Authors

• Stuart M. Butler

Publication: The JAMA Forum

Image Source: Burazin

Consolidation in health care has been increasing for some time and is now quite extensive in many markets. Some of this comes from mergers and acquisitions, but an important part also comes from larger organizations gaining market share from smaller competitors. The degree of consolidation varies by market. In California, most observers believe that metropolitan areas in the northern part of the state have provider markets that are far more consolidated than those in the southern part of the state. Insurer markets tend to be statewide and are less consolidated than those in many other states. The research literature on hospital mergers is now substantial and shows that mergers lead to higher prices, although without any measured impact on quality.[1]

The trend is accelerating for reasons that are apparent. For providers, it is becoming an increasingly challenging environment to be a small hospital or medical practice. There is more pressure on payment rates. New contracting models, such as Accountable Care Organizations (ACOs), tend to require more scale. The system is going through a challenging transition to electronic medical records, which is expensive and requires specialized expertise to avoid pitfalls. Lifestyle choices by younger physicians lead them to pursue employment in large organizations rather than solo ownerships or partnerships in small practices.

The environment is also challenging for small insurers. Multi-state employers prefer to contract with insurers that can serve all of their employees throughout the country. Scale economies are important in building the analytic capabilities that hold so much promise for effectively managing care. Insurer scale is important to make it worthwhile for providers to contract with them under alternative payment models. The implication of these trends is an expectation of increasing consolidation. There is need for both public and private sector initiatives in addition to anti-trust enforcement to foster greater competition on price and quality.

How can competition be fostered? For the insurance market, public exchanges created under the Affordable Care Act (ACA) and private insurance exchanges that serve employers can foster competition among insurers in a number of ways. Exchanges reduce entry barriers by reducing the fixed costs of getting an insurer’s products in front of potential customers. Building a brand is less important when your products will be presented to consumers on an exchange along with information on the benefit design, the actuarial value and the provider network. Exchanges make it easier for consumers to make informed choices across plans. This, in turn, makes the insurance market more competitive. Among public exchanges, Covered California has stood out for making this segment of the insurance market more competitive and helping consumers make choices that are better informed.

The rest of my statement is devoted to fostering competition among providers. I believe that fostering competition among providers is a higher priority because the consequences of lack of competition are potentially larger. In addition, a significant regulatory tool, minimum medical loss ratios, part of the ACA, is now in place and can limit the degree to which purchasers pay too much for health insurance in markets with insufficient competition.

Fostering competition in provider markets involves two prongs—broadened anti-trust policy and other policies to foster market forces. Anti-trust policy, at least at the federal level, to date has not addressed hospital acquisitions of physician practices. These acquisitions lead to higher prices to physicians because hospitals can negotiate higher prices for their employed physicians than the physicians were getting in small practices. Although not yet extensive, a developing research literature is measuring the price impact.[2] Hospital employment of physicians can also be a barrier to physicians steering patients to high-value providers (another hospital or a freestanding provider). To the degree that it reduces the chance of larger physician groups or independent practice associations forming, hospital employment of physicians reduces potential competitors in contracting under alternative payment models.

Another area not addressed by anti-trust policy is cross-market mergers. The concern is that a “must have” hospital in a multi-market system could lead to higher rates for system hospitals elsewhere. Anti-trust enforcement agencies have tended to look at markets separately, so this issue tends not to enter their analyses.

Many have seen price and quality transparency as a tool to foster competition among providers. Clearly, transparency has become a societal value and people increasingly expect more information about organizations that are important to them in both the public and private sector. But transparency is often oversold as a strategy to foster competition in health care provider markets. For one thing, many benefit designs have few incentives to favor providers with lower prices. Copays are the same for all providers and with coinsurance, the insurer covers most of the price difference. Even high deductibles are limited in their incentives because almost all in-patient stays exceed large deductibles and out-of-pocket maximums also come into play for many who are hospitalized. Another issue is that the complexity of comparing prices is a “heavy lift” for many consumers. Insurers and employers now have excellent web tools designed to make it easier for patients to compare prices, but indications are that the tools do not get a lot of use.

Network strategies have the potential to be more effective. The concept behind them is that the insurer is acting as a purchasing agent for enrollees. To the extent that they have the potential to shift volume from high-priced providers to low-priced providers, money can be saved in three distinct ways. The first is the higher proportion of services coming from lower-priced providers. The second is the additional discounts from providers seeking to become part of the limited or preferred network. Finally, if a large enough proportion of patients are enrolled in plans with these incentives, providers will likely increase the priority given to cost containment. In creating networks, insurers are increasingly using broader and more sophisticated measures of price as well as some measures of quality. Cost per patient per year or cost for all services involved in an episode is likely to have more relevance than unit prices. Using such measures to judge providers for networks has strong analytic parallels to reformed payment approaches, such as ACOs and bundled payments for episodes of care. Network strategies also create more opportunities for integration of care. For example, a limited network or a preferred tier in a broader network could be mostly limited to providers affiliated with a large health care system. Indeed, some health systems are developing their own health plan or partnering with an insurer to offer plans that favor their own providers.

In this testimony, I discuss two distinct network strategies. One is the limited network, which includes fewer providers than has been the norm in private insurance. The other is the tiered network, where the network is broad but a subset of providers are included in a preferred tier. Patients pay less in cost sharing when they use the preferred providers. Limited networks are a more powerful tool to obtain lower prices because patient incentives are stronger. If patients opt for a provider not in the limited network, they are subject to higher cost sharing and might have to pay the provider the difference between the charge and what the plan allows. Results of these stronger incentives are seen in a number of studies by McKinsey and Co. that have shown that on the public exchanges, limited network plans have premiums about 15 percent lower than plans with broader networks.

