PURCHASING IN MEDICARE:
JUST ANOTHER GIMMICK?
Describing current methods of paying health care providers as failing to ensure high quality of care and efficiency in the delivery of care, the Centers for Medicare & Medicaid Services (CMS) has proposed a variety of Medicare demonstration programs to test “value-based purchasing (VBP),” also known as pay-for-performance (P4P). The demonstrations are intended to evaluate health care providers according to specified performance measures and to pay additional money to providers that meet designated performance goals. But little evidence exists that P4P systems lead to significant improvement in quality of care for patients and there is some evidence that unintended negative consequences occur.
Although relatively new to the Medicare program, VBP is not a new idea in health care. More than 100 “reward and incentive programs” are now used by private insurance companies and employers.[i] Researchers who reviewed 17 studies of P4P noted that while it may seem “intuitive that paying more money for higher-quality services will improve health care…health care does not operate like a classic free market.”[ii] Even policy-makers who endorse incentive-based purchasing programs in principle recognize that there is little evidence that VBP actually leads to improved care for patients.[iii] As The New England Journal of Medicine notes in a recent editorial, the rationale behind pay for performance is “compelling,” but the evidence linking incentive payments to better quality of care is “thin.”[iv] The editorial calls for a recognition by CMS that P4P is “fundamentally a social experiment likely to have only modest incremental value.”[v]
While CMS touts the Premier Hospital Quality Improvement (PHQI) Demonstration now underway as an unqualified success,[vi] even the hospitals that received incentive payments in the Demonstration question the method of distributing incentive payments and the measures themselves.[vii]
The PHQI demonstrations proposed by CMS do not include additional Medicare funding. Instead, they are budget-neutral; money paid to one health care provider comes from other health care providers. Reductions in Medicare spending, primarily payments to hospitals, will fund incentive payments for the nursing home and home health demonstrations. If Medicare spending is not reduced, nursing homes and home health agencies will not receive incentive payments, regardless of their performance.
While there are differences in the VBP demonstrations that CMS is currently developing for nursing homes, hospitals, and home health, some common themes and consumer concerns emerge in all of them. Here are some questions to ask:
Which measures of performance will be used? Identifying measures of good performance is both critically important and complex. Criteria for selecting measures include such factors as importance of the measure, scientific validity, usability, whether the matter being measured is under the control of the provider, and whether collecting the required data to evaluate performance is unduly burdensome. Performance measures need to be clinically meaningful, and not simply those that can be calculated easily.
Generally, a limited number of outcome and process measures are selected. While too many measures would be confusing, too few measures are insufficient to evaluate complex areas of care. The Premier Hospital Quality Incentive Demonstration project uses 33 process and outcome measures for five clinical areas. An evaluation of the Demonstration’s first year found that patients with heart failure, the most complex clinical area of the five areas studied, who received more of the Patient Process Measures did not do better; in fact, they had worse clinical outcomes, higher average lengths of stay, and higher costs than patients receiving fewer of the recommended services.[viii]
The measures that are selected are likely to affect providers’ performance, possibly in unintended, even negative, ways. For example, providers might focus their attention on the selected measures, rather than on broader and more important, but less quantifiable, areas of performance. Or providers might avoid treating patients with difficult health care needs who will give them lower scores on certain measures, exacerbating access problems for certain beneficiaries. Indeed, the Institute of Medicine (IoM) identifies three such “unintended adverse consequences” of P4P programs – “decreased access to care, increased disparities in care, or impediments to innovation” – that need to be “closely monitored.”[ix] These concerns are not simply theoretical. As described below, they have been identified as consequences of existing P4P systems.
How is performance evaluated? If providers self-report their performance data, as is typical, some level of audit is necessary to confirm that the data are accurate. If evaluations are made on the basis of Medicare claims data, how will information be included about services rendered that are not billed to Medicare? The hospital VBP demonstration has identified a broad range of alternative methods for validating hospital-reported data, such as annually or quarterly auditing all hospitals or a sample of hospitals, permitting attestation by hospitals, or imposing financial penalties for the submission of invalid data.[x]
Which levels of performance should be financially rewarded? Paying only for performance at the highest levels discourages providers at lower levels that cannot hope or expect to be eligible for incentive payments. Paying only for excellence may result in providers getting rewarded because they perform at the highest levels, even if they do not improve at all, improve only slightly, or even (theoretically) decline, but remain at the highest levels. On the other hand, paying for improvement might result in the payment of incentive payments to providers that still fail to provide sufficiently high levels of care.
