The Proposed Accountable Care Organization Antitrust Guidance: A First Look
April 14th, 2011
 by
Joe Miller
|
Editor’s note: This post, by Joe Miller, is part of a series of Health Affairs Blog posts examining the proposed rules and guidelines implementing the Medicare Shared Savings Program, issued March 31 by the Centers for Medicare and Medicaid Services and other agencies. You can read other posts in the series by Mark McClellan and Elliott Fisher, Steven Lieberman, Douglas Hastings, and Ron Klar.
The Federal Trade Commission and Department of Justice (“the Agencies”) released a Proposed Statement of Antitrust Enforcement Policy Regarding Accountable Care Organizations Participating in the Medicare Shared Savings Program (“Statement”) at the same time the Notice of Proposed Rulemaking on Medicare Accountable Care Organizations (ACOs) was published by the Centers for Medicare and Medicaid Services (CMS). The Affordable Care Act provides no waiver authority to the HHS Secretary for the antitrust laws, and those who might consider forming ACOs were asking for regulatory clarity regarding their potential for antitrust risk.
Antitrust is a nuanced and complicated area of law, and CMS wisely delegated the antitrust function to the expert Agencies instead of doing the work itself or ignoring the subject completely. The resulting Statement sets up a useful framework for review that confirms that the ACO program will rely on competition to hold down costs and implicitly rejects the argument that market power and quality go hand-in-hand. However, as with all large programmatic undertakings, getting the objectives and framework right is only the first step. The Statement is essentially a screening mechanism to allow all parties to quickly discern whether a particular ACO requires closer antitrust scrutiny, and if the screens are set at a level that lets too much consolidation slip through, both consumers and the Shared Savings Program will be worse off.
The Statement attempts to balance the need for administrability, accuracy and speed for antitrust reviews, which is a challenging task in light of the inherent difficulty and complexity of the underlying legal and economic concepts. Does the Statement get the balance right? As I explain below, there is reason to worry that the Statement tilts too much toward allowing consolidation. Because of the vast size of health care markets, the anticompetitive effects of even a few ACOs with excessive market power could swamp any savings from the Shared Savings Program. Moreover, the screens laid out in the Statement do not address the potential market power of ACOs formed through mergers and acquisitions, the danger of anticompetitive effects from participation in the Shared Savings Program by providers who already have substantial market power, or the potential for cost shifting from Medicare to commercial markets. In addition, analyses that depend on Medicare data may not reveal the market power wielded by all providers that might participate in ACOs under the Shared Savings Program.
Summary of Proposed Statement of Antitrust Enforcement Policy Regarding Accountable Care Organizations Participating in the Medicare Shared Savings Program
The Agencies promulgated a proposed enforcement policy regarding the application of the antitrust laws to health care collaborations among otherwise independent providers and provider groups formed after March 23, 2010, the date the Affordable Care Act was signed into law. The Agencies are seeking public comments, which are due by May 31, 2011. The Statement recognizes that the perverse incentives of the Medicare fee-for-service payment system could be mitigated in part by providers forming ACOs under the Shared Savings Program, and further acknowledges that providers are more likely to coordinate patient care if they can use the ACO to serve commercially insured patients as well. At the same time, the Agencies recognize that ACOs have the potential to reduce competition and harm consumers through higher prices and lower quality of care, and the antitrust harm extends to both the Medicare and the commercial market.
The Statement asserts that ACOs that meet CMS’s eligibility requirements are reasonably likely to lead to provider collaborations intended to improve quality and reduce costs, and therefore will meet established clinical integration standards and avoid per se condemnation under the antitrust laws. Further, CMS will collect and evaluate cost, utilization and quality metrics over the three year agreement period, which will allow the Agencies to determine if in fact the CMS eligibility criteria lead to competitive benefits contemplated by the clinical integration metrics.
