The game that shows why value-based pricing is doomed

| April 10, 2017 Opinion

Those of us who teach negotiation often use a game, developed by Harvard Business School professor Michael Wheeler, called “Win As Much as You Can." It is the ultimate exercise in how difficult it is to achieve sustainable cooperation within a structure that pushes people to behave selfishly.

The game pits four players against another. In 10 successive rounds, each participant plays an "X" or a “Y" card. The payout to each player depends on his or her choice in that round, as well as the choices made by the other three players. Your chance of winning as much as you can for yourself depends on persuading everyone in the group to play “Y," but the structure of the game provides an overwhelming incentive for each individual to play “X." So players often engage in betrayal as they attempt to convince others that they will play “Y," while they then play “X."

Wheeler's game has implications for the healthcare policy arena — and helps explain why ambitious changes in payment structure haven't worked yet, and probably never will.

Good intentions, flawed assumptions

A key element of President Obama's healthcare policy was a push for "value-based pricing," using the authority of the Centers for Medicare and Medicaid Services to experiment with pricing incentives to reduce overuse in clinical care. Private insurers were likewise encouraged to pursue this direction. It's possible, though hardly certain, that the Trump administration will maintain this focus.

Let's review the concept. The status quo in the U.S. for years has been “fee-for-service" pricing, in which doctors and hospitals were paid for piecework. Believing that the persistent rise in healthcare costs in the United States was driven by overuse resulting from this financial incentive, some policy analysts decided that doctors and hospitals should be given a countervailing financial incentive to reduce certain kinds of diagnostic tests and to avoid medically unnecessary procedures.

The predominant form of value-based pricing to emerge from this policy assumption was the so-called “global payment." Medicare or private insurers would give each health system (now called an accountable care organization or ACO) an annual number of dollars per year per patient, based on the risk profile of that system's population. If the health system could treat the patient for less money, it would keep the surplus. If its spending exceeded that budget, it would suffer a loss.

In essence, the plan consisted of CMS and private insurers trying to transfer the actuarial risk of patient care to providers, counting on the new financial incentive to change behavior.

Did the global payment regime have a chance of success? Let's go back to Michael Wheeler's game. The Obama folks were trying to get doctors and hospitals to play “Y," but the structure of the business inevitably pushes people to play “X." Sustainable cooperation is highly unlikely.

Let's see why

Say you are a gastroenterology doctor in an ACO anchored by an academic medical center. You are paid in great measure by how many endoscopies you perform. Your health system has signed a global payment contract with its insurers, looking at data which suggests that many of these procedures can be avoided by “tincture of time," dietary advice, and observational treatment overseen by the primary care doctors in its network.

You are skeptical, but you're willing to be a team player because you've been told that, should a surplus emerge at the end of the year, you'll get a bonus on your paycheck. At the end of year one, your clinical volumes have declined because of the new care management protocols adopted by your health system. Your bonus arrives, but you notice that your net income is less than before because the bonus has been shared across the health system, with a portion being given to those primary care doctors in return for their diversion of procedural cases.

Now, let's imagine you're a PCP in the network. A patient shows up with gastric distress in the fee-for-service era. Had you said to the patient, “Let's just watch your diet for a few weeks and see if the symptoms subside," the patient might likely have said, “Don't you think I should see a specialist?" With your 18-minute appointment drawing to a close, you take the path of least resistance and refer the patient to a GI specialist, who some percent of the time will perform an unnecessary endoscopy to see if there is an underlying pathology that needs treatment.

Under the global payment plan, the PCPs, like the GI specialists, have been told that avoiding such procedures will likely yield a surplus for the health system. You are skeptical, but you're willing to be a team player because you've been told that, should a surplus emerge at the end of the year, you'll get a bonus on your paycheck.

At the end of year one, the time you have spent seeing patients has increased because of the new care management protocols adopted by your health system. You've received no extra payment for that time spent, and you discover that your share of the surplus is tiny relative to your previous income, while your quality of life has declined because you're spending more time in the clinic.

Whether you're the GI specialist or the PCP, do you choose to play “Y" and be a good team player over time, or do you choose to play “X" because the incentive to cooperate is so weak and distant from your needs?

