NEJM’s Fundamentals of U.S. Health Policy, Part 3: Do We Spend Too Much?

Photo by Scott R on

This next part in The New England Journal of Medicine’s fundamentals of health policy series is written by my favorite health policy writing duo–Drs. Baicker and Chandra. They both do amazing research independently, but when they work together to write an article, it seems to be extra insightful and interesting.

Their task with this article is to help people think deeper than the simplistic sentiment, “The U.S. spends way more than every other country on healthcare; we need to cut back, and any increase in spending is wasteful!”

Key insight: Only looking at the aggregate number obscures many important facts about our healthcare spending; digging a little deeper totally changes the conversation.

Here are my favorite examples of this from their article:

  • As nations grow richer, they spend more on healthcare. So, based on that alone, the U.S. would be expected to spend more than nearly every other country. Of course, this doesn’t explain all the difference between our spending compared to other countries, but it explains a lot of it, and it’s not a bad thing. A related side point they make about this is that it “highlights the challenge of putting all Americans, with very disparate incomes, into a single insurance plan.”
  • Studies that conclude our higher spending is purely due to higher prices (rather than higher quantity) aren’t able to account for all the inter-country differences in quality or intensity of care. This one was news to me, and you’ll see in my prior writings that I didn’t know this. You see, I thought that if a study shows that we generally have the same number of hospitalizations, doctor visits, prescriptions, imaging studies, etc., they would have controlled for differences in what kind of doctor visits (primary care vs. specialist) and scans (a 0.5-Tesla MRI vs. a 3-Tesla MRI) and such were being delivered. But that’s wrong–researchers don’t always have the data they would need to be able to do that. This means we need to look closer at exactly what we’re getting when we are paying a higher price than other nations, which will help us distinguish if it’s just plain overpriced in the U.S. or if it’s a substantially better service.
  • If we look at specific health outcome domains, we find out that we overspend on some of them and actually underspend on others. This means that sometimes increasing spending is actually a good thing, like for vaccines or other preventive care. When we look at spending this way, we can start to evaluate whether reforms’ spending impacts are effective not based on whether they increase or decrease aggregate spending, but instead we can base our assessments on how well they do at reducing overspending on the low-value care and increasing spending on the high-value care.

I hope that helps you think a little differently about (i.e., be a little more critical of) aggregate health spending references like it did for me.

And, to close, here’s a thought-provoking statement they drop in the middle of their article: “The debate about whether health care is a right sidesteps the more difficult and important question of how much health care is a right that should be ensured through public programs.”

Who Can Fix Irrational Healthcare Spending?

Photo by Karolina Grabowska on

A couple weeks ago, I wrote about an experience I had at work with a patient wanting to stay in the hospital for two extra days just so he could spend a few hours with his daughter from out of town before going to a skilled nursing facility where no visitors would be allowed. The cost of staying in the hospital those couple extra days was probably at least $4,000 total, but the patient and his family were (rationally) ignoring that cost because they weren’t going to have to pay for any of it directly themselves.

The solution I proposed to such system-level irrational spending was to have the person making the purchase decision bear at least part of the cost of that decision. That way, if they choose something more expensive, it will be done with a consideration for the additional cost that choice entails.

As context for my proposed solution to such a problem, remember three things: (1) every system is perfectly designed to get the results it gets, (2) the design of a system generates a set of incentives, and (3) incentives are what drive the results of every system. So does no one have an incentive to get the people making purchase decisions to bear at least part of the price differential?

This is how I look at it:

Anything that lowers the total cost of care (while at least holding quality constant) is desirable for whoever is bearing the burden of that total cost of care. And the parties who bear that burden are the insurer and the patient, to varying degrees.

But since the patient typically doesn’t have much of an idea about what their out-of-pocket cost will be or how those costs will be impacted by choosing a different procedure or a different provider, they are not in a position to do anything about it.

The insurer, on the other hand, has the capacity to know–or at least give a reasonable estimate of–the cost of something. Therefore, we need to look to insurers for the necessary changes.

But think about this proposal from an insurer’s perspective . . .

The young energetic intern pitches such a plan to the executives, “Hey, how about we offer a new and innovative insurance plan that gets rid of the typical deductible and instead makes enrollees pay a 50% coinsurance on each service they receive (subject to their out-of-pocket max, of course). We could set the total price of each service (and, therefore, their coinsurance portion) at our negotiated rates with each provider. And we’ll make an app that will allow them to compare prices beforehand.”

The room’s executives would immediately see all the problems with such a plan. “This is too complicated for people! They will hate it and never want to get an insurance plan from us again.” “We can’t divulge our negotiated rates, all the providers who find out they are getting paid less will be angry and demand more from us.” “What about things that are not shoppable?” “If a single bad publicity event comes out of this experiment, it will severely damage our reputation and we’ll lose market share as a result, so we stand to lose more than we could gain.” “We’re going to have to make premiums extra low for such a plan to be able to convince people to try it out, but there’s no way to easily communicate how this plan is different, so a lot of people will just choose it because it’s cheap and says it has a $0 deductible and then they won’t be prepared to pay all their coinsurance costs.” “What about out-of-network coinsurance rates?” “Nobody’s ever done something like this before, so it’s too crazy to try unless we are likely to make a killing on it, which we aren’t.”

