UNITED STATES OF AMERICA
FEDERAL TRADE COMMISSION
WASHINGTON, D.C. 20580
Bureau of Competition
STATEMENT TO THE NEGOTIATED RULEMAKING COMMITTEE ON THE SHARED RISK EXCEPTION TO THE ANTI-KICKBACK STATUTE
ROBERT F. LEIBENLUFT
Bureau of Competition
Holiday Inn Capitol
550 C Street, S.W.
Washington, D. C. 20024
JUNE 18, 1997
Good morning. It is a pleasure to be here today to discuss how the federal antitrust enforcement agencies -- the Federal Trade Commission and the United States Department of Justice -- understand and apply the concepts of risk-sharing and substantial financial risk in our analysis of provider networks.
As I understand it, the task of this committee is to flesh out the types of arrangements to which will apply the statutory exception to the anti-kickback statute for remuneration paid for items or services provided pursuant to certain managed care and risk-sharing arrangements.(1) My primary goal today is to explain the context in which the antitrust agencies consider risk-sharing, and what we understand the term to encompass. You, of course, are dealing with a different statute and will consider risk-sharing in a very different context. Accordingly, your conception of terms likely will vary significantly from ours. Nonetheless, we share with you a number of common goals as we interpret our quite different statutory mandates: avoiding the imposition of conflicting demands on health care practitioners; minimizing confusion about terminology used to describe risk-sharing arrangements; and avoiding restrictions that may preclude market arrangements that could be beneficial to consumers.
My comments relate only to the second part of the statutory exemption, that applies to items or goods provided pursuant to a written agreement that "through a risk-sharing arrangement, places the individual or entity at substantial risk for the cost or utilization of the items or services . . . which the individual or entity is obligated to provide."(2) Further, my comments describe the concepts of risk-sharing and substantial financial risk only as those terms are used in the 1996 Statements of Antitrust Enforcement Policy in Health Care issued by the Federal Trade Commission and the Department of Justice.(3) Copies of that document have been distributed to you. Before I begin, I must tell you that my remarks represent only my own views, and do not necessarily reflect those of the Federal Trade Commission or any individual Commissioner, or the Department of Justice.
The antitrust enforcement agencies typically deal with the concept of substantial financial risk-sharing in the context of evaluating agreements among competing physicians or other types of health care providers(4) concerning negotiation of contracts with insurers, HMOs, or other third-party payers that determine the prices the providers will be paid for services they furnish as members of a network. It is important to understand at the outset, therefore, that our analysis of risk-sharing focuses on agreements under which groups of health care providers share financial risk, not situations where individual providers independently assume risk for their provision of health care services. Our concern is with the competitive effect of joint activity among providers, not directly with the impact of risk assumption or other activities on total health care costs or service utilization.
To help you understand how these concepts fit into the antitrust analysis, I first will provide a very brief overview of the antitrust statutes and the conceptual framework of antitrust analysis.
The goal of antitrust law is to ensure that markets operate competitively, so that consumers may choose from a variety of offerings in accordance with their individual preferences. Neither the Federal Trade Commission as an institution, nor the antitrust laws as a discipline, favor managed care, fee-for-service, or any other particular form of health care delivery. Nor is antitrust designed to protect any group or type of competitor. Rather, our aim is to protect the competitive process so that consumers are able to choose among available offerings based on their needs and preferences. In health care, employers and other third-party payers generally contract with health care providers on behalf of groups of consumers, and we place considerable weight on payers' experiences and views in our efforts to assess the competitive impact of particular conduct.
Section 1 of the Sherman Act is most directly relevant to provider networks. It prohibits "contracts, combinations, and conspiracies" that restrain trade.(5) One crucial element of a violation of Section 1 is an agreement between two or more separate economic entities. For this reason, only collective conduct by one or more independent entities, such as is present when physicians join together to operate through a network, can violate Section 1. Where this collective conduct is absent -- when, for example, members of a single integrated group practice contract jointly -- Section 1 issues usually are not present. However, agreement among competitors need not be express or even formal to violate Section 1. An agreement can be implied -- based on a wink and a nod -- and can be inferred from circumstantial evidence showing a "conscious commitment to a common scheme" and that tends to "exclude the possibility" of independent action.(6)
Because most contracts "restrain trade" in some sense, Section 1 has been interpreted to prohibit only agreements that "unreasonably" restrain trade or competition. Many types of agreements are analyzed under what is called the "rule of reason," where the procompetitive aspects of the conduct in question are weighed against its anticompetitive effects. Normally, a case brought under the rule of reason requires a fact-intensive inquiry into whether the parties have "market power," which generally is understood to mean the power to raise prices, limit output, or restrict entry into a particular defined market.
