By Rita Rubin, M.A.
JAMA, February 21, 2018
Paying physicians on the basis of the quality of their care, not the quantity, sounded like a good idea when Congress passed the Medicare Access and CHIP Reauthorization Act of 2015 (MACRA), which changed the way the government pays physicians.
For more than a decade, Medicare had paid physicians based on the number of services they provided to fee-for-service beneficiaries, whether or not those services were needed. Under MACRA, however, Medicare would assess the quality, value, and results of care physicians provided to these beneficiaries and reward the top performers while penalizing the worst.
The problem, health policy researchers say, is that evidence about how best to evaluate health care quality is lacking and currently used measures fail to account for differences in patients’ socioeconomic and health status that could skew quality scores in favor of practices that care for higher-income, better-educated, and less-complex patients.
Medicare’s pay-for-performance scheme is a zero-sum game, so bonuses for practices that score higher on quality measures are offset by the penalties levied on practices that score lower or don’t even submit data to play the game. “For every winner, there has to be a loser,” said Donald Berwick, MD, MPP, an administrator of the Centers for Medicare & Medicaid Services (CMS) during the Obama administration.
In this game, the losers are more likely to be physicians who care for poorer or sicker patients, and, in turn, their patients. “We are literally taking money from providers that serve the poor and giving it to providers that serve the rich,” said Karen Joynt Maddox, MD, MPH, a cardiologist and health services researcher at the Washington University School of Medicine in St Louis.
Continuing the Cycle
With the incentives stacked as they are, medical practices could game the current payment system by cherry-picking the patients who are more likely to try to stay healthy and, as a result, have better outcomes, thus making it appear that their physicians provide higher-quality care, said J. Michael McWilliams, MD, PhD, a general internist and professor of health care policy at Harvard Medical School.
He said he is more concerned that pay-for-performance incentives could discourage large organizations from opening or acquiring practices in poorer areas.
In a recent study in Annals of Internal Medicine, McWilliams and his coauthors found that the PVBM (Physician Value-Based Payment Modifier Program) had no effect on the quality or efficiency of care provided and likely exacerbated health care disparities by disproportionately penalizing practices that care for lower-income or sicker patients.
As of January 2018, the PBVM was succeeded by the Merit-based Incentive Payment System (MIPS), which differs mainly by allowing medical practices to choose the quality measures on which they’d like to be assessed. “Essentially, we are going from providers knowing the test questions in advance to providers being able to pick the test questions,” McWilliams said. “It further weakens the incentives of the program and allows for a lot more gaming behavior.”
And so, the cycle will persist, said Austin Frakt, PhD, coauthor of the editorial accompanying McWilliams’ article. Without accounting for patients’ socioeconomic status, practices that are otherwise doing a good job but happen to be serving lower-income patients will continue to be disproportionately hit with penalties, which “makes it harder for them to do a good job,” said Frakt, who holds appointments with the VA Boston Healthcare System, the Boston University School of Public Health, and the Harvard T.H. Chan School of Public Health. “The resources matter,” he added.
“I could see members of Congress saying, ‘Oh, sure, we should be able to measure quality in medical practices,’” said (Karen Joynt Maddox, MD, MPH, a cardiologist and health services researcher at the Washington University School of Medicine in St Louis). “The reality of that is much further away than any of us would prefer. We don’t even have the ability to collect and report high-quality clinical data…and that’s worse in underresourced settings.” However, even in places where the data are available, “we don’t have the science behind the quality measures,” she added.
Adjusting for Case Mix
“The critics say case-mix adjustment is incomplete. It will always be incomplete. You can’t get all the right variables on the page,” said Berwick, president emeritus and senior fellow at the Institute for Healthcare Improvement. On top of that, he said, providers could manipulate data for case-mix adjustment by coding patient visits as more complex than they actually were.