I talked about these same issues in my Why Insurers Don’t Innovate post a year ago. And not all of them are immediately solveable. But if we could address at least a majority of those executives’ concerns, I believe we would start seeing some enterprising insurers try this out.

The big thing would be solving the novel complexities that such a plan would create and then finding a way to convey this information to potential enrollees, either in the health insurance marketplace or in employer benefits explanations. And we would have to hope that many people would be willing to use an app to check healthcare prices in compensation for paying a much lower premium.

There would be lots of challenges to solve, and apparently no insurer has yet determined that the potential benefits are worth the potential harms. But I have hope that this will happen and be successful sooner or later. Maybe if the first enterprising insurer to try it can identify some first-mover advantages and create some barriers to imitation, they would stand to gain a lot more.

Also, remember there’s the static-world benefits to patients making value-sensitive decisions (they get better deals on the care they buy), and then there’s the dynamic-world benefits (the market starts to evolve toward delivering higher value). This kind of insurance plan design change, as it starts and then spreads, will enact a big change in providers’ incentives. In other words, the system will be fundamentally changed, which will result in much higher value care being delivered as it evolves in response to that change.

Healthcare Experts Often Support Good Healthcare Reforms for the Wrong Reasons

This week’s post is a little later than usual, but next week will be back on track with a Tuesday post about Hayek’s book about socialism, The Road to Serfdom, and how it fits into my framework for categorizing governments.

Something I have noticed for many years now is that many good and important healthcare reforms are touted by experts for the wrong reasons. Supporting a good reform for the wrong reason may seem harmless, but without a clear understanding of the principles behind why the reform is important, the implementation may undermine much of the benefit of the reform, or it may not be evaluated based on the right expected impact (and, therefore, cause the reform to be incorrectly judged as a failure). Either one of these mistakes could ruin the reform.

Example 1 – Quality metrics reporting: This refers to making providers track and report a variety of quality metrics, which are then usually used to give quality-contingent bonuses. These quality metrics are also often reported publicly with hopes that it will add some accountability to providers and motivate the lower-quality ones to improve.

What many experts don’t realize is that quality-contingent bonuses are not going to make a big dent in our healthcare problems. They also don’t realize that the main purpose for quality metrics should be to help people make value-sensitive decisions, which means the tracked and reported metrics need to enable people to do this. Commonly used metrics these days, such as aggregate mortality numbers and overall patient satisfaction scores, aren’t super useful at achieving this goal.

Example 2 – High deductibles: Some experts say that if people have high deductibles, they’ve got some “skin in the game” and will therefore stop being such spendthrifts, which will decrease overutilization and total healthcare spending.

It’s true that a high deductible will reduce healthcare spending;, although, unfortunately, people tend to decrease unnecessary AND necessary care. That’s what the classic Rand Health Insurance Experiment demonstrated. But lowering spending is not the main purpose of high deductibles. The primary benefit of them is that they make people actually consider price when they are choosing where they will get care, which allows people to start preferentially choosing higher-value providers (i.e., make value-sensitive decisions). Of course, this only applies to services that cost less than the deductible.

Example 3 – Bundled payments: These are seen as a way to get providers to integrate more and, through that integration, “trim the fat” (what’s with all the flesh metaphors?). Usually the reduction in total episode costs comes from providers becoming less likely to discharge people to skilled nursing facilities.

Bundled payments do get providers to send fewer patients to nursing facilities and to find other superficial ways to decrease total episode costs, but the primary benefit is that they allow people to compare, apples to apples, the total cost of a care episode. Again, it’s all about removing barriers to value-sensitive decisions. This will lead to complete care process transformations as providers become motivated to improve value relative to competitors and are assured they will win greater profit as a result. So implementing bundled payments with a single provider in a region will likely result in only very modest benefits, which will come from those superficial low-hanging-fruit types of changes.

That’s enough examples for this week! Merry Christmas, and may everyone do good things for the right reasons.

An Example of How Evidence Can Be Misleading (Bundled Payments Version)

Photo by Pixabay on

Bundled payments are a proven strategy in non-healthcare industries, only we call them by different names. We say we’re paying for a “hotel room” when really we’re paying for the use of the room, the cleaning of the room, the “free” wifi, the “free” continental breakfast, access to the concierge, etc. We say we’re paying for a “cruise” when really we’re paying for the cabin, the unlimited food, the port fees, the access to the shows, use of the onboard pools, etc. These terms (“hotel room” and “cruise” are just two examples) are shorthand terms that refer to the bundle of goods and services you get when you pay that single price. That’s how it works in every industry.