Accordingly, the Agencies will apply the rule of reason to an ACO if the ACO uses the same governance and leadership structure for the commercial and Medicare markets. The Agencies will undertake a streamlined analysis that calculates an ACO’s share of services in each ACO participant’s Primary Service Area, or PSA.(note 1) The higher the PSA share the greater the risk the ACO will be anticompetitive by reducing quality, innovation and choice for Medicare and commercial patients, in part by reducing the ability of competing ACOs to form. High PSA shares may also allow ACOs to exercise anticompetitive effects against commercial health plans. The Agencies are careful to explain that PSAs are not necessarily relevant antitrust geographic markets.
The Statement articulates three levels of antitrust scrutiny indicated by an ACO’s PSA share: ACOs with PSA shares of 30 percent or less fall into a “safety zone”; those with over 50 percent shares require formal antitrust review by the Agencies; and ACOs with PSA service shares of between 30 percent and 50 percent may voluntarily seek Agency review as they apply to become ACOs under the Shared Savings Program.
A. Safety Zone
If a proposed ACO’s PSA share is 30 percent or less it falls into an antitrust safety zone, meaning it is highly unlikely to raise significant antitrust concerns, and no initial competitive Agency review is necessary. The safety zone also includes a “Rural Exception” for providers that serve sparsely populated areas, and also has a “Dominant Provider Limitation,” meaning the ACO may include a provider in the group with over a 50 percent PSA share, if (1) no other ACO participant provides that service in the PSA, and (2) the provider participates in the ACO on a non-exclusive basis.
B. Mandatory Review
If the ACO’s PSA share exceeds 50 percent it will be subject to mandatory Agency review. The review will be based on several categories of information the ACO must provide to the Agency, including internal strategies and plans, documents that relate to competitors’ ability to compete with the ACO, and contact information for commercial health plans or other payers. The review is promised to be completed within 90 days, and will culminate in a letter from the Agency saying it either has no present intent to challenge or recommend challenging the ACO, or it is likely to challenge or recommend challenging the ACO if it proceeds. CMS will not approve the ACO for participation in the Shared Savings Program if it receives a challenge letter from the Agency.
C. PSA shares between 30 percent and 50 percent
If an ACO is below the 50 percent mandatory review PSA share, but above the 30 percent PSA safety zone, it may apply to become an ACO but will remain subject to investigation and potential challenge if it appears its formation or conduct may be anticompetitive. To provide additional guidance to potential ACOs in this middle category, the Statement lists categories of conduct to avoid to reduce the risk of an antitrust investigation. The categories to avoid include: preventing or discouraging commercial payers from steering patients to choose certain providers; tying sales of the ACO’s services to the commercial payer’s purchase of other services outside the ACO; contracting with other ACO providers on an exclusive basis; and restricting a commercial payer’s ability to make available the provider’s cost and quality information to the payer’s enrollees if similar information is used in the Shared Savings Program. If an ACO desires further certainty as to its antitrust risk it may request expedited antitrust review from the Agencies.
Initial Thoughts on the Statement
1. A Screening Mechanism
The Agencies have articulated when an ACO will be subject to an antitrust review; the Statement does not state the standards or norms of an antitrust analysis. It appears to be designed to encourage ACOs to get private antitrust advice as they form, and thereby deter plainly problematic collaborations.(note 2) The Statement itself makes the point well: “This Policy Statement provides guidance to allow ACOs to determine whether they are likely to present competitive concerns. It does not reflect the full analysis that the Agencies may use in evaluating ACOs . . .” (Statement, footnote 9)
Likewise, the Primary Service Area is not intended to identify an antitrust relevant geographic market. (Statement, footnote 22) Defining relevant geographic antitrust markets is a fact intensive inquiry requiring data and information that an ACO will find difficult to obtain in the public domain. Instead, the PSA test looks to patient draw statistics, a test that closely resembles the “Elzinga-Hogarty” test that has been repudiated by the FTC itself as a test for geographic markets, but may suffice as a “good enough” solution to identify those ACOs in need of further review. The PSA test favors administrability over precision; a potential ACO can determine, at relatively low cost and with tools and resources available to the general public, whether it is likely to get a closer look by the Agencies. The PSA test is not designed, however, to predict the outcome of the more intensive antitrust review. It is too simple and crude a tool for understanding the competitive dynamics of a provider market. The question for the Agencies is what, if anything is better suited to achieve the multiple goals of speedy review, ease of administration and protecting consumers from market power.
