Exploring the Ethics of Value-Based Purchasing

By Philip Nathanson for VBP monitor

The ethics of value-based purchasing (VBP) seem straightforward enough, don’t they? A purchaser of healthcare ties compensation to how providers perform on various measurable aspects of the triple aim. Who could have an ethical problem about rewarding providers for better population outcomes, or enhanced patient satisfaction, or more cost-effective care?

Actually, as soon as VBP began to gain serious traction in health policy and industry circles a decade ago, providers began publicly worrying that there might be serious ethical problems with it. Over the years official statements of the American Medical Association (AMA), American Hospital Association (AHA), and many other provider groups have supported the concept in theory, but have also expressed strong concern about its implementation. They have pointed to the potential for VBP or pay-for-performance (P4P) programs to have adverse impacts on patients, to treat providers unfairly, or to be too methodologically weak to use to reward or punish performance.

Today VBP is accepted as healthcare’s future, and we are hearing fewer voices raising ethical issues about it. However, that doesn’t mean that some of the concerns that have been raised by providers in the past aren’t valid. Let’s look at three of the most compelling of them.

• Ethical Concern No. 1: Are VBP and P4P programs risky for patients? “Initiatives that provide incentives for a few specific elements of a single disease/condition may neglect the complexity of care for the whole patient, especially the elderly patient with multiple chronic conditions,” warned the American College of Physicians (ACP) in a 2007 “Ethics Manifesto.” ACP added that “P4P sets up a potential conflict between the physician’s duty to the patient and a competing interest in achieving a performance measure – whether the measure is a priority for that patient or not.

  “That is one example of how VBP program critics suggest that its incentives could tempt providers to act unethically toward patients – and there certainly are enough single disease/condition measures in VBP programs for this to be an issue. Another example is this: inherent in VBP programs that penalize poor performance is a strong incentive for providers to “deselect” the sickest patients with chronic conditions. There is anecdotal evidence that this has occurred in U.S. VBP programs, and there exists suggestive statistical evidence that it also occurred in a British National Health Service P4P pilot program that had a 20-percent performance incentive for quality performance.

  However, one could also point out that potential conflicts of interest are built into any kind of provider reimbursement system – and that some percentage of providers will always succumb. For example, fee-for-service medicine in general has its incentives to give more care than needed, and its incentives to give less, with examples of each abound. One could also argue that if providers yield to these temptations, the problem may be less in the incentives than in the providers themselves, whose ethics should require them to give the right amount of the right care for the patient – no more, no less – regardless of whether there is more money to be made by acting otherwise.

  But it also is fair to suggest that, providers being human, it makes sense wherever possible to eliminate potential conflict by aligning the providers’ interests as closely as possible with those of the patients. In fact, today’s VBP programs are balancing single-condition measures with those that get at the overall health and well-being of each patient receiving care. For example, the current Centers for Medicare & Medicaid Services (CMS) hospital VBP measure set includes Surgical Care Improvement Project measures such as prophylactic antibiotic received within one hour prior to surgical care. The current Medicare accountable care organization (ACO) measures set includes such measures as all-cause unplanned admissions for patients with multiple chronic conditions and health and functional status.

  These measures help, but fuller alignment of provider and patient incentives will require more transparency in care delivery, more education of patients as consumers, more meaningful use of data, and more assumption of significant risk by providers – all of which, according to both public and commercial payors, are coming in the years ahead.

• Ethical Concern No. 2: Is it ethical to penalize poor performers that arguably are in most need of funding for improvements? The pain of penalties for poor performers is compounded by another bite out of their revenue “off the top.” VBP programs are funded by taking away a small amount of total reimbursement to most providers (including many struggling hospitals and systems) to create a pool to pay out to high performers. Unfairness to providers with limited resources is one of the reasons that provider groups have historically (and unsuccessfully) opposed the “pool” approach to funding.