Except healthcare. When we get a bill for a knee replacement, it only contains the surgeon’s cost (okay okay, including the hardware). But there’s also a bill for the anesthesiologist. And for the operating room. And for the hospitalization if you need to stay.

Why would all those things be listed individually if you know you will need a surgeon, an anesthesiologist, and an operating room every single time you get a knee replacement?

Enter bundled payments. They finally turn healthcare a little bit more into a normal industry by allowing patients to pay a single price for the bundle of services that should always be included in that one upfront price.

Sidenote: yes, healthcare is and should always be different from other industries in many important ways. For more details on that, read Arrow’s seminal article on the topic. None of this is incompatible with those insights.

But apparently bundled payments are not very helpful according to this Health Affairs review of the evidence. What gives?

When I see a paper like that, I appreciate the effort to summarize the evidence, but I cringe thinking about all the headlines and misinterpretations and misconceptions that it will perpetuate. People are going to start saying that our large-scale efforts to implement bundled payments are a waste of money . . . and they have evidence to back them up!

But that’s totally the wrong conclusion. This is how I interpret this study: I first think about the overall purpose of bundled payments. When you put their role into the context of the Healthcare Incentives Framework, their purpose is to allow patients to know up front the full price for accomplishing the job they have (get their knee functional again). This enables the patient to compare the price of different options. If they also have quality information of those options, now patients have the ability to shop for the best value (Value = Quality / Price). And when patients start choosing which provider to do their knee replacement based on value, market share starts to shift to the higher-value providers, thus forcing the lower-value providers to change in ways that either raise their quality, lower their price, or (hopefully) both! This is the potential benefit of bundled payments–it is an essential component in stimulating a newfound evolution toward higher quality and lower prices in the market for that specific service.

Compare that to how these studies evaluated the utility of bundled payments: They implemented them for a single procedure and usually with just one of many insurers a provider contracts with. Thus, in most cases the providers were still reimbursed the old fashioned way by all their other insurers. And I would guess that very few competitors in any given region were participating in the same bundled payment program. Therefore, there was no way providers were going to completely shift how they deliver the service because of the narrow scope, a lack of uniform incentives, and a lack of any strong financial imperative to do so (their biggest risk was only of losing a percentage of revenue on small portion of their patients–enough to motivate them to try to do some things a little differently but not to completely redesign how they deliver care–and the risk of losing market share to their competitors due to low value was almost nil). In short, these studies did not get even close to creating the environment for an evolution toward higher value.

With this as context, it’s a surprise to me that any of these bundled payment studies found any benefit at all!

Now, if I could perform my dream study (the design details here and here), that would make a splash. It would get us much closer to the true estimation of how impactful bundled payments could be in healthcare. And until a study like that is done, remember the importance of context, and take any evidence on the impacts of bundled payments with a large grain of salt.

Next week I’ll start looking at Joe Biden’s healthcare plans, so look forward to that!

Evaluating the ACP’s Vision for Our Healthcare System, Part 3 of 3: Coverage and Cost of Care

The last in my series evaluating ACP’s “vision for a better U.S. health care system.” Check out the intro to the series, Part 1, context for Part 2, and Part 2.

This is the paper that generated the most media attention, much of which incorrectly asserted that the ACP endorsed “Medicare for All.” So let me, at the outset, clear this up: The ACP doesn’t endorse Medicare for All in this paper. It endorses a couple different options to achieve universal coverage, one of which is a single-payer system. There are many ways to achieve a single-payer system, and Medicare for All is just one way to do that. Maybe this sounds like semantics, but I think it’s important to be accurate here because the term “Medicare for All” carries with it a lot of specific ideas about how a single-payer system should be implemented, and it also carries with it specific political affiliations. The ACP was not committing to any of those specific ideas or political affiliations; the writers were only endorsing the general single-payer system approach as one of two options for how to increase insurance coverage.

Now, on to assessing the details of the paper.

Part 1 of this paper reviews the state of insurance coverage and healthcare spending, which sets the stage for Parts 2 and 3 to talk about ways to expand coverage and lower care costs.

Part 2 is mostly what I want to talk about. First, it asserts unabashedly that the ACP feels that universal health insurance coverage is essential. And since having insurance coverage is pointless if its spending requirements are not affordable or if providers are unavailable, it adds those access details in as also being essential.

Then it lists two options that it sees as being able to achieve that goal: a “single-payer financing approach” and a “publicly financed coverage option.” Let’s look at each one.

