|
H O S P I T A L S & P H Y S I C I A N S :
G A I N S H A R I N G
5 December 2006
Gain Sharing:
A Good Concept Getting A Bad Name?
Allowing physicians to share in the savings they help produce
could lead to greater accountability in health care.
by Gail R. Wilensky, Nicholas Wolter, and Michelle M. Fischer
ABSTRACT:
The introduction of diagnosis-related groups (DRGs) created a clear misalignment between the incentives facing hospitals and those facing physicians. The interest in gain sharing that developed in the 1990s represented an attempt by physicians to extract and hospitals to offer some of the savings being produced by physicians. Advisory bulletins by the Office of Inspector General (Department of Health and Human Services) quickly put a stop to further interest in these strategies. Newer, narrowly defined types of gain sharing have been under consideration. More broadly defined strategies that will be tested under a new Centers for Medicare and Medicaid Services demonstration are more promising. [Health Affairs 26, no. 1 (2007): w58–w67 (published online 5 December 2006; 10.1377/ hlthaff.26.1.w58)]
Health care costs and the imperative to improve patient safety and quality represent two of the most important issues on the U.S. health care agenda. Health care spending, which continues to grow more than two percentage points faster each year than the rest of the economy, in real terms per capita, is creating an unsustainable strain on private and health care budgets and exacerbating already major access problems for millions of Americans.1 In spite of these high spending levels, the Institute of Medicine (IOM) and others continue to highlight the need to improve patient safety and quality performance.2
One opportunity involves the care of complex, high-cost patients. The fragmentation of the current health care delivery system, along with cultural and regulatory barriers, frequently inhibits physicians’ and hospitals’ ability to engage in collaborative approaches that both reduce cost and improve quality.
Some policymakers believe that insurance companies that form and manage physician and hospital networks can increase their focus on cost and quality improvement, perhaps through network competition. Some believe that forming accountable provider units consisting of physicians and hospitals will be a key ingredient to improving cost and quality.3 Large multispecialty group practices and integrated delivery systems (IDSs) can focus on care coordination and chronic disease management to produce measurable improvements in patient safety, quality, and cost—although not all do so equally well.
The current health care delivery system, however, is marked more by fragmentation than by organization. The majority of physicians are in small single-specialty groups, which makes it difficult to devise coordinated approaches to managing highly complex, high-cost patients. This is exacerbated by the lack of underlying information technology (IT) and by disparate systems, where they do exist, that do not share information across settings.
The challenge of forming organizational approaches to care management and cost control are magnified by a reimbursement system that reinforces silos of care rather than system approaches to tackling high-volume, complex, and high-cost areas of medicine. The problem is made worse by legal and regulatory barriers that essentially prohibit the use of financial incentives to motivate cooperation and coordination. Additionally, high-volume, more-costly care is rewarded in the physician reimbursement system, and there are no rewards for those who provide more-efficient but high-quality care. Indeed, our current reimbursement systems continue to pay for unnecessary hospital admissions or for physician work related to caring for problems resulting from medical errors and complications.
Our vision for gain sharing. Within this context, we advance the concept of rethinking gain sharing as a transitional strategy to allow increased focus on reducing costs and improving quality, while also encouraging the development of formal or even “virtual” group practices and, ultimately, more IDSs. Properly constructed, gain sharing could both provide short-term gains and incentives for new organizational structures, more capable of effectively managing both cost and quality.
We also raise the possibility that gain sharing might be a transitional alternative to the explosion of physician-owned entities and ancillaries, where some have raised concerns about the likelihood that the conflict of interest represented by physician self-referral will create more rapid increases in use and even more fragmentation of care. Whether or not gain sharing can be constructed as part of a serious alternative to physicians’ owning facilities is unclear.4
Shortcomings of recent gain-sharing efforts. If gain sharing were to have a true chance to play a role as a transitional strategy, we think that it is unlikely to be the kind of gain sharing that has occurred recently. These efforts have focused primarily on limited and narrow high-cost technology—not without value, but too limited to address broader systemic organizational redesign, which might produce more-sustainable approaches to cost reduction and improvements in patient safety and quality. Furthermore, the very short (usually one year) time limits on gain-sharing arrangements, as they have been allowed, would not support the longer-term goal of major physician-hospital realignment into accountable, integrated-care practices.
