they do not change with service volume. But their use may lead to underestimates of what is being paid to providers in the commercial sector, mask market power by distorting observed revenues per service delivered, and may complicate enforcement of price regulation. Understanding how much revenue is flowing through non-claims channels, and how it is structured, may give a more complete picture of how markets are functioning and what is rewarded in health care.
6. Are Pay For Performance Systems Worth It?
Many payment systems (in commercial and public sectors) pay providers partly based on quality metrics. These systems strive to incentivize value, not volume, but are only justified if payments lead to better quality. There is a growing body of evidence suggesting this is not the case. Quality measures are often not closely enough tied to health outcomes to merit additional payments. In addition, operating these models is expensive and may distract from other activities. Understanding costs and benefits of pay for performance programs (whether stand-alone or part of an alternative payment model) is critical given the investment we have in them. Based on available evidence, we believe it would be reasonable to conclude that some of these systems should at least be scaled back, maybe even abandoned, until better, more targeted approaches to eliminating substandard care and improving quality can be designed.
7. To What Extent Do High Prices Reflect Higher Costs Of Production In The US, And Why?
Prices reflect, in part, the costs of production. Production costs in the US may be higher than in other countries for several reasons. First, health care prices likely reflect high labor costs in the US. While these high labor costs in health care may reflect details of the American health care system, they may also reflect a broadly different structure of the labor market. Reforming only the health care sector may not alter wage profiles more broadly, and efforts to lower incomes in the health care sector may, over time, create an imbalance between health care and other sectors. Relatively little is known about these dynamics.
Second, while the US uses more of some technologies and less of others compared to other countries, in general, prices of technologies (including drugs and devices) are higher in the US compared to other nations. These technology prices may be associated with higher quality (or innovation), but more work is needed in that area.
Additionally, the complexity and fragmentation of the American health care system create higher administrative costs, driving higher prices. While we know market power is an important determinant of higher prices in the US, further understanding of production costs of health care services in the US, and how and why these differ from other systems, would be valuable.
Conclusion
With health spending high and rising in the US, attention to health care prices – particularly in the commercial sector – will only continue to grow. Policy solutions will involve tradeoffs between spending, access, and quality, so understanding the impacts of various policy options on these tradeoffs will be central to better decision-making.