Single-payer financing approach. This means a single government-run insurance company provides insurance for everyone in the country. It doesn’t necessarily prevent people from purchasing private supplemental insurance, but it would cover everything considered to be essential. The paper then discusses some of the well-known benefits and concerns that come along with having a single-payer system. The interesting part to me, which makes sense coming from a physician organization, is their particular concern about how this would affect physicians. It could decrease administrative burden and uncompensated care, but it could also decrease autonomy. And the big concern is that if it relies on “Medicare’s flawed payment system,” it would perpetuate a few undesirable things, including bad incentives (such as an over-reliance on fee for service), the overvaluing of certain services (procedures), and unsustainably low reimbursements (that would make some providers go bankrupt). Any prior “Medicare for All” proposal I’ve seen relies on Medicare’s payment system, so this is a good example of why the ACP is not endorsing Medicare for All specifically.

Publicly financed coverage option. Another name for this is a “public option.” Basically it’s a government-run insurance plan that will be offered alongside private insurance plans. It would be available for employees to opt into rather than use their employer-sponsored insurance as well. An essential part of this insurance plan is that it would have premium and cost-sharing subsidies (so that, again, lower-income people don’t end up with useless coverage). Later on in this paper, they say that no matter which option is chosen, it needs to have included with it a mandatory or automatic enrollment component. I’m not sure how premiums would be paid for by people auto-enrolled into the public option plan–possibly through filing their taxes–but isn’t this looking more and more like the Affordable Care Act? Require people to have insurance coverage and subsidize lower-income people to be able to buy it? Yes, that’s the ACA. The one difference is that one of the insurance plans people can select from will be run by the government (which, incidentally, was originally part of the ACA plan).

So those are the two options the ACP supports. They also mention another approach: convert all insurers to non-profits and require everyone to have health insurance (another variation of the ACA, but this time there’s no public option and instead all the insurers are non-profits). But they say there is not enough information on how such a system could be applied to the U.S., so they don’t endorse it. And all the other options for reform that are out there (such as “market-based approaches”) would not achieve universal coverage, so they were eliminated from consideration.

Even though the ACP is politically neutral, you can see that their foundational beliefs and priorities line up much more closely with modern liberal thinking rather than conservative thinking/beliefs/priorities. This is what I expected, and I discussed it in my intro to this series.

Part 3 discusses strategies they support to lower healthcare costs. There are 5 of them: invest more in primary and comprehensive care, reduce excessive pricing and improve efficiency, reduce low-value care, rely more on global budgets and all-payer rate setting, and use more reference pricing.

Rather than get into the merits and drawbacks of these different policies here, lets finish this series with a brief discussion about whether the two options they endorse to attain universal coverage will get in the way of the ideas we talked about in Part 2.

If you will recall, the second ACP paper we discussed gave 6 different recommendations, the two main ones being to (1) give patients price and quality information to help them choose higher-value providers and (2) getting providers to shift to value-based purchasing arrangements (especially different forms of capitation) with insurers.

With a single-payer system, usually the implementation involves setting a uniform nationwide price for each service (adjusted by region for cost of living). This would completely ruin their idea of getting patients to choose providers based on value because the price would be the same for all providers. I will not get into detail on how, but setting a price floor like that completely distorts providers’ innovation incentives. It would kill pretty much any cost-lowering innovation that results in slightly reduced quality, even if the innovation could result in drastically lower cost.

But, on the upside, since there’s only one insurer to deal with, a single-payer system would allow for perfectly uniform incentives for any provider that chooses to enter a VBP arrangement!

With a public option like they have described, there is the exact same problem with distorted/ruined incentives due to uniform pricing, but at least the extent of the problem is limited to the number of patients who are on the public option plan. And, as for providers’ efforts to enter into VBP arrangements, it would probably make it slightly easier to get uniform incentives because I assume VBP arrangements would be harmonized between Medicare and this public option. But any provider who wants to design their own VBP arrangement is going to have to have an even harder time attaining uniform incentives because I suspect they would have to convince the plan administrators of Medicare and the public option independently, which means they’re trying to win over yet another public payer to be able to move forward with their idea.

Am I saying that the ACP’s efforts to accomplish universal coverage will interfere with their efforts to improve the care delivery system? Yes. The two options they endorse don’t necessarily need to be implemented in a way that does that, but I see nothing in what they’ve written that specifies that they should be implemented in a way that mitigates those problems. (For details on how this could be done, see what an optimal single-payer system would look like here and what an optimal ACA-type system would look like here.)

I think different groups of people were working on each paper, and even though they were reporting their findings and recommendations to the same group of ACP leaders for approval, it’s quite a difficult task as one of those ACP leaders to be presented with complex recommendations from two different groups and figure out exactly how they might conflict with each other.

Well, there we have it. The ACP endorsed some things that would be very beneficial, but they fell into the same trap of not thinking about these things from a complete system perspective, so their efforts to get more people into the system will interfere with the changes the system needs to deliver better value over time. This is why any healthcare reformer first needs to have a firm understanding of the core causes of low value and how those can be resolved before deciding on the details of how to enact universal coverage.