Again, these concerns are real, not theoretical. A recent evaluation of the Premier Hospital Quality Incentive Demonstration, during its first two years of operation, found that “the lion’s share of bonus payments were [sic] made to hospitals with the highest baseline performance.”[xi] Thus, the highest-performing hospitals were better before the demonstration and received incentive payments primarily for remaining better. In the third year of the demonstration, payments will be reduced to hospitals that “failed to exceed the performance of hospitals in the lowest two deciles, as established during the program’s first year.”[xii] The result of the demonstration may be the transfer of money from facilities that improved the most to the highest-performing facilities that showed little or no improvement.
How large should incentive payments be and how should they be structured? Most incentive payments are quite small at present, generally only 1-2% of reimbursement rates. It is an open question whether such small incentive payments are sufficient to affect provider behavior, especially when compared with the incentives inherent in the rest (98%) of the provider’s reimbursement rate. However, not enough is known about the effectiveness of QBP to justify shifting larger portions of Medicare reimbursement into larger incentive payments.
Do quality-based purchasing programs improve care for patients? This is the most important question. As noted at the outset of this Center for Medicare Advocacy Alert, there is little evidence that pay-for-performance payment systems leads to significant improvements in quality of care for patients[xiii] and some evidence that unintended negative consequences occur. Researchers who reviewed 17 pay-for-performance studies found, in a study involving people with substance abuse, that some providers denied care for the most severely ill patients.[xiv] They reported that providers that improved the most received the smallest amount of performance pay when programs paid solely for the achievement of excellence, not for improvement, and that rewarding processes of care (e.g., documentation of smoking cessation advice) rather than outcome of care (e.g., cessation of smoking) may encourage gaming.[xv] They also found that improved performance may reflect improved documentation, rather than changes in care practices.[xvi] A hospital participating in the Premier Demonstration reported that “the scoring system rewarded it for following the recommended medical treatments more frequently but did not take into account the need to make overall improvements that actually improve care.”[xvii] Pay-for-performance also exacerbates health care disparities when poorly-performing health care providers from whom money is taken to pay for other providers are those who provide care and services to poor and minority populations.[xviii]
Conclusion: Pay-for-performance is not an obvious cure for poor quality of care or for reducing the high costs of health care. Bruce Vladeck, former Administrator of the Health Care Financing Administration (predecessor agency to CMS), finds little to support in P4P, which he describes as “the kind of seductive focus group-tested catch phrase that . . . is largely devoid of real content.”[xix] He writes:
Is the increasing “commodification” of health care, especially as embodied in “pay for performance” schemes, consistent with a thoughtful, long-term strategy to maximize quality? A comprehensive quality improvement strategy needs to focus on reinforcing the norms and values of professional responsibility, rather than on undermining them through the exercise of economic muscle.[xx]
[i] Institute of Medicine, Rewarding Provider Performance: Aligning Incentives in Medicine page 3 (2006).
[ii] Laura A. Petersen, “Does Pay-for-Performance Improve the Quality of Health Care?” Annals of Internal Medicine, Vol. 145: 265-272 (Aug. 15, 2006), http://www.annals.org/cgi/content/full/145/4/265.
Premier, Exploring the Nexus of Quality and Cost; Methodology and Preliminary Findings (Aug. 31, 2006), http://www.premierinc.com/p4p/press/quality-cost-methods-paper3.pdf.
[iii] Medicare Payment Advisory Commission (MedPac), Report to the Congress; Medicare Payment Policy pages 183-10- (March 2005). MedPac recommended that CMS begin pay-for-performance programs with hospitals, home health agencies, and physicians.
[iv] Arnold M. Epstein, “Pay for Performance at the Tipping Point,” N. Engl. J. Med. 356:5 (Feb. 1, 2007) (editorial).
[v] Arnold M. Epstein, “Pay for Performance at the Tipping Point,” N. Engl. J. Med. 356:5 (Feb. 1, 2007) (editorial). CMS announced the year two results of the Premier Hospital Demonstration as finding substantial improvement in hospital quality.
