Thursday, January 29, 2015

Marching, but where? Moscow, I fear.

Melanie Evans and Bob Herman at Modern Healthcare report that "a new task force made up of providers, insurers and employers has committed to shift 75% of its members' business into contracts with incentives for health outcomes, quality and cost management by January 2020."

What's up? Well, the theory is that risk-based payment mechanisms like "accountable care, bundled payments and other contracts with the potential for rewards or penalties based on quality performance and better cost control" will bring about greater efficiency and higher quality in the health care system.  It is argued that the current system, mainly based on fee-for-service schemes, leads to overtreatment and waste.

In the midst of all the excitement, there are very few thoughtful observers who raise questions about the march.
I've devoted a lot of columns here to the unanswered questions and potential unintended consequences associated with risk contracts.  I've also been hard on the health care industry for a lack of analytical rigor when it comes to designing public policy prescriptions to deal with rising health care costs.  I don't want to go through all those arguments now, but let's just outline some as yet unanswered concerns:

1) As risk is shifted from insurers to providers, what adjustment will be made in the insurers' cost of business, and how will those adjustments be passed along to consumers.  A reduction in risk should be accompanied by a reduction in capital reserve requirements of insurers:  How and when will those huge investment accounts be reduced and redistributed to the public?  Likewise, if payments are made on the basis of annual population rather than individual claims, when will the insurers stop adjudicating claims data and reduce the size of their staff involved in these functions?  When and how will those savings be passed along to consumers?

2) The decisions by ACOs to take on more risk is already driving mergers and acquisitions as those entities try to expand their risk pools.  That market concentration acts to increase the leverage of the providers over local insurers, driving up rates (whether fee-for-service, capitated, or bundled).  How can we be sure that whatever savings might emerge from risk contracts are not offset by the greater market power of provider groups vis-a-vis insurers?

3) What will be the internal distribution of risk and reward for the various physician specialties within ACOs?  How will that negotiation take place? Is there any reason to believe that the current fee schedules which offer higher relative payments to proceduralists than cognitive specialists will not form the base for ACO internal transfer pricing?  If there is a surplus within an ACO, which doctors will get which share?  Likewise, if there is a deficit?

4) Similarly, what will be the internal distribution of risk and reward for the various entities within ACOs--the tertiary care centers, community hospitals, multispecialty clinics, and post-acute care facilities.  If there is a surplus within an ACO, which facilities will get which share?  Likewise, if there is a deficit?

5) What will be the governing structure of ACOs?  If, as is often the case, the large tertiary centers hold the cards, how can the other players within the ACO rest assured that they will be treated fairly in matters of risk and reward allocation?

6) And finally, if quality metrics are not properly drawn, don't risk-based contracts offer the potential for undertreatment of patients, swinging the pendulum too far in the other direction?

“If all you're going to do is you're going to put providers at risk for cost, but you don't have a robust system for measuring quality, then you're not leaving patients better off,” said Dr. Ashish Jha, a Harvard University health policy professor.

We are all pleased, of course, to see the following kind of optimism:

“As a doctor, I am very excited about the direction this is going,” said Dr. Stephen Ondra, chief medical officer of Health Care Service Corp. “For much of my career, payers and providers had an adversarial relationship that often created win-lose choices.”

But I fear that Dr. Ondra and his like-minded colleagues are naive.  The adversarial relationship he sees today may very well be replaced by a new set of adversarial relationships--between physician specialties; between tertiary and other health care facilities; and, sadly, between patients and ACOs. The main drivers of health care costs are not overuse related to fee-for-service care.  As I have noted here, they start with the changing demographics of society.  Fee-for-service is way down the list.  Quoting myself:

If we were being rational and rigorous about policy prescriptions, we would rank order these causes and determine the costs and benefits of policies that might offset them.  For example, we cannot change demographic patterns, but it could make sense to introduce public health programs to promote exercise and proper nourishment.  We could change the compensation system for primary care doctors so they could spend more time with patients.  We could subsidize physician education so they wouldn’t have to earn so much to pay off loans.  We could reform malpractice laws to reduce defensive medicine. And we could certainly engage in full-scale process improvement training of doctors and implementation of those techniques in hospitals to reduce the extra medical costs associated with harming patients.  (Those of us who have done the latter have demonstrated conclusively the cost savings, not to mention the mortality and morbidity benefits.)

But, our public policy leaders have not done this.  Instead, they assert that pricing-based over-treatment is the key problem, and they offer capitated rate plans and bundled payments as the solution.  If you look closely, you will find that most of those proposals come from payers, either insurance companies who have a corporate desire to shift risk to providers or government officials who are trying to reduce appropriations.  Or from economists, who have a tendency to simplify market behavior and blame everything on pricing regimes.  As I have said, when you have a hammer, everything looks like a nail.