But I am happy the ACP jumped directly into this arena. It shows that powerful physician groups are also interested in figuring out how to fix the healthcare system. Maybe the biggest benefit from their efforts of researching and releasing these papers (and then having them critiqued) is that it will make the ACP and other provider groups more likely to recognize and support policies that will truly move us closer to fixing our healthcare system.

My Ideas Versus Congress’ Ideas to Transition to Alternative Payment Models

Next week I’ll finish looking at the ACP’s papers on fixing healthcare. But this week, I wanted to say one more thing about alternative payment models (APMs)–how to transition to them.

First, let’s think about this task without using the Healthcare Incentives Framework. How would a policy maker go about trying to get our healthcare system to shift from FFS to APMs (ignoring the fact that this is the wrong way to look at it)?

They would probably start by saying, “We need to find a way to give incentives to providers and payers to try out these different APMs.” This would be fairly easy to do through Medicare, so they would create some Medicare APM programs and structure them in a way that makes the benefits of joining large enough that lots of providers will want to participate.

They would also encourage private insurer-provider diads to start using APMs. How could they do that? They would probably just have to offer them money to do so.

These two approaches are what we’ve seen policy makers do. Medicare has the Medicare Shared Savings Program to get providers to enter into ACO contracts with Medicare, and the Bundled Payments for Care Improvement initiative to get providers to enter into bundled payment contracts with Medicare.

And to encourage private insurer-provider diads to start using APMs as well, I am only aware of two initiatives:

  1. Medicare joined up with a bunch of private payers to institute a patient-centered medical home program called Comprehensive Primary Care Plus.
  2. In 2015, congress passed the Medicare Access and CHIP Reauthorization Act (MACRA). A major part of this law is something called the Merit-based Incentive Payment System (MIPS). MIPS applies to all providers receiving reimbursements from Medicare, and it says they will now get a bonus or penalty based on a few criteria (quality, cost, EHR use, and quality improvement efforts) UNLESS they are actively participating in enough APMs (including qualifying programs with private payers!), in which case they are exempted from MIPS bonuses/penalties and instead they get an automatic 5% bonus.

I don’t think congress has passed any direct-to-private-insurer incentives to create APM programs (unless you count the Comprehensive Primary Care Plus program), but congress is influencing private insurers indirectly through providers because providers who want to get the 5% bonus and be exempted from MIPS will be pressuring private insurers to sign APM contracts with them.

So there we see the evidence of how policy makers not using the Healthcare Incentives Framework are approaching this effort to shift to APMs.

What would I do, knowing the principles illustrated in the Healthcare Incentives Framework?

First, I would discard the assumption that, for APM usage to increase, artificial incentives need to be created. The Healthcare Incentives Framework makes it clear that if an APM could truly offer increased value to patients, it would naturally arise in the market IF there are no barriers to it doing so. Next, I would go about looking for barriers and eliminate them. And only after doing that, if I want to accelerate the uptake of APMs, I could also offer artificial incentives.

APMs are a contract between an insurer and a provider, so let’s look at both parties.

Providers: Their incentive is to provide care of all kinds–this is how they make money. And any investment that enables them to raise their value relative to their competitors and that is not too risky will be desirable to them AS LONG AS they can be reasonably assured that patients will be able to identify their value as being higher than competitors’ and also have an incentive to choose the highest-value option. So, from a provider perspective, as long as any APM contract being offered by an insurer meets those requirements, they will be happy to participate.

Insurers: Their incentive is to minimize the total cost of care because they are getting a fixed amount of money in premiums, so any expenditure that is prevented is money that stays in their pocket (assuming those frustrating medical loss ratio rules instituted by the Affordable Care Act don’t come into play). The problem is, insurers don’t have much control over the total cost of care. Sure, they can try to negotiate the lowest prices possible, but providers are the ones that largely determine the total cost of care because they are the ones with the ability to prevent care episodes and to determine how much care is needed for care episodes that cannot be prevented. What I am saying is that insurers have the incentive to reduce the total cost of care, but providers are the ones able to make that happen. Therefore, insurers need to pass along their incentives to providers with these APM contracts. And insurers are happy to give money to providers to institute and run these programs if they can reasonably expect to save a lot more money than what they are giving.

With all that as context, what would I do to transition our healthcare system to APMs?

First, I would make sure providers are willing to join APM contracts by reasonably assuring them that if their investments into the program successfully increase their value, they will win more market share (and, therefore, profit).* How can I do that? By enabling patients to identify the highest-value provider up front and also ensuring that they actually have an incentive to choose the highest-value provider. I won’t go into details here on how to accomplish those things because I’ve written extensively about them before. But the result of those changes is that it would make providers see APMs as a potential for being very beneficial not only to their patients but also their profitability, which would probably result in them taking the lead in designing many of the APMs since they’re the ones who know best what changes could make a difference.

Next, I would make sure insurers are also willing to sign on to these APM contracts. Since insurers don’t like investing a lot of money into a program and then being required to give away all the financial benefits of that investment, I would eliminate the Affordable Care Act’s medical loss ratio requirements.