However, the antitrust laws have long recognized that certain types of conduct, including "naked" price-fixing, some group boycotts, and market allocations among competitors, are so unlikely to have procompetitive effects that they are deemed to be inherently anticompetitive, and are presumed to be illegal. In cases involving so-called per se illegal conduct, no analysis of the actual competitive effects of the specific conduct is needed, and the government is not required to demonstrate that the parties to the agreement have market power. Conduct that would be per se illegal standing alone, however, will be evaluated under the rule of reason if it is "ancillary" to an otherwise permissible joint venture or collaborative undertaking. In determining whether an arrangement is ancillary, we look at whether the conduct -- for example, price fixing -- accompanies and is reasonably related to a significant integration of productive assets that has the potential to produce efficiencies.
For purposes of illustration, let me mention three hypothetical situations. First, assume that a group of physicians or hospitals decides to negotiate jointly with managed care plans in order to create a "united front" in dealing with managed care. All the providers do is agree on the fees they will charge, on the payer utilization review terms they will accept, and on which providers will serve patients from certain insurers. This agreement, because it has no potential to create procompetitive effects, is per se illegal under the Sherman Act, and carries a risk of criminal prosecution. The law conclusively presumes that the agreement will reduce competition, even if the parties are mistaken in their belief that they can suppress discounting, and even if they are totally unable to prevent entry of managed care into the market as they intend. The conduct cannot be justified by arguments that it is necessary to assure patient quality, to "level the playing field," or to prevent providers from being driven out of business.
On the other hand, assume that two hospitals create a joint venture to operate a new PET scanner, because neither hospital alone has the volume of patients necessary to support the equipment. Each facility contributes capital to purchase the machine and shares in profit or losses generated by its operation. Both hospitals participate in management, including setting the fee for use of the machine. In a sense, the agreement "fixes" the price of the PET scan technical fee. Nonetheless, this type of agreement generally will be evaluated under the rule of reason, rather than being deemed illegal per se, because it is "ancillary" to the joint venture: a price must be established for the services provided.
Third, assume that a group of doctors, or doctors and hospitals, jointly establish, fund and operate an HMO. Through the HMO, the providers agree on premium prices and on how participating providers will be paid for specific services provided to enrollees. While the providers have agreed on the fees they will receive through the HMO, the agreement will be judged under the rule of reason.
What is the fundamental difference between these examples? In the second and third examples, unlike with the first, the parties have integrated productive assets in a manner likely to benefit consumers, and the price agreement is ancillary to the legitimate goals of the venture. These actions create a new choice for consumers, and thereby may promote competition, and thus are not "naked" restraints of trade. They must be examined under the rule of reason to see what their likely competitive effects will be on the market, given the particular facts and circumstances involved. In the first example, in contrast, prices are likely to be higher than they otherwise would, consumer choice has been restricted, and the antitrust laws presume there is little likelihood that the conduct will offer any benefits for consumers; thus the arrangement can be condemned summarily without a detailed examination of market conditions.
The Antitrust Enforcement Policy Statements are intended to give guidance to the public and the provider community about how we analyze certain kinds of collaborative activity among health care providers, and to offer specific hypothetical examples showing how we apply that analysis in particular cases. One aspect of this guidance involves identifying which types of provider arrangements should be evaluated under the rule of reason, as opposed to being treated as per se illegal "naked" restraints. It is here that the agencies have considered the role of substantial financial risk-sharing.(7)
SUBSTANTIAL FINANCIAL RISK-SHARING
One factor we look at in determining whether a price agreement in the context of a provider network should be evaluated under the rule of reason is whether competing providers are sharing with one another substantial financial risk for the services provided through the network.(8) By that we mean, are there in place mechanisms that make the network providers as a group accountable for the total cost of defined services delivered to a group of covered individuals, so that the providers have incentives to cooperate in controlling costs and improving quality by managing the provision of services? This type of risk-sharing is contrasted to traditional fee-for-service payment, where each individual provider stands to maximize his or her revenues by increasing the number of services provided to patients. It also is contrasted to situations where providers individually assume risk; assumption of individual risk by members of a network, regardless of its effectiveness in reducing health care costs, is not the type of shared risk that justifies, under the antitrust laws, agreement among members of the network on the prices they will receive for services rendered through the organization.