What Do We Mean By ‘Gain Sharing’?
When we refer to “gain sharing,” we mean the ability to share savings with physicians that result from the more appropriate use of imaging and testing services; or the careful and appropriate prescribing of therapeutics that use the least-costly appropriate therapy available; or activities around medication reconciliation, which reduce medication errors and subsequently more-costly care resulting from these errors. We also mean, for example, the ability to share the savings that come from using outpatient services rather than inpatient services where appropriate, or from providing disease management services that keep congestive heart failure (CHF) patients from having to be admitted to hospitals for acute episodes of illness if they can be safely maintained in an ambulatory setting or at home. System approaches to coordinating end-of-life care or the care of complex chronic illness, including diabetes mellitus and coronary artery disease, also offer opportunities to improve quality and safety while reducing costs over time.
Savings of the type described above might be more readily accomplished within IDS settings. However, most physicians practice in small single-specialty groups. Even multispecialty groups have difficulty sharing savings produced in settings outside of their direct purview, such as hospitals. Some of the current barriers are legal and regulatory, although cultural barriers between specialties and between physicians and hospitals play a role as well.
Although encouraging the formation of IDSs is desirable, strategies must be devised that encourage desired behavior in the world as it currently exists, perhaps incentivizing the development of physician groups or “virtual groups” and integrated health care networks over a longer period of time. “Virtual groups” are groups of physicians who are not formally connected but who choose to associate with each other informally to promote coordination and improve efficiency.5 We believe that gain sharing, used in a broader context, provides a strategy that might be very useful during a transitional period in U.S. medicine.
Gain Sharing And Current Law
Gain sharing, as the OIG has been using it, looks at narrowly defined savings generated by using one medical device over another, rather than in the broader sense of rewarding clinicians who create savings and improve quality by varying practice style and decision-making strategies. Narrowly defined gain sharing does not provide a mechanism for rewarding physicians who participate in broader system approaches to quality, patient safety, and cost management.
The concept of gain sharing took root during the time when the payment systems for hospitals and physicians became misaligned, beginning with the Social Security amendments of 1983 (P.L. 98-21), which established the statutory framework for the Medicare hospital prospective payment system (PPS) that pays hospitals based on diagnosis-related groups (DRGs).6 This type of payment system creates strong incentives for facilities to contain costs to be financially successful, within the flat amount received for each discharge. Conversely, Medicare generally pays physicians a separate fee for each service, which does not create any incentive for containing a physician’s outpatient costs or costs within the hospital.
In seeking ways to realign the incentives, and in response to physicians’ complaints that they were being asked to produce the savings but not being offered any of the gains, hospitals began to consider “gain sharing” as a strategy wherein the hospital would share with physicians some of the cost savings achieved through their involvement with programs designed to control hospital costs.
Antikickback statute. Although this approach appears to make sense from a financial standpoint and could lead to the practice of more cost-effective medicine, there are several restrictions on these types of arrangements, including the Social Security Act Civil Monetary Penalties Law (CMP), federal antikickback statutes, or the Stark laws dealing with self-referrals. Internal Revenue Service (IRS) regulations regarding private inurement and private benefit are also concerns.
The primary concern regarding the CMP is the potential negative impact of any gain-sharing arrangement on the quality of care provided in Medicare and Medicaid. Hospitals are prohibited from knowingly making a payment either directly or indirectly to a physician as an inducement to reduce or limit items or services furnished to Medicare or Medicaid patients under that physician’s direct care.7 Policymakers’ concern when constructing this broad prohibition was that hospitals would have an economic incentive to financially reward physicians to discharge patients too soon or otherwise compromise care. Therefore, any hospital gain sharing that could influence physicians to reduce or limit clinical services violates the CMP. This has led, in part, to the narrow scope of gain-sharing arrangements that have little or no potential to reduce services but that are also limited in their cost-saving potential. These include not opening supplies until needed or other similar activities; however, even they must be closely scrutinized to ensure that they do not cause any decrease in patient services.