Next, lest you worry that insurers will forever keep all the savings generated by these APM contracts, I would enhance the ability for patients to compare the cost and quality of different insurance plans. That way, insurers will want to lower premiums because they will be assured that patients shopping for insurance will see that they are offering higher value (particularly in the form of lower premiums), so more patients will choose them, thus raising their profit as a result of increased market share.

Next, I would probably continue the programs Medicare is already doing, but I would also allow Medicare to sign on to other APM contracts happening between private insurers and providers. This would enable providers to get all insurers to reimburse them using the same contract, which would give them uniform incentives and make a huge difference in how much they are able to optimize toward that program.

Finally, if I do all that and am still unsatisfied with how fast this shift to APMs is happening, I would offer bonuses paid through Medicare to providers who are working hard to implement APMs (like the 5% bonus Medicare already implemented, described above). And if all that fails at getting this shift to happen as fast as I want, I would consider also offering grants to providers and insurers to try out APMs so that they don’t have to risk their own money designing and implementing them.

This approach is very different than the one currently being used by policy makers, and it would require changes that might be more difficult to make, but it would also not be limited by our current understanding of the “best” APMs. Instead, it would create the right environment for our healthcare system to continually shift toward better and better payment models as they are invented and refined.

* I know that an increase in market share does not automatically increase profitability, but this is my shorthand way of saying that it will increase their market power, which leads to increase profit either through a low-margin high-volume pricing strategy or through a high-margin low-volume pricing strategy. And my prediction is that, given where prices in healthcare are these days, the vast majority of providers would find that the profit-maximizing pricing strategy would be a lower-margin higher-volume option inasmuch as provider capacity allows.

Building a Healthcare System from Scratch, Part 8: Fixing the American Healthcare System

Image credit:

This is the final installment of the Building a Healthcare System from Scratch series, and it begins with the same caveat as Part 7—to understand the rationale behind the ideas presented, you need to read and understand all the prior parts of the series, beginning at Part 1.

Now that I have explained the Healthcare Incentives Framework (Parts 1 through 6) and described what various types of systems would look like that have implemented the principles of it (Part 7), we are ready to look at something more complex: rather than building an optimal system from scratch, how can this framework be applied to an existing system that needs fixing? The American healthcare system will be a perfect case study for this.Picture1

Where it sits right now, the American healthcare system seems to land somewhere in between the libertarian-type system and the single-payer system described in Part 7, except that it fails to implement the majority of the principles of the framework. It is also overly complex, which contributes greatly to its impressive administrative expenses. So what could a much-simplified American healthcare system look like that also maximally implements the principles of this framework? With consideration for and much guessing about Americans’ preferences regarding healthcare, here is an imagined description . . .

The United States eventually chose to strengthen its individual mandate. The government now mandates all people, without exception, to have a healthcare insurance plan that covers all services included on its list of “essential health benefits.” For those who forego insurance coverage, they pay a tax penalty that ends up being nearly the same as the premiums they would have paid. If uninsured individuals receive care that they cannot afford, they usually end up having to either go on long-term repayment plans or declare bankruptcy because there is no bailout available for them.

The website has become the ultimate source for healthcare insurance shopping. All qualified insurance plans (i.e., plans that cover all the essential health benefits) are listed there, along with their coverage level (according to the metal tiers), prices, and a short description that highlights any other benefits the plan offers that may help prevent healthcare expenditures.

Premium subsidies are available to all whose premiums will exceed a certain percentage of their income, and the subsidy amount is pegged to the second-cheapest qualified insurance plan available to them. The subsidies are automatically applied at the time people are choosing a plan on Because this system worked so well and was basically duplicative of Medicare and Medicaid, both programs were slowly phased out, which decreased insurance churn and increased time horizons. However, a vestige of Medicaid remains in that, depending on an individual’s annual income, there are also limits on how much they can be required to pay out of pocket for care.

The government also did away with the employer mandate and somehow found the political willpower to repeal all tax breaks for healthcare expenditures, which eventually led to employers getting out of the business of providing healthcare insurance for their employees and just giving that money to employees directly as regular pay.

Altogether, these changes mean that all Americans shop for their healthcare insurance on For anyone who is unable to do this themselves, there is a phone number to call to connect with someone who can assist them in selecting the plan that seems best for them.

A few changes were also made to encourage more insurance options. First, many regulations were eliminated, including the medical loss ratio requirement and state insurance department approvals of rates. These became unnecessary after people began to be able to compare the value of different insurance plans and choose based on that because overpriced or low-value plans (either from too much overhead or too high of rates in general) lost market share and profit. State-specific insurance regulations were standardized so that insurers can easily expand to new markets. And a law was passed requiring transparency of all price agreements between providers and insurers, which made the process of forming contracts with providers in a new region easier. These policies led to almost all markets having multiple options for each coverage tier.