The risk-sharing standard found in the Policy Statements is founded in large part on Supreme Court precedent. In Arizona v. Maricopa County Medical Society,(9) the Court found that agreement among independent doctors on the prices they would accept from insurance companies as payment in full was per se illegal price-fixing. The Court noted that persons who "pool their capital and share the risks of loss as well as the opportunities for profit" may be permitted to price their services jointly.(10) In that case, however, the agreement was among competing doctors on the prices each would accept for his or her own services, and there was no sharing of risk.
The current Antitrust Policy Statements discuss four general types of arrangements currently used by networks that can entail the sharing of substantial financial risk. However, the Statements explicitly recognize that other types of risk-sharing mechanisms may develop, and the federal antitrust enforcement agencies will consider other types of arrangements through which network participants may share substantial financial risk. Both the Federal Trade Commission and the Department of Justice have advisory opinion programs that permit parties to request the agencies' opinion on whether a proposed arrangement would be considered one that involves sharing of substantial financial risk.
The kinds of risk-sharing arrangements discussed in the Policy Statements normally are a clear indicator that the network using them involves significant cooperation and integration among its participants, and that the joining together of the physicians in the network has the potential to create efficiencies in the delivery of care through the network that may benefit consumers. Risk-sharing in this sense provides direct incentives for the participating providers to cooperate in controlling costs and improving quality by managing the provision of services by network members, and thus to achieve the overall efficiency goals of the venture. Moreover, the setting of price on a collective basis is integral to the use of these kinds of risk-sharing arrangements. Thus, the price agreement is considered ancillary to the overall agreement to operate the venture, and is evaluated under the rule of reason rather than the per se rule that applies to "naked" or free standing price-fixing agreements.
The first two types of risk-sharing discussed in the Policy Statements are arrangements under which the network agrees to provide a defined set of services to covered persons in exchange for a capitated rate, or a percentage of health plan premium or revenue. Both of these arrangements require the venture to provide covered services within a predetermined budget. It should be noted that the Statements contemplate arrangements where the network as a whole is capitated, not simply agreements among physicians who each will be individually capitated by a health plan. In the latter case, the physicians are not sharing risk, and thus are not directly motivated to work together collectively to provide care more efficiently.
A third type of risk-sharing arrangement discussed in the Policy Statements involves "use by the venture of significant financial incentives for its physician participants, as a group, to achieve specified cost-containment goals."(11) These arrangements usually involve an agreement that participating providers will be paid by the third-party payer on a fee-for-service basis. Two common ways of implementing this kind of incentive are the risk-withhold, and the bonus or penalty based on the network's success in meeting predetermined cost or utilization targets. The risk-withhold usually is structured so that a substantial part of the compensation due to all physician participants is withheld by the network or by the payer, with distribution of all or a portion of the withhold occurring only to the extent that the group as a whole meets predetermined cost-containment goals. In other words, a portion or all of the withhold would be forfeited to cover the amount by which actual costs exceeded the target, with any remaining funds eligible for distribution to the participating providers. We have not prescribed precisely what must be done with the withheld funds, though the normal practice is for the funds to revert to the payer. It should be noted that it is not sufficient for the network itself to withhold the funds and keep any amounts not distributed for its own use, either to cover operating expenses or to create reserves against future losses. In these cases, the physician members of the network indirectly would benefit from the withheld funds.
A mechanism similar to the risk-withhold involves negotiation with the payer of overall cost or utilization targets, with the group paying a penalty or receiving a bonus to the extent that it meets the target or fails to do so. This mechanisms differs from a risk-withhold in that the money is payable at the end of some accounting period, rather than being withheld in advance. Such systems may be easier to administer than a withhold, but can create similar incentives for the group as a whole to regulate the conduct of its members in order to enhance the providers' collective performance.