2. How Should the Screen Work?
Any screening mechanism should work to minimize the sum of false positives (in this case, scrutinizing ACOs that are competitively benign) and false negatives (not catching ACOs that deserve meaningful antitrust review), or what social scientists call Type I and Type II error. To set this screen appropriately however, the Agencies should recognize that the two types of error do not carry equal weight. A false positive imposes a minimal burden of a mild antitrust review(note 3), while a false negative imposes significant and lasting consumer harm by ignoring an antitrust violation, with its attendant price and quality effects. Accordingly, the screen should be set to err on the side reviewing more ACOs, particularly at the outset of the program. This approach is consistent with the policies of allowing competition to bring consumer benefits to health care markets.
To put some context and metrics around the benefit/burden calculus, consider that CMS predicts on page 349 of its proposed rule (page 19366 in the April 7 Federal Register) that the Shared Savings Program will generate about $510 million in savings over three years. Compare that number with the total annual operating revenues for hospitals in one medium sized city of over $7 billion. Multiply a small price increase on $7 billion by scores of jurisdictions and it’s easy to see the concern. If ACO policy encourages the formation of entities with market power, the resulting anticompetitive effects will likely far outweigh the financial benefit to the Medicare Trust Fund. Put another way, even the negative effects from a few ACOs with market power can call into question the premise of the Program.
Whether the Statement sets the screens at the right levels is an empirical question, but there is reason to question if they are too permissive. I would encourage the Agencies to articulate the rationale for the 30 percent safe harbor, and when the Shared Savings Program starts, for the FTC to use its subpoena power (note 4) to gather the necessary information to test that rationale and publish the results in a peer review economics journal. Until we have a better understanding for how the PSA test works in practice it is consistent with law and policy to err on the side of preventing consumer harm. Even shares of a single specialty group in the 20 percent to 30 percent market share range can create enough market power to warrant a closer look; a well-intentioned program does not justify ignoring the potential for harm in the name of expediency. This is the real danger of the Agencies’ approach; it has the right objective but the test may not be set accurately to achieve the right results.
Aside from setting the right PSA levels, the test itself contains one potentially problematic data bias. Due to the general difficulty in obtaining publically available commercial payment data, potential ACOs will likely rely on Medicare fee for service payment data. Those data, however, will not identify providers who choose not to accept Medicare patients because the reimbursement levels are too low. Similarly, those data may identify as appropriate for further review providers who have practices heavily weighted towards Medicare patients, even though—if all patients were considered—they would not cross the fifty percent threshold. This may be an unavoidable problem due to the lack of available data on the combination of private and public markets, but the Agencies should acknowledge the problem and attempt to correct for the biases.
3. Antitrust Problems are Best Avoided, not Fixed
Antitrust policy generally acknowledges that market power can be effectively deterred and avoided, but it is very difficult to remedy once entrenched. For this reason, mergers that violate the antitrust laws are blocked in court, and the Hart Scott Rodino Act sets up a screening mechanism similar to the Statement to identify potentially problematic deals before it’s too late. Transactions for which antitrust issues are capable of being remedied by selling off assets are allowed to proceed with such a “structural” remedy, but the Agencies generally disfavor remedying antitrust violations through after-the-fact restrictions on behavior. Regulating parties through the crude tool of antitrust is rarely effective, so behavioral restrictions are generally reserved for those violations where there is no other good option. The Agencies deserve recognition for drafting the Statement consistent with long-standing principles of prevention, and in applying those principles should maintain the traditional approach of prevention (either outright or through structural remedies) over regulation (through behavioral remedies)..