We shouldn’t dismiss a change in the payment system just because it might benefit the insurers or the government, but we also shouldn’t adopt it just for that reason -- or because it fits into economists’ idealized models.  Instead, we should determine how big a portion of the over-treatment problem comes from the payment system versus other causes.  And then we should rigorously review the experience of such regimes and evaluate their costs and benefits.  We should also determine how practical it is to implement a new pricing regime. 

I have a deep seated feeling that the march that is seen as so optimistic today might start to feel more like Napoleon's attack on Moscow, nicely represented in the chart above from Edward Tufte.


Anonymous said...

Well put. Napoleon's march hangs in my office. I will observe it differently now. Let's hope we don't have to make the return trip from Moscow. I'm not convinced we know what Moscow looks like. We sure can march, however.

Anonymous said...

Hi Paul,

Our current health care system is the most expensive in the world, by far, approximately double on a per capita basis compared most other OECD countries with worse outcomes

[Worse outcomes as you have shown many times on charts you published but most easily seen in American longevity stats, out health system doesn't produce healthy people who live a long time compared to most of the "rich" world.]

The pricing system as currently designed has totally failed to control costs and not produced better outcomes for most Americans.

Since, in theory, unlimited amounts of money can justifiably be spent to "save people", how do you propose that society, say "that is too much".

You have published many times yourself, the stats showing spending on health care as a percentage of GDP.

In the 1990's after capitation was tried the first time and cause a great deal of agitation, it was quietly "decided" to let the market try to control costs.

What we have seen is market failure.

And since health care spending is now approaching 20% of GDP and crowding out many other types of beneficial governmental spending and hurting the competitiveness of the American economy, which reduces income and jobs, it is important that something finally be done.

We shouldn't let the quest for the perfect prevent the good.

Paul Levy said...

Agree that "We shouldn't let the quest for the perfect prevent the good." But we haven't shown this to be good.

I note, too, that you don't address the questions I raise.

All the points you make are truisms, but they are not dispositive of what the appropriate policy prescription should be. BTW, other countries with much lower healthcare costs as a percentage of GDP use fee-for-service pricing. That might suggest that the problem is not so much FFS as other elements I have raised.

Iyah Romm said...

From Facebook:

Important piece from Paul Levy expressing skepticism about the race to VBP. I am a strong proponent of this shift in payment and am thrilled to see CMS' putting a firm stake in the ground with clear and near goals, but only view payment reform as a tool to shift incentives and to seek to crack the nut of broken care delivery systems for more efficient, high quality, accessible care.

Paul raises important concerns and considerations for all of us in this world - whether your view is that these observations are the guardrails that government and others should seek to build parameters around (e.g., funds flow, governance, etc.), if you are a die-hard proponent of payment reform, or if you are wedded to FFS, very interested in hearing your thoughts on these challenges. Paul's six points effectively captured all the issues I wrestle with every day right now.

I also am surrounded by health care wonks and providers all the time - really interested in non-health care folks perspectives on these issues - do you think about them at all? Do you care if you go to a PCMH or an ACO? Do you know?

FrancieGrace ‏@FrancieGrace said...

From Twitter:

I'd add: #health task force w/ no patients is blind.

Anonymous said...


I am responding to your post of 10:55 AM.

You list 6 reasons why a move to bundled payments in some form is complicated and has financial risks to important stake holders.

I agree.

And I don't know how to address them.

But isn't process improvement about taking steps and then making course corrections along the way? Not letting oneself be paralyzed into inaction because all possible issues haven't been addressed.

Based on some of your past writings, and please correct me if I am wrong, I think your primary concern about bundling payments is it will require building competing networks of providers, and add to the concentration of market power and thereby result in even higher health care prices?

But here in Massachusetts at least, we already have the highest health care prices in the universe.

We also have in CHIA, Health Policy Commission along with many other entities like our non-profit health plans - a great deal of talent focused on measuring and raising the alarm if health care costs jump again.

What was once called Romneycare, was an experiment to see if we could insure everyone (as most civilized countries do) and keep health care costs under control.

Our current FFS system has no mechanism to say no to additional spending. There is no real "working" market with clear prices to consumers and consumers caring enough about saving "their" money that they make the "value tradeoffs" necessary for a market to work.

If we still want to make use of market forces in heath care there are two primary options.

1) try to get health consumers involved in the cost and value of every health care decision -- even when dying themselves, or a loved one is dying or when very complex tradeoffs between complex medical treatments are needed and now we just "trust the doctor".

I think this type of market focused on individual consumer decisions for every "health care decision" is not likely to work well except in particular circumstances.

there is another type of market solution that "might" work.

2) I will use the Wallmart analogy. We walk into a wallmart expecting everyday low prices, but we still expect to get high quality goods. When we walk into a Walmart, Target etc we don't expect to get high end goods like a Bloomingdales or the types of service we see in a store like Macy's.