If the premise of global payments is that doctors are economically rational creatures who will respond to financial incentives, then the financial incentives have to be substantial, immediate, and transparent to be effective. Under a global payment regime, however, the incentives are minor, delayed, and fuzzy. Even those doctors who might want to be good corporate citizens will find themselves inexorably pushed to play “X," and thereby will undermine the hoped-for results.

Thus global payment regimes are hoisted on the petard of their own underlying assumption: The rationality of the participants.

In the case of CMS, the global payment plan was doomed for other reasons. For one thing, early versions had no downside: While there was the potential to garner surpluses, health systems were protected against losses. Further, under federal law, Medicare patients are mobile: They have no obligation to stick with one health system if they feel they can get more of what they want from another.

The real problem with fee-for-service

But let's go back one more step: Why is the overwhelming focus of U.S. policy based on removing fee-for-service payment schemes? In virtually every other sector of our economy, we employ fee-for-service pricing. As consumers, we evaluate the value of something against the price per unit that we will pay for that service or good. No one questions this payment regime in the rest of society. Why do we do so in healthcare?

Perhaps because, in healthcare, consumers don't make the purchase choice, or are not informed of prices, or do not have a good sense of the underlying value of what they're buying. Intermediaries usually make those decisions for them. Given these challenges, value-based pricing, however well-intentioned, is likely to be an energy-sapping distraction, while we fail in the major task of addressing the disenfranchisement of consumers in their treatment decisions.

While policymakers fuss with ill-conceived centralized policy initiatives, entrepreneurs in the healthcare sector will focus on disintermediation. That's where we may truly find value — and rational choices for all involved.

Paul F. Levy is the former CEO of Beth Israel Deaconess Medical Center from 2002-2011. He is the author of “Goal Play! Leadership Lessons from the Soccer Field."

Comments

Add a comment

Add a comment

Required fields are marked *

Your email address will not be published.