The federal antikickback statute, in effect since 1972, is intended to protect federal health care programs and patients from fraud and abuse by limiting the negative influence of money on health care decisions. It prohibits payments in any form made purposefully to induce or reward the referral or generation of federal health care program business. Therefore, gain-sharing arrangements can violate this statute if the cost-saving payments influence referrals, although the OIG has published some safe harbors to protect certain financial relationships between referring parties. Safe harbors were introduced because of providers’ concerns regarding the broad nature of the antikickback law and the implications on business practices. Some safe harbors include specialty referral arrangements between providers, investments in group practices and ambulatory surgical centers (ASCs), and joint ventures in underserved areas.8
Stark laws. The Stark legislation prevents physicians from referring Medicare and Medicaid patients for designated health services to entities with which they (or an immediate family member) have a financial relationship. Physicians must evaluate any economic benefits they might receive from entities to which they are referring Medicare and Medicaid patients, to determine whether they meet any of the exceptions described in the statutes, which can be very complicated.
Early Attempts At Gain Sharing
CMS demonstration. Several dozen hospitals experimented with gain-sharing arrangements in the 1990s that proved effective at reducing operational spending within a particular hospital service line. Gain sharing seemed a particularly attractive concept for cardiology but was not limited to this area. A CMS demonstration project conducted between 1991 and 1996 was the most detailed of any experiments in demonstrating the effectiveness of gain sharing.9 Four selected hospitals agreed to accept a global rate that would cover Medicare Parts A and B services for each beneficiary undergoing coronary artery bypass graft (CABG) surgery. The participating hospitals and physicians could apportion the global fee as they desired. Two hospitals capitalized on the opportunity to use this global, or bundled, payment arrangement to align physicians’ incentives with those of the hospital by implementing gain-sharing programs, resulting in sizable reductions in costs in intensive care, laboratory, routine nursing, and pharmacy services costs. There also was a reduction in the duration of operating room procedures, intensive care unit (ICU) stays, and post-ICU stays, and patient outcomes improved overall. Participants concluded that “aligning surgeons’ goals with hospital incentives to reduce costs was absolutely critical in changing practice patterns and improving department efficiency.”10
OIG ban. As a result of what was regarded as a successful CMS (then HCFA) demonstration project, many hospitals began creating similar arrangements, leading to an influx of requests for OIG opinions on proposed gain sharing. The result was the July 1999 OIG Special Advisory Bulletin that effectively banned gain-sharing arrangements, which indicates that in almost any form, they violated the CMP law and that legislative change would be necessary before the OIG could approve any such arrangements.11 The OIG recognized in its July 1999 bulletin that appropriately structured gain-sharing arrangements “may offer significant benefits where there is no adverse impact on the quality of care received by patients” but also stated that such arrangements were a clear violation of Section 1128A(b)(1) of the Social Security Act and that legislative changes would be necessary to provide a reprieve from the CMP prohibition.12 However, in 2001 the OIG issued an advisory opinion allowing a gain-sharing arrangement of a limited nature. In 2005 it issued six advisory opinions allowing gain sharing in cardiology and cardiovascular services at four hospitals. These opinions stressed safeguards that the OIG felt would mitigate the risk of reducing needed patient care, selection of lower-acuity patients, or payment to physicians for referrals.13 Many experts feel that even with some definition around these and other safeguards, institutions wishing to proceed with gain sharing should seek their own OIG advisory opinion to lessen risk—a costly and time-consuming process.
MedPAC report. Another potential stimulus to reconsidering the role of gain sharing occurred when the Medicare Payment Advisory Commission (MedPAC) issued its March 2005 report on physician-owned specialty hospitals. In this report, MedPAC made a series of recommendations to Congress, including one stating that “Congress should grant the authority to allow gainsharing arrangements between physicians and hospitals and to regulate those arrangements to protect the quality of care and minimize financial incentives that could affect physician referrals.”14
MedPAC’s stated rationale for this recommendation was that “properly structured, gainsharing arrangements have the potential to encourage physician and hospital cooperation to lower costs and improve care, but there should be safeguards to ensure that cost saving measures do not reduce quality or inappropriately influence physician referrals. These arrangements could serve as an alternative to physician owned specialty hospitals.”15 However, legislative change allowing more widespread use of gain sharing was not included in the 2005 Deficit Reduction Act, although the conference agreement did allow for the establishment of gain-sharing demonstration pilot projects.
Are Past Concerns Still Relevant?