[vi] CMS, “Groundbreaking Medicare Payment Demonstration Results in Substantial Improvement for Hospital Patient Care,” (Press Release, Jan. 26, 2007), http://www.cms.gov/apps/media/press/release.asp?Counter=2076&intNumPerPage=10&checkDate=&checkKey=&srchType=&numDays=3500&srchOpt=0&srchData=&keywordType=All&chkNewsType=1%2C+2%2C+3%2C+4%2C+5&intPage=&showAll=&pYear=&year=&desc=&cboOrder=date.
[vii] Reed Abelson, “Bonus Pay by Medicare Lifts Quality,” The New York Times (Jan. 25, 2007), http://query.nytimes.com/gst/fullpage.html?sec=health&res=9401E3D9163FF936A15752C0A9619C8B63. (“The hospitals involved say that payment method is not the best model for a nationwide system. Hospital payments should focus on encouraging hospitals to make significant strides in improving care, and should not be based on rankings.” Questioning the measures, “One hospital, for example, said it found itself focusing on how it was delivering care instead of looking at how patients were actually faring.”)
[viii] Premier, Exploring the Nexus of Quality and Cost; Methodology and Preliminary Findings (Aug. 31, 2006), http://www.premierinc.com/p4p/press/quality-cost-methods-paper3.pdf. Patients in the four less complex clinical areas (pneumonia, coronary artery bypass graft, acute myocardial infarction, and hip and knee replacement procedures) had fewer complications, lower lengths of stay (and lower costs) when they received the recommended services.
[ix] Institute of Medicine, Rewarding Provider Performance: Aligning Incentives in Medicine pages 4-5 (2006).
[x] HHS, Medicare Hospital Value-Based Purchasing Plan Development, Issues Paper, 1st Public Listening Session, page 18 (Jan. 17, 2007), http://www.cms.gov/AcuteInpatientPPS/downloads/hospital_VBP_plan_issues_paper.pdf.
[xi] Peter K. Lindenauer, et al, “Public Reporting and Pay for Performance in Hospital Quality Improvement,” N Engl J. Med 2007; 356: 486-96.
[xii] Peter K. Lindenauer, et al, “Public Reporting and Pay for Performance in Hospital Quality Improvement,” N Engl J. Med 2007; 356: 486-96.
[xiii] The Premier Demonstration resulted in an improvement of 2.9%. Arnold M. Epstein, “Pay for Performance at the Tipping Point,” N. Engl. J. Med. 356:5 (Feb. 1, 2007) (editorial).
[xiv] Laura A. Petersen, “Does Pay-for-Performance Improve the Quality of Health Care?” Annals of Internal Medicine, Vol. 145: 265-272 (Aug. 15, 2006), http://www.annals.org/cgi/content/full/145/4/265.
[xv] Laura A. Petersen, “Does Pay-for-Performance Improve the Quality of Health Care?” Annals of Internal Medicine, Vol. 145: 265-272 (Aug. 15, 2006), http://www.annals.org/cgi/content/full/145/4/265.
[xvii] Reed Abelson, “Bonus Pay by Medicare Lifts Quality,” The New York Times (Jan. 25, 2007), http://query.nytimes.com/gst/fullpage.html?sec=health&res=9401E3D9163FF936A15752C0A9619C8B63.
[xviii] Vincent Mor, “Driven to Tiers: Socioeconomic and Racial Disparities in the Quality of Nursing Home Care,” The Milbank Quarterly, Vol. 82, No. 2 (2004), http://www.milbank.org/quarterly/8202feat.html (finding that lower quality nursing homes service predominantly Medicaid and African-American beneficiaries); Peter K. Lindenauer, et al, “Public Reporting and Pay for Performance in Hospital Quality Improvement,” N Engl J. Med 2007; 356: 486-96.
[xix] Bruce Vladeck, “Ineffective Approach,” Health Affairs, Vol. 23:285-286 (March/April 2004), http://content.healthaffairs.org/cgi/reprint/23/2/285?maxtoshow=&HITS=10&hits=10&RESULTFORMAT=&author1=Bruce+Vladeck&andorexactfulltext=and&searchid=1&FIRSTINDEX=0&resourcetype=HWCIT.
© Copyright, Center for Medicare Advocacy, Inc.