The fourth category of risk-sharing explicitly discussed in the Policy Statements includes certain kinds of global fees or all-inclusive case rates. Included are agreements by a network:
to provide a complex or extended course of treatment that requires the substantial coordination of care by physicians in different specialties offering a complementary mix of services, for a fixed, predetermined payment, where the costs of that course of treatment can vary greatly due to the individual patient's condition, the choice, complexity, or length of treatment, or other factors.(12)
As is the case with individual capitation arrangements, agreement by a group of competing physicians on a case rate for a bundle of services performed in substantial part only by each individual physician -- for example, agreements by obstetricians to provide prenatal care and delivery -- does not involve shared substantial risk. Instead, this provision contemplates situations where the venture has significant incentives for its members to cooperate in order to use the most cost-effective mix of resources in each case, and where the venture spreads among its members the risk of loss and possibility of gain on any particular case. In such cases, the arrangement has many of the features of a capitation arrangement, except that the provider is not at risk for the patient's health care status because payment is received for each patient requiring treatment of certain types of conditions, rather than for each patient enrolled in a health plan.
If the payment arrangement negotiated between the network and the payer subjects group members collectively to penalties or rewards based on the performance of the group as a whole, we usually are not concerned about how revenues are divided among members of the group. For example, networks that are paid on a capitation basis often use some type of fee-for-service mechanism to pay their members for services rendered to individual patients. As a general rule, we do not inquire into the nature of those payment arrangements, so long as the group as a whole has agreed to the fixed budget inherent in the capitation arrangement, and therefore has an incentive to regulate the conduct of its individual members.
We have not attempted to define on a categorical or quantitative basis what level of risk is considered "substantial." It depends on the context: in general terms, we regard as substantial a level of risk that is sufficient to provide real incentives for members of the network to adjust their behavior to achieve the collective goals of the network, and to work cooperatively with other network members to accomplish the same goals. In evaluating these issues, we consider standard industry practices as a rough guide, at least, to some basic parameters of what is considered commercially reasonable. For example, most risk-withhold proposals we have seen involve withholds in the 15 - 20% range.
Let me emphasize, however, that in analyzing particular risk-sharing arrangements, we look at the substance of an arrangement and the incentives it creates, not simply at its form. For example, in extreme cases, we might find that a risk-withhold or bonus system did not in fact (or was not likely to) create incentives to achieve the stated goals.(13) For example, a 20% risk withhold from a fee schedule set at 120% of the physicians' current charges would not likely create much incentive for the participants to modify their behavior to permit the group to meet the stated goals. It also is possible, in a bonus or penalty arrangement, for the targets to be set at a level that neither party thinks likely will be met, thus effectively negating the impact of the arrangement. For example, if the network were to receive a bonus if costs for an employee group over a year were 30% lower than the prior year's experience, and to pay a penalty if costs exceeded the prior year's experience by 30% (with neither rewards nor penalties accruing for experience between those ranges and with virtually no possibility that either threshold would be met), the arrangement likely would have no effect on the performance of the group.
In all the arrangements that we recognize as involving sharing of substantial financial risk, there is shared risk of loss or possibility of gain for members of the group based on the performance of the group as a whole. Many networks use evaluations of individual providers' performance on certain quality and efficiency measures in allocating funds internally, in order to provide incentives for those providers to conform their own practices to the group's goals. These mechanisms likely are useful to a network in managing risk that is assumed under these types of arrangements, but if used in isolation they do not establish shared financial risk by members of the group.
The newly mandated anti-kickback exception mentions four factors that should be taken into account in determining what constitutes "substantial financial risk":
the level of risk appropriate to the size and type of arrangement;
the frequency of assessment and distribution of the incentives;
the level of capital contribution; and
the extent to which the risk-sharing arrangement provides incentives to control the cost and quality of health care services.
As you can see, in our analysis the last factor subsumes the others; the first three factors we view as indicia of the extent to which the arrangement, considered as a whole, provides incentives for individual providers to meet the quality and efficiency standards of the group. We do not consider capital contribution, by itself, as establishing the sharing of substantial financial risk by the members of a network. Any network -- indeed, any price-fixing cartel that administers the prices charged by its members -- will have organizational and operating costs. Thus, mere contribution of money by providers, either through equity investment or periodic dues or assessments, does not serve to distinguish arrangements that have real potential to produce efficiencies from those that are merely likely to raise price and limit choice. Most physician-owned networks, by their nature, are designed to benefit their participants primarily by increasing their sales of the professional services provided through the network, rather than through a return on investment in the network entity. This does not make those networks suspect; it does, however, mean that the existence of an investment interest in the network, by itself, does not offer adequate assurance that the network is designed to provide real incentives for network members to work together to improve quality and efficiency.