MA gives PCPs substantial incentives to "manage care", and lowers the total claims costs. The bonuses paid to the PCPs are not diluted as in an ACO, where the hospital can get most of the savings. Why can't we replicate the MA model to all patients?
Name: 
tim oliver
Email: 
tim.oliver@nexcoregroup.com
I respectively believe your analysis is flawed due to several issues with it. 1) The basic assumption you make that value-based payments equate only to a global payment or capitation ignores other concepts in value-based payment schemes such as episode-based payments. 2) In your specific argument, you assume implementation of global payment from the payer but then assume that the provider payment scheme would not adjust and remain FFS. If so, perhaps your argument holds true, but more on that below. Importantly, it is more likely that in order to be effective, provider payments too will have to switch from FFS to something comparable to global payments, in this case salary. Many providers are already on salary and it is an incomplete assessment of global payment schemes to only assume one thing would change in the system. 3) There is little doubt that a radical change from FFS to something like global payments will be tumultuous to the system. There will be some transition time as the new incentives shape the market. I contend that your argument above for year one needs to be extended and we need to consider what is likely to happen in the subsequent years as the system returns to 'steady state.' I would argue that over time with a lowered demand for the services of the GI (or other) specialists, their compensation would decrease, and conversely as we begin to again value the PCP care more, demand for that service will increase and their compensation will rise. Over time we're likely to see a reduction in disparity between the compensation for the two categories of providers and further more interest in providers becoming PCPs vs. specialists. This would be especially true under a salary system where the provider's employers could make this adjustment more directly through control of the salary. Put simply, we'll need less GI specialists and more PCPs and the market forces will likely create that. 4) Finally, the most likely reason that value-based pricing is doomed isn't because of your argument above, but rather from that simple fact that anytime we talk about creating savings in the healthcare system, that means we are creating losers somewhere in that system. In this case, we would be taking away profits from providers and returning them to the consumer. Unfortunately your industry is much better organized than the consumers and as a consequence has much more influence in the creation of the laws, regulations and policies. It is far more likely the case that if value-based pricing fails it will be due to special interests and not because of the economics. Thank you for taking the time to discuss this important subject. I agree wholeheartedly with your brief comments on the need for better information in the market, but do not feel that that is the only answer to this difficult problem. The payment scheme incentives need to change as well.
Name: 
Rob Davis
Email: 
robdavisjea@gmail.com
With all due respect to Dr. Levy , he is missing the point. Value Based Purchasing programs include penalties and incentives for hospitals to reduce adverse events, population based programs such as ACO's, and bundled payments for specific procedures or conditions. The available data for population health has generally not shown decreases in specialist incomes, it is rather the hospital volumes and revenues that are reduced. Additionally, in the case of GI, there is often a resulting increase in volumes of screening colonoscopy from the population-based efforts to improve colon cancer screening. He is certainly correct that Medicare beneficiaries can seek care anywhere they want. This encourages a trusting and longstanding relationship with their physician, and the requirement that needed care be available locally and in a timely way. When needed care is not available locally, this lack would typically be reflected in the elevated costs attributed to the ACO in the baseline period from which future costs are derived. Price transparency comes into play in this situation when the physician talks with the patient about their options for care. There certainly are situations where physicians at risk for the global expense for a population do refer patients to a medical center providing high value care- very good outcomes and experiences, and lesser cost than the competitor down the street- or in the next city. This represents a major treat to many referral centers and is beginning to result in more attention being paid to "bundled pricing". These trends are being pushed not only by CMS and commercial payers, but also by large employers and employer groups. A problem with the game analogy is that the VBP does not depend on cooperation of all the parties to achieve a successful outcome- if success is defined as "reducing cost and maintaining or improving quality". As Dr. Levy knows- Medicare does not negotiate rates, so the cooperation of the principal player in the game is not needed for its' success. Specialists and Hospitals are paid their standard rates in the population based ACO's and the goal is reduction in the rate of rise measured against an expected trend. Even for the "Bundled Payment" programs, all participants receive their standard individual payments- but the success of the total effort is measured after CMS deducts 2% from the estimated total price. Since the hospital is usually the group "holding the bag", success often demands on convincing the physicians to standardize their processes in the hospital to reduce hospital expenses and to change the use of downstream resources- often those used because of tradition rather than patient/family requirements or desires. In conclusion, physician leadership of, and cooperation in, these efforts is certainly required to achieve maximal improvement in patient care and in reducing costs. Leadership from all other participants in health care financing and delivery is also needed- and a cooperative effort is much to be preferred to any other method. However, while cooperation of all parties would not appear to be an absolute requirement for success-- it certainly improves the chances of success.
Name: 
Bruce Hamory MD
Email: 
bhhamory@gmail.com