Concerns that have migrated into statute and regulation via CMPs, the federal antikickback law, and Stark regulations were based on worries about incentives leading to less-than-adequate or -appropriate evidence-based care, inducement by hospitals for increased physician referrals, and physician self-referral.
Previous work has shown that there is tremendous regional variation in the amount of care provided for given conditions and that there is either little correlation or even an inverse correlation between the amount of care and the resulting patient safety and quality measures in use.16 This leads us to argue that gain sharing should be benchmarked not only to the historical costs of care of a particular institution but also to evidence-based research and independent benchmarks, where available. Initially, this might be defined by combining historical costs with the below-average costs in the geographical area. Properly constructed and evaluated, gain-sharing arrangements might even be research laboratories for new benchmarks around best practices for cost and quality management.
Reinforcement of fragmented system. The Stark regulations are under the jurisdiction of the CMS and have not been addressed in OIG opinions. In our view, the regulations have become an inconsistent patchwork of prohibitions and safe harbors, preventing physician-ownership and self-referral in many areas but allowing rampant growth in others. Gain sharing, where true accountability for cost and quality measures is in place, might provide better incentives and outcomes than are now in place in some physician-ownership venues.
The federal antikickback provisions are particularly troubling. At a time when many would argue that tighter alignment between key physician specialties and hospitals could dramatically improve quality and possibly have a strong impact on current cost problems, providers find that they cannot use a valuable transition strategy that could create long-term realignment and new organizational approaches to care delivery. Creating incentives for the development of tighter, more organized, accountable physician-hospital units creates an important opportunity for improvement. Current incentives reinforce an individualistic, artisan, cottage industry delivery system, which perpetuates care fragmentation and marked regional variation in resource use and quality performance.
Need for legislative changes. Given the imperatives the United States faces regarding cost and quality issues, it seems time to reconsider the wisdom of prohibiting shared savings where physicians and hospitals are willing to coordinate on system approaches and other strategies focused on improvement, with the requirement that appropriate measurement systems be in place to ensure accountability around cost savings, quality, and patient safety improvement and with the understanding that such claims will be subject to audit.
Partial recognition of this quandary has led to some relaxation of the prohibition against any form of gain sharing outside of groups that are at financial risk, which have been previously exempted from various prohibitions. Unfortunately, under current arrangements, the only type of gain sharing that has been allowed is a sharing of savings generated from using certain medical devices and supplies. Even this limited gain sharing is closely scrutinized and approved by the OIG case by case, and approval can take years. Although the empirical evidence regarding the effects of gain sharing in general is not robust, the results of the CMS CABG demo indicate cost reductions associated with no reduction in quality and some improvement in mortality and patient satisfaction.17
When the OIG has evaluated particular gain-sharing arrangements, it has generally focused on three areas: accountability, quality controls, and safeguards against payments for referrals.18 Accountability is defined as an arrangement wherein actions will result in cost savings that are clearly and separately identified. If we want to reward quality and efficiency, then measures need to be in place that define how these efforts will be rewarded, and physicians will need to be held accountable for demonstrating the savings generated. We contend that accountability for improved patient safety and quality should have the same rigor.
Few would question the importance of quality controls, given the potential adverse impact on patients if only cost savings were tracked and rewarded. Gain-sharing safeguards primarily focus only on maintaining quality, when so much research points to the urgent need to improve it.
The third category is to ensure that gain-sharing payments are not being used to reward or induce patient referrals that would violate the antikickback statute. Although it is understandable that the OIG would need to apply this criterion unless and until the law is changed, this part of the safeguard needs to be changed if a wider range of gain-sharing arrangements are to be allowed and if we wish to create incentives for the development of more coordinated, even integrated, accountable provider units of care.
Without legislative change. In the absence of legislative change that would allow for the wider use of these arrangements, the only avenue available is to use the new federal demonstrations outlined in the Deficit Reduction Act of 2005.19 Section 5007 of the Conference Agreement allows for the establishment of a program to “test and evaluate methodologies and arrangements between hospitals and physicians designed to govern the utilization of inpatient hospital resources and physician work to improve quality and efficiency of care provided to Medicare beneficiaries.” Thus, it represents an explicit opportunity to test the effects of better aligning incentives between hospitals and physicians.