Race might be an appropriate variable to use for case-mix adjustment, Berwick said. After all, “blacks have a different disease burden. They’re of lower wealth. Overall, in the United States, the consequences of racism are still with us.”
However, that approach, while more difficult to game, has its own problems, he said. “What you’re then doing is saying, ‘If I am treating a population of more African-Americans, then I can have worse care, and you’ll adjust for that. It’s kind of like excusing the performance issues. How do you feel about that if you’re black?”
Seeking the Patient Perspective
According to Berwick and others, the current approaches to assessing the quality of care physicians provide leave out a key player: the patient. “We need to shift the question from ‘How are we [physicians] doing?’ to ‘How are you [patients] doing?’” said Berwick.
“It’s a paternalistic argument to say that we—CMS, scholars—are going to decide what high-quality means, and we’re just going to try to impose that or incentivize that,” Frakt added.
Berwick recently served on a panel convened by the Organisation for Economic Co-operation and Development (OECD), whose 35 member countries include the United States, that recommended using patient-reported indicators to strengthen the international comparison of health system performance.
“There remain substantive gaps in what is known about the experience of patients and the outcomes of care from the patient’s point of view,” panel members wrote in a report released in January 2017.
However, patient satisfaction should not be confused with outcomes, Frakt said. “Maybe I had a great visit with a physician, and I give him 5 stars out of 5. But what I really care about is getting my tendonitis resolved so I can run again.” No matter how likeable the physician, if a patient’s tendonitis has not improved by the time of a follow-up visit, the patient is not going to consider that a good outcome, Frakt said.
In the end, though, no matter what approach Medicare uses to pay physicians, it cannot be counted on to eliminate health care disparities, Berwick said.
“Pay-for-performance on the whole may well have aggravated disparities,” he acknowledged. But, he noted, “We have problems in social justice and income inequality and unfairness that need to be addressed directly, not indirectly, through public policy.”
By Don McCanne, M.D.
It seems that these days almost every forum or treatise or journal article on health care financing reform begins with the admonition that we must start to pay for the value of health care delivered rather than for the volume, or the quality instead of the quantity. But then just try to find in the discussion that follows precise methods of demonstrating true global quality or precise methods of demonstrating how quantity can be reduced by eliminating selectively only that care that is not beneficial. Yet this aspiration has been repeated so often that it has become an empty meme.
Today’s article by Rita Rubin demonstrates that trying to comply with this meme fails in achieving the goals of reducing volume while increasing quality, but, worse, it may have adverse consequences that negatively impact patients and the professionals who care for them.
Years ago, in discussing the medical-industrial complex, Arnold Relman was one of many who pointed out that paying fee-for-service incentivizes greater volume regardless of the value of the services, whereas paying capitation (a fixed amount per patient regardless of the amount of care rendered) incentivizes a lower volume of care that increases margins by reducing overhead. Some have said that physicians should be paid a salary instead, but Relman pointed out that salaries incentivize sloth – in some ways a greater evil than greed.
The policy community is spending too much time looking at policies that might change the behavior of the health care professionals, even though plenty of studies have shown that there is not much impact from their schemes. They underrate the professionalism of the practitioners of the art of medicine. Give them the infrastructure and practice environment in which they are free to apply their skills, and they will do the right thing with and for the patient. Quantity is controlled by the resources available and quality is incentivized by the favorable practice environment. Sloth is tempered by by reducing the causes of burnout, such as having to comply with intrusive protocols and measurements that have little positive impact on patient care.
Donald Berwick reminds us that it really is all about the patient, and not just in the health arena. As he says, “Pay-for-performance on the whole may well have aggravated disparities,” but “we have problems in social justice and income inequality and unfairness that need to be addressed directly, not indirectly, through public policy.”
The next time a presenter opens with “value instead of volume” or “quality instead of quantity,” please feel free to scream – as loud as you can.
Stay informed! Visit www.pnhp.org/qotd to sign up for daily email updates.