In this way, the United States achieved universal access to affordable healthcare insurance relying on the private market while preserving the ability and incentive for all people to select the highest-value insurance plan for them. As a result, insurance plans aggressively innovate to find ways to prevent care episodes so that they can offer lower premiums and attract higher market share. Insurers also found that implementing differential cost-sharing requirements led people to start choosing lower-priced providers, which also enabled the insurer to lower premiums further. Finding provider prices has become easier ever since the price transparency law was passed.

The government has had to help overcome the problems caused by a multi-payer system by enforcing some standardization, including uniform insurance forms/processes, standardized bundles of care that all insurers in a region either agree or disagree to implement together, and standardized quality metrics that providers are required to report. These quality metrics are not used for bonuses, so they have been changed to be more focused on what patients need to know to choose between providers for specific services.

These quality metrics are now reported on an additional section on that lists all providers, their quality metrics, and their prices (seen as your expected out-of-pocket cost if you log in) in an easy-to-compare format. Due to patients’ differential cost sharing requirements for most services, as well as broad common knowledge of the existence and utility of this website, most patients have begun to refer to it before choosing providers. This part of has even been developed into a highly rated smartphone app.

In response to these changes, providers found that their value relative to competitors largely determined their market share and profitability, which unleashed an unsurpassed degree of value-improving innovation. The cost of care in the United States was previously so high that the majority of those initial innovations led to much cheaper care, which led to much lower insurance premiums and eased the premium subsidy burden on the federal government. Thanks to these changes, the federal deficit has begun to sustainably diminish quicker than any budgetary forecasting model could have predicted, which has also helped stabilize the American economy.

There are still barriers to people being able to identify and choose the highest-value insurers and providers. There are many important aspects of quality that are unmeasurable. Many people do not have the health literacy required to figure out which insurance plan or provider would be best for their situation and preferences, despite the ease of comparison enabled by There are medical emergencies that do not allow shopping (although the number of these has turned out to be much less than was previously thought because most of what used to present to emergency departments were not actually emergencies).

In spite of these lingering barriers, enough patients are choosing the higher-value options that providers and insurers still have a strong incentive to innovate to improve their value so they can win the market share and profit rewards available to higher-value competitors. And the result is that Americans are being kept healthy more often and are receiving care that is higher quality and more affordable.

That concludes my imagined description of how the American healthcare system could look with the principles from the Healthcare Incentives Framework fully applied. Just as a reminder, it represents merely a guess of how it could end up given the current system. This is by no means the only way to apply the principles of this framework, nor is it my secret idealized version of how it could end up. But I hope it was useful and thought provoking as a case study!

We have come a long way in this series. Throughout it, I worked hard to make the principles of the Healthcare Incentives Framework clear, and I hope the concrete examples have helped solidify those as well as demonstrate their potential for sustainably fixing healthcare systems around the world. If you want a more academic treatment on this framework (at least the part of it that applies specifically to providers), I published that as a medical student.

I intend to help policy makers of all types find ways to apply the principles discussed in this series. Please contact me if you have questions or would like me to help work through potential applications. Contact info is on my About Me and This Blog page. In the meantime, I hope you will follow along as I continue to blog about how to fix our healthcare system!

The 3 (Actually, 2) Problems with the U.S. Healthcare System

Image credit: Gary Larson

Before something can be fixed, the problem must be defined and the causes of that problem diagnosed.

The generally accepted problems with the U.S. healthcare system are that its prices are too high, its quality is too low/patchy, and not enough people have insurance coverage.

I’m going to change that a little bit. Since price and quality are the two variables that determine value, we could combine them into only two problems: suboptimal value and not enough insurance coverage.

The insurance coverage piece is kind of separate from the value piece because solving it primarily relies on government policies that subsidize the purchase of health insurance for those who wouldn’t be able to afford it otherwise. Those government policies are a (purposeful) distortion to the healthcare market. Distorting a market for a good reason is fine, but before you do it, you need to understand how the market should be structured to optimize value so you don’t accidentally ruin its potential to have higher value in the process.

And the nice thing is that value improvements will primarily come in the form of lower prices —> insurance coverage will become cheaper —> fewer people will be priced out of the market. Therefore, fixing value partially fixes the insurance coverage problem too!

If We Lower Total Healthcare Spending, Who Will the Money Come from?

Image credit:
Image credit:

At the recent AcademyHealth Annual Research Meeting in Baltimore, I went to a session on the accomplishments and challenges of community collaboratives. A community collaborative is a pretty cool idea that goes something like this: for a specific community (i.e., city), let’s get all the leaders of the providers and payers in a room (plus a bunch of other stakeholders committed to improving health) and make some decisions collaboratively on how we can fix healthcare in the community. The Robert Wood Johnson Foundation has provided the money to make these things happen in 20 different communities in the U.S. (see Aligning Forces for Quality, and Value-Based Payment Reform).