The three-year demonstration will involve six sites, two of which must be located in rural areas, beginning 1 January 2007 and running through 31 December 2009. The projects must meet certain requirements laid out to maintain or improve quality while achieving cost savings and are designed to improve financial and operating performance by sharing some hospital cost savings with physicians. Restrictions on incentive payments under current law are waived for this demonstration so that their effectiveness can be evaluated.
Under the demonstration, hospitals will receive their regular Medicare reimbursement for patient care, but unlike under current law, they can pay physicians up to 25 percent of the documented cost savings generated from quality improvements. Payments must be capable of being audited, must be uniform across physicians, and cannot be based on volume or referrals.20 For the CMS to be able to assess the effects of gain sharing, it is requiring a large sample of participants, with preference given to consortia of health care groups with affiliated hospitals and also requiring some follow-up on patients, at least beyond the hospital episode.
Federally sponsored demonstrations represent a useful first step as long as positive results lead to legislation that allows the action being demonstrated to be applied in a more general way across the delivery system. Unfortunately, the history of even successful demonstrations’ becoming law is not promising. The special Medicaid program in the Arizona Health Care Cost Containment System (AHCCCS) is a well-known example of a program that took more than a dozen years to go from research and development status to an allowable program under the law, even though it was widely regarded as having provided good results for both the recipients of the program and the state of Arizona. Many others never resulted in allowable legislation.
Next Steps
The focus of efforts to moderate spending while improving outcomes, which we believe gain sharing supports, needs a much broader focus than choosing one device over another. Creating gain-sharing arrangements that align the incentives of hospitals and physicians and improve quality of care will require careful planning and collaboration between these two groups, which has frequently not occurred to date. Informal comments by physicians who were participating in the CMS CABG demonstration during the 1990s suggest that the use of single bundled payments, which allowed savings to be shared in ways determined by the individual group, encouraged greater collaboration among the participating physicians. It will be important to assess whether similar reports of increased collaboration occur in the gain-sharing demonstrations scheduled to start in 2007.
Many groups have been working on defining measures of quality, and some health plans are already starting to use some of the measures, either in identifying good performers or in paying for results. Although much of this activity is focused on developing or actually testing pay-for-performance (P4P) measures, many of the same issues are relevant for gain sharing.
Payers and employers have been reporting clinical process measures from the Health Plan Employer Data and Information Set (HEDIS) for more than ten years, and some payers have begun using the data in their payments, including the Integrated Healthcare Association (IHA).21 The CMS has been working with the Ambulatory Care Quality Alliance (AQA) and the Hospital Quality Alliance to develop common sets of quality measures for both outpatient and inpatient care. A detailed discussion of the current state of performance measures, the desired set of performance measures, and a roadmap on how to get from where we are to where we want to be is available in the first of the volumes in Clinical Pathways to Quality Health Care recently published by the IOM.22
More effort and resulting agreement seems to have been made in developing measures of quality than has occurred in developing measures of efficiency. Although the gain-sharing demonstrations might accelerate work in this area, one possibility is to use an efficiency type of screen. An example would be to require gain-sharing partners to be below the median charge in their geographical area, although it might be more desirable to use a combination of improvement in cost combined with an absolute threshold measurement.
Some measures are also available reflecting patient satisfaction, an additional measure that can be included reflecting the impact of gain sharing, as was the case with the CABG demos in the 1990s. These measures primarily come from the Consumer Assessment of Healthcare Providers and Systems (CAHPS).
Gain sharing, of the type we have described, is envisioned as an important transitional strategy, leading U.S. health care to higher levels of accountability and improved performance. It would allow the physicians that produce the savings and that can do so in ways that maintain or improve quality of care to share in the results of their efforts, even when not formally aligned in IDSs. Although safeguards need to be put in place to protect against abuses, the results of the current system are abundantly clear: continued unsustainable increases in spending and unacceptable levels of quality.
The authors thank the reviewers for many thoughtful and constructive suggestions and revisions.
NOTES
1. H.J. Aaron and J. Meyer, “Health,” in Restoring Fiscal Sanity 2005: Meeting the Long-Run Challenges, ed. A.M. Rivlin and I. Sawhill (Washington: Brookings Institution Press, 2005).
2. L.T. Kohn, J.M. Corrigan, and M.S. Donaldson, eds., To Err Is Human: Building a Safer Health System (Washington: National Academies Press, 1999).