Public and private exchanges are an ideal environment for limited network plans. The fixed contributions or subsidies to purchase coverage mean that consumers’ incentives to choose a plan with a lower premium are not diluted—they save the full difference in premium. Exchanges do not have the “one size fits all” requirement that constrains many employers in using this strategy. If an employer is offering only one or two plans, it is important that an overwhelming majority of employees find the network acceptable. But a limited network on an exchange could appeal to fewer than half of those purchasing on the exchange and still be very successful. In addition, tools provided by exchanges to support consumers facilitate comparisons of plans by having each plan’s network accessible on a single web site.

In contrast, tiered networks have the potential to appeal to a larger consumer audience. Rather than making annual choices of which providers can be accessed in network, tiered networks allow these decisions on a point-of-service basis. So the consumer always has the option to draw on the full network. Considering the greater popularity of PPOs than HMOs and the fact that tiered formularies for prescription drugs are far more popular than closed formularies, the potential market for tiered networks might be much larger. But this has not happened. In many markets, dominant providers have blocked the offering of tiered networks by refusal to contract with insurers that do not place them in the preferred tier. This phenomenon was seen in Massachusetts, where 2010 legislation prohibiting this practice led to rapid growth in insurance products with tiered networks.

Some Californians are familiar with a related approach of reference pricing due to the pioneering work that CalPERS has done in this area for state and local employees. Reference pricing is really an “extra strength” version of the tiered network approach. An insurer sets a reference price and patients using providers that charge more are responsible for the difference (although providers sometimes do not charge patients in such plans any more than the reference price). So the incentive to avoid providers whose price exceeds the reference price is quite strong. While CalPERS has had success with joint replacements and some other procedures, a key question is what proportion of medical spending might be suitable to this approach. For reference pricing to be suitable, the services must be “shoppable,” meaning that they must be discretionary with the patient and can be planned in advance. One analysis estimates that only one third of health spending is “shoppable.”[3]

While network approaches have a lot of potential for fostering competition in health care markets, including those that are consolidated, they face a number of challenges that must be addressed. First, transparency about networks must be improved. Consumers need accurate information on which providers are in a network when they choose plans and when they choose providers for care. Accommodation is needed for patients under treatment if their provider should drop out of a network or be dropped from one. Network adequacy regulations are needed to protect consumers from networks that lack access to some specialties or do not have providers close enough to their residence. They are also important to preclude strategies that create networks unlikely to be attractive to patients with expensive, chronic diseases. But if network adequacy regulation is too aggressive, it risks seriously undermining a very promising tool for cost saving. So regulators must very carefully balance consumer protection with cost containment.

Some consider the problem of “surprise” balance bills, charges by out-of-network providers that patients do not choose, to be more significant in limited networks. This may be the case, but the problem is substantial in broader networks as well, and its policy response should apply throughout private insurance.

Another approach to foster competition in provider markets involves steps to foster independent medical practices. Medicare has taken steps to ease requirements for medical practices to contract as ACOs. It recently took some steps to limit the circumstances in which hospital-employed physicians get higher Medicare rates than those in office-based practice. Private insurers have provided support to some practices to incorporate electronic medical records into their practices. To the degree that independent practice can be made more attractive relative to hospital employment, competition in provider markets is likely to increase.

Additional restrictions on anti-competitive behavior by providers can also foster competition. These behaviors include “all or nothing” contracting requirements in which a hospital system requires insurers to contract with all hospitals in the system and “most favored nation” clauses in which insurers get providers to agree not to establish lower rates for other insurers.

Although the focus of discussion about policy in this testimony has been about fostering competition, regulatory alternatives that substitute for competition should not be ignored. At this time, two states—Maryland and West Virginia—regulate hospital rates. Some states, mostly in the Northeast, have been looking at this approach. Although I respect what some states have accomplished with this approach in the past, I need to point out that the current environment poses additional challenges for rate setting. The notion that rates would be the same for all payers, a longstanding component in Maryland, is unlikely to be practical today because rate differences between private insurance, Medicare and Medicaid are so large. So differences would likely have to be “grandfathered.” More practical would be to limit regulation to commercial rates, as West Virginia has done since the 1980s.

Another challenge is that with broad enthusiasm about the prospects for reformed payment, those contemplating rate setting need to make sure that the mechanism encourages payment reform rather than blocks it. Maryland has been quite careful about this and its recent initiative to broaden its program seems promising. But with the recent emphasis on multi-provider approaches to payment, such as ACOs and bundled payment, the limitation of regulatory authority to hospital rates could be a problem.

So what are my bottom lines for legislative priorities? I have two. States should address restrictions on anti-competitive practices such as anti-tiering restrictions, all-or-none contracting restrictions, and most favored nation clauses. My second is to regulate network adequacy wisely. It is a potent tool for fostering competition, even in consolidated markets. Network strategies do have problems that need to be addressed, but it must be done while preserving much of the potency of the approach.

A concluding thought involves acknowledging that provider payment reform approaches are likely to contribute to consolidation. Small hospitals and medical practices are not well positioned to participate, although virtual approaches can often be used in place of mergers, for example as California’s independent practice associations have enabled many small practices to participate. But I see payment reform as having major potential over time to reduce costs and increase quality. So my advice is to proceed with payment reform but also take steps to foster competition. Rate setting is best seen as a “stick in the closet” to use if market approaches should fail to control costs.

Authors

• Paul Ginsburg

Authors

• Paul Ginsburg

Publication: JAMA