Sounds like a great idea, right? Well, an interesting challenge has arisen. More and more, these collaboratives are expected to find ways to reduce the total healthcare spending in that community. But so far, they’ve pretty much failed miserably. Why? Well, think about it. Here are all the leaders of payers and providers in that community sitting in a room together saying, “We need to reduce total spending,” but the savings are going to have to come from someone in that room, and none of them are going to say, “Sure, my organization will take one for the team! I’ll have to cut everyone’s pay, but because it’s for a good cause, they’ll love it.”

Does this mean these kinds of collaboratives are utterly useless in terms of lowering total spending in communities? That was the question I (carefully) asked at the end of the session, and one of the panelists gave a really insightful answer. To paraphrase/translate/elaborate on what he said, his answer went something like this:

Yeah, we’re not going to convince anyone in that room to just give up money like that. But what we can do is come up with standardized ways of reporting prices and quality. And when those are standardized across all payers and providers, patients will be better equipped to choose higher-value payers/providers, which, in the U.S., usually means ‘cheaper’ payers/providers. So this standardization will allow total spending to go down by getting more people receiving services from cheaper competitors. Thus, the higher-priced competitors will be the ones who are losing money when total spending goes down, all because we helped standardize quality and price reporting.

I agree. There are still many barriers to getting patients to choose these “higher-value” providers/payers, but this would help solve one of the biggest ones. And with each barrier we overcome, more patients will be enabled to receive higher-value care, which is what everyone wants, right?

Should We Regulate Prices of Hospitals? All-payer Rate Setting’s Allure

Image credit:

The Bitter Pill article has received a lot of press lately. People reading it have often turned to a simple solution: regulate prices. The most straightforward approach to this is called “all-payer rate setting,” which has been experimented with before in some places in the U.S. and is still used in Maryland. The basic idea is that the government says, “When any provider performs this certain service, he/she will be paid this much for it no matter who the payer is.” And they set prices for every single service. Think of how this would instantly make all chargemasters a thing of the past. And no more worrying about hospitals increasing their bargaining power as they join together to form ACOs. And all that administrative complexity that would be gone (thus decreasing costs a fair amount)!

But there are downsides, too, which are not as obvious and may lead people to jump on the bandwagon of all-payer rate setting ignorantly. First, back to basics:

Total spending on healthcare = price * quantity

Yes, we probably have some quantity problems (running too many scans, etc., which regional variation literature attests to quite thoroughly), but the main reason we spend so much more than other countries is because of the prices. So, here’s the prices equation:

Price = Cost + Profit

What’s making prices too high? Brill makes a strong case that, at least in a lot of hospitals, profit is part of the problem [Update: Turns out most hospitals lose money on average, so it’s not as big of a deal as we thought]. But what about costs? Is the actual cost of care too high as well? YES, costs are the major problem, as shown by looking at the average profitability of healthcare organizations. More evidence of this: even in countries that do a pretty good job minimizing unnecessary services and regulating profits to reasonable levels, healthcare spending growth is still unsustainable, which only leaves cost as the primary culprit. Therefore, any policy (whether it’s meant to regulate profits, improve access, improve quality, or whatever) that creates barriers to cost lowering should be reserved as a last resort.

So, would all-payer rate setting create a barrier to cost lowering? If yes, I don’t like it. If no, let’s consider it.

First, since I’m reading The Wealth of Nations lately, let’s ask Adam Smith what he thinks about the subject:

I shall conclude this long chapter with observing, that though anciently it was usual to rate wages, first by general laws extending over the whole kingdom, and afterwards by particular orders of the justices of peace in every particular county, both these practices have now gone entirely into disuse.

By the experience of above four hundred years [says Doctor Burn] it seems time to lay aside all endeavours to bring under strict regulations, what in its own nature seems incapable of minute limitation: for if all persons in the same kind of work were to receive equal wages, there would be no emulation, and no room left for industry or ingenuity.

Particular acts of parliament, however, still attempt sometimes to regulate wages in particular trades and in particular places. (Emphasis added)

What’s he trying to say? All-payer rate setting would leave “no room left for [cost-lowering] industry and ingenuity”? (If you’d like to see my explanation for why I assume innovations by providers are generally cost-lowering, see here.)

I’ve explained before how taking away the freedom to set your own prices also removes much of the rewards for cost-lowering industry and ingenuity. In short (and simplified), lowering costs without sacrificing quality means you can lower prices more than others and therefore offer higher value than others, and higher value will eventually be rewarded with market share and profits. (Another assumption I’m making: patients preferentially choose higher-value providers, which is starting to be more true, but there are still many barriers to it.)

Back to the big picture: All-payer rate setting reduces the potential rewards for cost-lowering innovations, which I can guarantee will reduce the amount of cost-lowering innovation that goes on. So, yes, all-payer rate setting will be a barrier to cost-lowering innovation. And that’s a huge problem, so let’s look for other ways to fix egregious profits and costs. More to come . . .