3. F.J. Crosson, “The Delivery System Matters,” Health Affairs 24, no. 6 (2005): 1543–1548.
4. U.S. Department of Health and Human Services Office of Inspector General, “Gainsharing Arrangements and CMPs for Hospital Payments to Physicians to Reduce or Limit Services to Beneficiaries,” Special Advisory Bulletin, July 1999, http://oig.hhs.gov/fraud/docs/alertsandbulletins/gainsh.htm (accessed 18 October 2006).
5. IOM, Rewarding Provider Performance: Aligning Incentives in Medicare (Washington: National Academies Press, 2006), 93–94.
6. Library of Congress, House Resolution 1900, Conference Report Summary, 24 March 1983,
http://thomas.loc.gov/cgi-bin/bdquery/z?d098:HR01900:@@@D&summ2=m& (accessed 18 October 2006).
7. Lewis Morris, chief counsel to the inspector general, Office of Inspector General, testimony before the House Ways and Means Subcommittee on Health, hearing on gain sharing, 7 October 2005,
http://waysandmeans.house.gov/hearings.asp?formmode=view&id=4671(accessed 18 October 2006).
8. OIG, “Fact Sheet: Federal Anti-Kickback Law and Regulatory Safe Harbors,” November 1999,
http://oig.hhs.gov/fraud/docs/safeharborregulations/safefs.htm (accessed 18 October 2006).
9. Ibid.
10. Ibid.
11. OIG, “Gainsharing Arrangements and CMPs”; and E. McGinnity, “Fact and Fiction about Gainsharing,” Healthcare Purchasing News, May 2005.
12. OIG, “Gainsharing Arrangements and CMPs.”
13. OIG, “Fraud Prevention and Detection/Advisory Opinions,” http://oig.hhs.gov/fraud/advisoryopinions/opinions.html (accessed 16 November 2006.
14. Medicare Payment Advisory Commission, Report to the Congress: Physician-Owned Specialty Hospitals (Washington: MedPAC, 2005), 47.
15. Ibid., 47.
16. M.R. Chassin et al., “Does Inappropriate Use Explain Geographic Variations in the Use of Health Care Services? A Study of Three Procedures,” Journal of the American Medical Association 258, no. 18 (1987): 2533–2537; E.S. Fisher et al., “The Implications of Regional Variations in Medicare Spending, Parts 1 and 2,” Annals of Internal Medicine 38, no. 4 (2003): 273–298; and E.A. McGlynn et al., “The Quality of Health Care Delivered to Adults in the United States,” New England Journal of Medicine 348, no. 26 (2003): 2635–2645.
17. Centers for Medicare and Medicaid Services, “Demonstration Projects and Evaluation Reports,”
http://www.cms.hhs.gov/DemoProjectsEvalRpts/MD/itemdetail.asp (accessed 16 November 2006).
18. OIG, “Gainsharing Arrangements and CMPs.”
19. Health Policy Alternatives, S. 1932, Deficit Reduction Act of 2005, Summary of the Conference Agreement: Medicare, Medicaid, and Other Health-Related Provisions, 23 December 2005,
https://www.aagponline.org/uploads/abstracts/HPA%20confgreement%20sum%20122305%20FINAL.pdf (accessed 16 November 2006).
20. CMS, “Physician-Hospital Collaboration Demonstration,” http://www.cms.hhs.gov/DemoProjectsEvalRpts/downloads/PHCD_646_Solicitation.pdf (accessed 16 November 2006).
21. IOM, Rewarding Provider Performance, 168.
22. IOM, Performance Measurement: Accelerating Improvement (Washington: National Academies Press, 2006), 21.
Gail Wilensky (gwilensky{at}projecthope.org) is senior fellow at Project HOPE in Bethesda, Maryland, and a former administrator of HCFA. Nicholas Wolter is CEO of Billings Clinic in Billings, Montana. Michelle Fischer, a major in the U.S. Army, is administrator of the Department of Surgery/Department of Anesthesia and Operative Services at the Brooke Army Medical Center, Fort Sam Houston, Texas.
Read related articles
DOI:
10.1377/hlthaff.26.1.w58
©2006 Project HOPE–The People-to-People Health
Foundation, Inc.
|