Unless making a very expensive purchase, with lots of time and energy to research trade offs, we just trust Wallmart to have the best prices, because they normally do,

We need "the Wallmart" of health care networks. You know you will get high quality goods that meet your need at the lowest price possible.

In our health care market Partners is Bloomingdales.

We need some high quality, low cost hospital chains like Wallmark and Target.

Currently if a health care consumer walks into Mass General or Brigham and Womens he or she mostly pays the same price as if he or she had walked into Tufts Med, BID or Lahey, even though they all cost 20% to 30% less on average than MGH or BW on a case mix adjusted basis per discharge (similar price structure disparities exist for Partners community hospitals).

If consumers pay the same whether at Bloomingdales or Wallmart, the halls would be empty at Wallmart.

Letting hospitals build competing networks is not without risks. There will be more concentration of market power.

But with watch dogs like CHIA and the HPC maybe we can get the Wallmart effect on health care prices.

We just need health care pricing that makes consumers care enough about "their money" to pursue the value of a Wallmart like health network.

In this scenario health care consumers make one big decision when they have the time and they are healthy. They trust if they go to wallmart they will receive high quality at low cost.

We need health care chains like that.

Barry Carol said...


This is a very interesting topic.

I certainly agree that sensible tort reform to reduce defensive medicine and subsidizing physician education to reduce the debt they incur would be helpful. Paying primary care doctors more so they could spend more time with patients or work with them more easily via phone or e-mail could be very helpful in managing chronic disease like diabetes and asthma as it could keep patients healthier and reduce the number of emergency room visits and, probably, cut the number of specialist referrals as well.

For surgical procedures that lend themselves best to bundled pricing, as a patient, I would like to better understand how costs differ among hospitals. For example, if one needs a hip or knee replacement, all hospitals probably follow a similar pre-admission testing protocol, surgeons do their job pretty much the same way and use comparable medical devices, there are probably the same number of people in the OR and post-operative rehab is also likely to be similar.

So, if the actual care provided is similar, where are the cost differences? Some hospitals could pay their staff more than others, some could provide single rooms instead of double rooms, some may offer more amenities like plush lobbies, marble floors, art collections, bigger rooms, higher nurse staffing ratios and more staff doing functions unrelated to direct patient care. Differences in service levels and amenities are similar to what we see in five star vs. three star hotels. Patients may value the five star hotel and be willing to pay for it or maybe not but they need transparent prices to make a judgment.

As for consolidation among providers, various categories of retailing provide plenty of competition with only two main competitors. Think Home Depot vs. Lowe’s; Wal-Mart vs. Target; Costco vs. Sam’s Club; CVS vs. Walgreen, etc. If hospital prices get too high, ambulatory surgical centers could enter the market if CON laws don’t get in the way or people could travel outside their home areas. Non-hospital owned providers that join an ACO will have to work out acceptable contract terms so they get their fair share of the bundled payment.

Defining and measuring quality is a big challenge yet to be tackled and, of course, patients need price transparency to make intelligent choices and tradeoffs.

Mitch said...

Your analysis is spot-on. There is a saying in the fitness community that "You can't out exercise a bad diet." A corollary it seems is "The sick care system can't out-treat the impact of bad diet, lack of exercise, too much stress, poor sleep and economic unfairness."

Bob said...

This is great post on the problem of the optimism of those who will benefit by the change in payment schemes. The ACO movement started with Dartmouth, and posited that the hospitals and the tertiary centers would be the organizers (and beneficiaries) of the "natural" medical neighborhoods. Ain't nothing natural about it at all. I'd call it a Ptolemaic system, perhaps. One would predict that ACOs would reinforce the nefarious trend toward more and more administrators -- the same process that has led university tuitions to skyrocket -- and for primary care not to do primary care so much, but rather to act as the lower-cost (and lower paid) outreach of specialist care. The vaunted preventive services of primaries don't save money in the short run, and are thus bound to be short-changed.

And then there is the issue of quality of care. We have come a long way in assessing quality, it's true, but people don't understand how far there is to go. Quality assessment is still in its infancy, and so many very important aspects are still unmeasured, and will continue to be unmeasurable. Just one easily understood aspect: one of the most important aspects of quality is persistence in achieving a difficult diagnosis. This is completely unmeasured, yet what could be more important?

So resting on quality assessment ain't gonna cut it, I'd say. In the end, what you need to counter financial incentives under both FFS and capitation is professional ethics, as buttressed by lots of organizational supports. Leadership within the organizations, and a clear concept of primary care goals, and power vested within the primary care system as opposed to the procedural specialists. It's complicated. And power in primary care, which is what's needed, and a concept of true prevention which is pretty much absent in adult primary care as currently practiced, are so deficient, that the ascension of ACOs will only further the deleterious consequences of our current specialty based system.

nonlocal MD said...