4. What the Statement does not Address
The Statement only applies to collaborations among “otherwise independent providers,” so by its terms offers no guidance to extant provider groups that may have market power but do not further consolidate the market to form an ACO. For example, the Massachusetts Attorney General issued a lengthy report identifying “market leverage” as a primary driver of high health care costs in that state, and further noting that price variations among providers are not explained by quality of care, illness burden of the patient population, complexity of the cases, or cost differences. A hospital in that market that wishes to contract with Medicare as an ACO is left without guidance by the Statement. The Agencies should consider clarifying the antitrust issues associated with allowing a provider with significant market power to participate in the Shared Savings Program. Moreover, CMS should carefully consider favoring ACO applications from provider groups without market power while it calibrates and refines the program so as to avoid potential consumer harm.
The Statement is also silent on how to address the issue of cost shifting between the Medicare program and the private sector. If an ACO aggregates market power, even though it is not subject to, or falls within the limits set forth in, the Statement, it has the potential to satisfy the metrics of the Shared Savings Program, reap the extra revenue from the Trust Fund, and simultaneously reduce quality in the private market. This is not the intent of the Program and should be addressed. CMS will gather cost, quality and utilization data to test whether the Program’s eligibility criteria in fact further Program goals; those data should be sufficient to identify and deter the potential for cost shifting. The Statement should clarify that the data gathering tools in the ACO Rule can and should be shared with the Agencies for antitrust enforcement purposes.
Finally, the Statement does not address the antitrust treatment of mergers and acquisitions, presumably because those transactions will be discovered by the normal Hart-Scott-Rodino reporting process and investigated under the Clayton Act. That assumption is fine as far as it goes, but many ACOs formed by merger or acquisition will be below the HSR reporting thresholds and slip through without review. The Agencies should make clear that the Statement applies to mergers and acquisitions that fall below the HSR reporting limits.
Conclusion
The Agencies have attempted to balance the need for administrability and quick review with traditional and time-tested antitrust processes. The system is based on the same deterrence paradigm that works for other industries in the normal course of business – almost all of the influence exerted by the antitrust laws is done through private counseling. The Agencies bring only a handful of cases each year, and private plaintiffs tend to bring cases on a per se theory of harm. Almost all antitrust “enforcement” is done privately between private business and their counsel. The Statement is a good start to the discussion, but the Agencies should carefully consider potential refinements, improvements or alternatives that hold the potential to improve the process and its outcomes for consumers, rather than the pecuniary interests of various interested parties.
Notes
Footnote 1.The PSA calculations are explained in an appendix to the Statement. The PSA is the lowest number of contiguous postal zip codes from which the health care provider draws at least 75 percent of its patients for service. The PSA shares are calculated by a physician’s primary specialty using Medicare Specialty Codes, or if for hospitals, Medicare MDCs, and for outpatient facilities, by outpatient category. The shares may be calculated using an all claims database if available or Medicare Fee for Service data.
Footnote 2. Antitrust practitioners will find the screening mechanism to be familiar; the Hart-Scott-Rodino (“HSR”) Act requires all acquisitions of stock or assets above certain minimum thresholds to face antitrust scrutiny. The Agencies routinely evaluate transactions under the HSR Act and fewer than 5 percent of transactions typically require extensive investigation. The vast majority of business deals pose no competitive threats to consumers and the Agencies clear them quickly.
Footnote 3. A 90 day review is relatively light by antitrust standards. A full merger investigation typically takes 6 -12 months, requires the production of millions of documents, all relevant electronic information in the company, depositions of senior executives, and so on.
Footnote 4. The Antitrust Division can only issue civil investigative demands for law enforcement purposes but the FTC can compel information for the purposes of studying an industry.
Email This Post
Print This Post