The situation is even worse than Dr. Levy suggests. The major driver of rising health care costs is insurance company profits. Insurers largely define what health care services are available, how and when. Medicare is copied from commercial insurance. Their payment policies create the market for health care, and have produced the remarkable growth in proceduralists incomes, and the decline of primary care. As we all know, primary care improves quality and reduces costs. Because insurance profits go up (after rate increases) in line with cost increases, the insurers have a dominant strategy: anything they do that eventually increases costs will increase their profits. (They even get their usual profit on the 30% overhead waste they create! ) Conversely, any insurer whose costs rise less will have lower profits, and will be swallowed by more successful companies. Their CEO will be fired, and their employees RIF’d. This does not require that insurers be evil or callous, only that they act (consciously or not) in their own interest. Congress and others have convinced themselves that small and inconsistent financial incentives will change professionals’ behavior in the direction they desire. There is no good evidence that the kind of financial rewards and punishments in ACOs or MACRA have ANY desirable effects. Indeed, there is considerable evidence that they are harmful. In every setting where financial inducements for cognitive work have been used, the results range from ineffective to catastrophic. (E.g., Morgan Stanley bond trading debacle.) P4P Fallacy-The Vanguard Method-https://www.youtube.com/watch?v=0BACEu4m4H0 The reality is that small financial incentives for cognitive work do not produce improvement, and large incentives distort behavior. David U. Himmelstein, Dan Ariely, Steffie Woolhandler Himmelstein, Ariely, and Woolhandler: “Findings from the new field of behavioral economics challenge the traditional economic view that monetary reward either is the only motivator or is simply additive to intrinsic motivators such as purpose or altruism. Studies have shown that monetary rewards can undermine motivation and worsen performance on cognitively complex and intrinsically rewarding work, suggesting that pay-for-performance may backfire.” http://journals.sagepub.com/doi/10.2190/HS.44.2.a Pay-for-Performance is a race down a rathole.
Name: 
Peter J Liepmann MD MBA FAAFP
Email: 
PeterL@hvc.RR.com
Thanks so much for these comments. Would it be that, in the long run, the trends hypothesized by Rob Davis might occur; but I think this is a case in which the long run result simply won't happen. There are much stronger forces in effect than the minor impacts that might come from global payments. By far overshadowing all of the items mentioned is the market power if certain provider groups that that exists in certain geographic areas--which has a tendency to make sure that even in a global payment environment, those health systems get disproportionately larger rate increases from insurers. On the private insurer side, there truly is no incentive to reduce costs and premiums, as the companies are assured that a fixed percentage of the premium dollar is kept as profit. Finally, even if the financial incentive scheme were to be directionally correct, much psychological research shows that financial incentives are not determinative of sustained improvements in efficiency. The focus of policymakers on fiddling with crude financial incentives is a diversion from (1) patient empowerment and (2) implementation of clinician-led clinical process improvement. When you have a hammer, everything looks like a nail. The government had the hammer of CMS payment dollars and decided to base a scheme on that. There was a lack of rigor and thoughtfulness in this policy approach--and it was a distraction from far more valuable things. Just one final point, concerning Dr. Hamery's assertion that "Price transparency comes into play in this situation when the physician talks with the patient about their options for care." In Massachusetts, which as the strongest law I know concerning price transparency, experience indicates that it is not working. See this recent report from the Pioneer Institute: http://pioneerinstitute.org/healthcare/mass-healthcare-price-transparency-law-still-not-a-reality/
Name: 
Paul Levy
Email: 
plevy0808@gmail.com
Dr. Levy. This is a great analysis and I appreciate you bringing up these concerns. I have been a part of implemented value based programs and worked to initiate a BPCI program at a large academic institution. Although I am optimistic about payment reform, I have seen first hand that increasing quality, in an environment that already strives for the utmost quality, and reducing cost, in an marketplace that is historically complex and nontransparent, is a very difficult mission. In most industries the more the cost decreases, there is a reciprocal reduction in quality. Admittedly there are lessons to be learned from lean process improvement and six sigma theory. There is waste and inefficiency in healthcare but current value based pricing/care programs are likely too convoluted and impractical.
Name: 
Dustyn Severns
Email: 
triseverns@yahoo.com
Thanks, Dustyn. I agree with many of your points. Our experience at BIDMC was that Lean principles were very helpful in clinical settings if the process improvement process were truly driven by the clinicians and focused on those aspects of work flows that were more helpful to them in their day-to-day lives. We never, ever emphasized cost savings: It was always about how we could make life better on the floors or units for the staff. In so doing, the quality and safety of care improved and, not so surprisingly, we also saved lots of money as waste was taken out of the system. It is simply not motivational to doctors and nurses to talk about or give directives about saving money. They have more important concerns in getting through the day.
Name: 
Paul Levy
Email: 
plevy0808@gmail.com
Paul, Your analysis is spot-on. At a lecture I attended last week, ACO-expert John Michael McWilliams from Harvard Medical School detailed how anemic ACO programs have been in changing practice patterns and reducing costs. On top of the weak financial incentives, the multiplicity of ACO models introduces unnecessary complexity and friction into medical delivery. This incremental drag further weakens positive program returns. In a nutshell, ACOs have lots of squeeze, but little juice. At best, these programs provide a modest improvement on fundamentally-flawed FFS payment models. More ACOs won't get us to the promised land. New business models and disintermediation (as you suggest) have much greater transformative power. Enhanced primary care companies (like Boston-based Iora) provide care management services that build trust, focus on wellness and preventive care as well as facilitate shared decision-making and second opinions when members require acute services. The result is the right care at the right time in the right place at the right price. The also eliminate unnecessary care, manage chronic disease, incorporate behavioral health services when necessary and provide round-the-clock access. Better outcomes. Lower costs. Great customer experience. What's not to like? Care management is the health system's biggest deficiency. It's also its biggest business opportunity. Incumbent health companies clinging to FFS payment should be scared to death, but they aren't scared enough to truly put patient needs ahead of revenue optimization. As consumers figure this out, they will flock to companies that truly deliver better outcomes at lower prices with better services. For those that continue to play the traditional FFS game, "a hard rain is gonna fall." Paul, you may remember me. We overlapped when you were COO at Harvard Medical School and I was on the Visiting Committee. All the best, Dave Johnson
Name: 
David Johnson
Email: 
david.johnson@4sighthealth.com

Add new comment

The game that shows why value-based pricing is doomed