  A counter to the unfairness argument is to point out that VBP is a purchaser initiative. So long as we have a free-market, competitive healthcare system, should payors be expected to keep poor performers in business? Au contraire, many would argue; driving poor performers out of business is in everyone’s long-term interest.

  But don’t government payors have a responsibility to keep subpar providers afloat and work with them to get better if they are, say, the only providers for miles around? CMS does recognize that responsibility and has exempted critical access hospitals (CAHs) from the Readmissions Reduction Program. Participation in the Hospital Value-Based Purchasing Program is limited to hospitals with enough admissions so that the metrics are meaningful. Should government VBP programs do more than that to protect struggling but critically located providers? That’s a question that is sure to continue to be debated going forward.

• Ethical Concern No. 3: Is it fair to reward or penalize performance using measures that may not be methodologically sound enough to be reliable? The answer is obviously no, but there are plenty of arguments over what constitutes “sound enough,” and whether the perfect should be the enemy of the good.

  One line of attack tries to disqualify all measures with specific numerical metrics because although they are all right in theory, they don’t work with live patients. A good example is the analysis of the venerable hemoglobin A1C level metric in the Society for General Internal Medicine’s 2009 white paper on the ethics of P4P programs. SGIM concedes that on the basis of the best evidence, “it seems reasonable to require that diabetic patients achieve below 7.0. However, in patients with previous hypoglycemic episodes this target might in fact be dangerous. Or, in a particularly difficult-to-control patient, a decline in hemoglobin A1C from 10.0 to 9.0 might be a remarkable achievement and more validly represent high quality care than a decline from 7.3 to 6.9.”

  The rebuttal to this analysis – and it’s applicable to any attack on a “best evidence” metric – is that the measure is not intended as a guideline for treating individual patients. Like all good population measures, the hemoglobin A1C level metric is built on solid studies that show that most diabetic patients can and should achieve the 7.0 target.

  Another ethical criticism of some VBP measures is that they hold providers accountable for aspects of care over which they don’t have control. A prime example is CMS’s Readmissions Reduction Program, which on its face assumes that if a readmission is preventable, then hospitals have the power to prevent it. Evidence is beginning to show that this is not necessarily the case. For instance, it turns out that the majority of readmissions for total hip or total knee replacement (in the readmissions program beginning in 2015) are for surgical site infections that occur outside the hospital.

  So is the Readmissions Reduction Program fair? Supporters argue that one of the purposes of the program is to force effective coordination of care across the spectrum of care. If the hospital can’t address the cause of the readmissions by itself, it behooves it to reach out to primary care physicians, community resources, and patients themselves, and to grow its patient education and care management resources as necessary. And it must be said that many if not most hospitals are making those investments today.

To sum up, our review of key ethical concerns raised about VBP has shown that some aspects of these issues are being dealt with effectively today; efforts to deal with others are works in progress. All the issues have moral weight and will need to be addressed through careful design of VBP programs, their incentives, and their measures going forward.

“Total risk” population health management contracts such as Massachusetts Blue Cross Blue Shield’s Alternative Quality Contract (AQC) may be the bellwethers, from both financial and ethical perspectives. The AQC gives physician groups a global budget for providing all needed care for patients and tracks a number of quality metrics, most of which have been in place since 2009. The latest AQC evaluation shows that spending has slowed by about 10 percent while quality performance continues to improve. As one program evaluator put it: “the more you include (in the budget), the more the incentive for the provider to really think holistically about the patient’s care … so that it’s high-quality and affordable.”

About the Author

Philip Nathanson is the president of Nathanson Consulting LLC. Phil has held quality leadership positions at CMS, Aetna, and NCQA. His consulting clients include hospital systems, HIM firms and biotech companies. His articles on quality and healthcare management have appeared in Becker’s Hospital Review, H&HN Online, Topics in Healthcare Financing, and other journals.

Contact the Author

philip@nathansonconsulting.com