Bob's insightful comments generate the thought 'Physician, heal thyself." I believe most docs within the system know very well what is wrong with it, but have either retreated into a ball of complaining helpless unhappiness, or are out there maximizing their personal gain so they can get out quicker. (see 'professional ethics' in Bob's comment)

I place the blame for this vacuum squarely on the backs of the professional societies and the AMA, which act as trade associations rather than thought and ethics leaders for our formerly esteemed profession. Inspirational leadership and enthusiasm would go a long way toward effecting change from within, but the lack of it leads to exactly what we have now.

Anonymous said...

Barry Carol,

Nice to chat again.

If you want to get costs comparison on hip and knee replacements, an analysis was done this year by CHIA or HPC [more info below] - I forget which. It compared academic medical centers, teaching hospitals and community hospitals as a group and individually. They also compared New England Baptist which is a orthopedic specialty hospital.

One factor increasing costs was unneeded post-surgical care and rehab. Some hospitals extensively used rehab others less so, with no benefits or detriments.

Another study was done {again by HPC or CHIA} showing an important factor for the higher cost of care in Massachusetts was BOTH higher levels of use and higher cost for post-hospitalization care including things like home care and rehab etc.

See the CHIA (center for health information and analysis) and health policy commission web sites. For more details.

Ricky said...

Just a quick note: the "infographic" you have at the top was not created by Mr. Tufte, but rather was made famous (for some definition of made and famous) by Mr. Tufte. I believe the creator of the graphic was named M. Minard.

Paul Levy said...

True. If you follow the embedded link, that is made quite clear.

Anonymous said...

1. Unless I am misreading your concern here, this seems like a sunk cost fallacy. While it would be unfair that we as consumers lose out on those capital reserves, that should not affect our decisions moving forward.

2. This is definitely a legitimate question. But mergers and acquisitions were happening before ACOs and would likely still be happening without them. ACOs likely are accelerating the rate of M&As, but if ACOs work well, I don't see this as prohibitive.

3-4. Large corporations consistently have to deal with different business units with different cost structures and different revenue streams. I believe through good management and leadership organizations can work out these issues internally.

5. I am not convinced that tertiary care centers will hold all the cards in ACOs. If these organizations are on the line to provide high quality care at a low cost, I would see this shifting power away from tertiary care over time.

6. This is a huge question. From reading your blog and Ashish Jha's work, I worry that we still have not figured out these quality metrics. This along with question 2 are of the largest concern in my opinion.

All that being said, I still believe that the shift to ACOs and value-based payment models is the right move. You are right that there remain many unanswered questions, but reality is constantly fraught with uncertainty and tradeoffs, and sometimes we have to make tough choices with insufficient knowledge. It is not as if we have a large number of other evidence-based policy prescriptions to choose from here. Even though you provided a number of other opportunities for policy fixes, are those more evidence-based than changing payment models, and is there any political ability or will to act on any of them at scale? Given what we currently know and the dire situation we are in, I see this as a reasonable path forward, but I could be wrong. Hopefully, instead of Napoleon's march on Russia, this becomes Eisenhower's invasion of Normandy.

Paul Levy said...

Here's some more on the topic from Budd Shenkin:

Anonymous said...

The hypothesis is that all ACO's and hospital networks are "run" by the Tertiary care hospital.

But TUFTS Medical Center & Lowell General Hospital have established a new type of network.

It is NOT based on a tertiary care hospital taking over a community hospital.

There is a common ownership entity, in this case called Wellforce dot org, that is jointly managed equally by both Lowell General and Tufts Med.

It is in it's early stages, but the strategy is very interesting if strengthening community hospitals and primary care is important to the Boston health care community

Currently Tufts Med and Lowell General get one vote on a joint board of directors. But you can see how if there were multiple community hospitals there could outvote Tufts Med.

As more community hospitals and other entities are added, you can see how the tertiary hospital becomes a "service" to the community hospitals and community physician networks and possibly other entities like rehab centers.

This is a very different model of hospital network and eventually probably ACO.

I personally think this is a model that should be supported by those who worry about too much centralized power by entities like Partners, that "suck in" patients to their tertiary care hospitals.

[Note: for the rest of the country, Boston has the highest use of Academic Medical Centers for community hospital care in the country which is one of the reasons health care in Boston is the most expensive in the country and important reason why community hospitals in and around Boston have closed and continue to close. The most recent example being Quincy Medical Center in Dec 2014. ]

A network of community hospitals, physicians networks etc that draws on the resources of a tertiary care hospital to help them succeed and thrive is an important experiment that should be helped to succeed, in my view.

I hope Massachusetts health regulators and academics are paying attention.