However, capital investment can be relevant to the potential of a network to produce efficiencies. We do not inquire into the existence or nature of capital investment in cases where risk-sharing as defined in the Policy Statements exists; however, the nature and extent of investment of capital (both monetary and human) in the infrastructure necessary for a network to manage risk successfully certainly is relevant to whether the network likely will be able to produce high quality care in a cost-effective manner. Thus we consider investment in infrastructure as one factor, among a number, in our assessment of provider networks that do not involve substantial financial risk-sharing, but involve other types of integration among their provider members that warrant analysis of the ventures under the rule of reason.
The purposes of the antitrust enforcement agencies' consideration of risk-sharing by providers differs markedly from those of this committee, and consequently, it is quite possible that the understanding of risk-sharing that emerges from this process will differ from that reflected in the Antitrust Policy Statements. Differences in the analysis of risk-sharing for the purposes of different statutory frameworks are neither inappropriate nor inherently troublesome. However, there are three important considerations that I urge the committee to bear in mind.
First, we need to work together to assure that providers do not face conflicting demands between what is necessary to qualify for the risk-sharing safe harbor and the requirements of the antitrust laws. The antitrust laws and the anti-kickback statute have very different purposes; but none of us wants to put providers in a situation where legitimate ventures cannot comply with both.
Second, it is important to minimize confusion in the provider community arising from differing understandings of similar or identical terminology. While the purposes and contexts of our consideration of risk-sharing are very different from yours, there is a clear potential for misunderstanding. My remarks today are an initial effort to prevent unnecessary confusion, and we can do more.
Finally, this committee will need to be alert to the possibility that its work will have unintended negative effects on the future development of market structures that could be beneficial to consumers. One of our major concerns in developing and refining the Policy Statements was to craft them in a way that would permit continued evolution in health care markets, and would not stifle innovation that could benefit consumers. We were very sensitive to concerns that prior statements of our enforcement intentions, particularly with respect to provider networks, may unintentionally have limited market developments. We deliberately opted to offer guidance in the form of general principles, with more specific illustrations of the applications of those principles to particular facts, so that we would not drive ventures exclusively into predetermined molds that would leave little room for market change.
I recognize that this committee will be expected to develop more specific guidance, and your job will not be easy. Nonetheless, weighing the conflicting demands of certainty and flexibility also will be a major concern for you. We all share a common interest in assuring that our actions neither impose unnecessary burdens on providers nor push them into business structures that are inefficient or otherwise harmful to consumers.
My staff and I are prepared to work with you in any way to address these concerns or otherwise further our common interests.
I welcome the opportunity to respond to your questions or comments.
1. Section 216 of the Health Insurance Portability and Accountability Act of 1996, Public Law 104-191, directed the Secretary to establish standards for a statutory exception to Section 1128B(b) of the Social Security Act, 42 U.S.C. 1320a-7b(b), commonly known as the "anti-kickback statute." The exception applies to some types of remuneration paid by qualified HMOs and certain other organizations, and to remuneration paid pursuant to certain risk-sharing arrangements.
2. Section 216 of the Health Insurance Portability and Accountability Act of 1996, Public Law 104-191.
3. Statement Eight discusses physician networks, and Statement Nine discusses other types of provider networks.
4. For the sake of convenience, I will refer to physicians, hospitals, and other professional and institutional health care services providers as "health care providers" or simply "providers."
5. 15 U.S.C § 1.
6. Monsanto Co. v. Spray-Rite Service Corp., 465 U.S. 752, 764 (1984).
7. The Policy Statements also establish antitrust "safety zones" for certain conduct that is presumed not to be anticompetitive, and therefore does not need to be evaluated under the rule of reason. Sharing of substantial financial risk is one of the requirements for the antitrust safety zones for physician networks.
8. Networks that exhibit certain kinds of non-financial integration among participating providers also are entitled to rule of reason treatment.
9. 457 U.S. 332 (1982).
10. 457 U.S. at 356.
11. Policy Statements at 69.
13. See ACMG, Letter to Paul W. McVay, July 5, 1994 (FTC staff opinion letter) (staff could not determine whether a compensation arrangement involving a 15% withhold was sufficient, taken as a whole, to alter physician incentives).