Relative to our February 2010 projections under prior law, average annual growth in national health spending over the projection period is estimated to be 0.2 percentage point higher than our previous estimate. The health care share of gross domestic product (GDP) is expected to be 0.3 percentage point higher in 2019.
The problem, some have argued, is that reform is supposed to be reducing costs and spending on heath care, not pushing spending up, if only by a small amount. To borrow an exceedingly overused expression from health policy circles, reform ought to "bend the cost curve" downwards, meaning it should stop the runaway health spending that's characterized our economy for decades. At the very beginning of the legislative debate over health reform last year, Gallup released a poll examining public attitudes. As you might imagine, the relative importance of a person's concern about health costs or expanding coverage (i.e. cutting down the number of uninsured) depended on whether or not they had insurance:
Most people do have insurance and you can see that most of them are more worried about their costs than helping the uninsured, which makes sense (though they also tend to fear losing their own coverage).
So we're back at a common complaint: "this law does nothing to control costs, it focuses entirely on coverage." As this is a pretty important allegation, it's worth thinking about in detail. But we need to take a policy-philosophical detour first.
Below you'll see a crudely constructed figure showing what's sometimes called the health policy triumvirate: cost, quality, and access. These are the three key factors that define our health care system. In physics, there's something call the "no-hair theorem" that tells us black holes are completely described by only three physical characteristics: mass, electric charge, and angular momentum. I'll go out on a limb and suggest a health care system is like a black hole in some ways (particularly when it comes to money). We might as well propose an analogue of the no-hair theorem for health care systems: three characteristics--cost, quality, and access--will characterize your system completely.
You can see that, at the risk of possibly violating that principle, I've gone a step further in that graphic--I've started making secondary colors (you know, figuratively). The way the various tips of the triangle interact are themselves noteworthy. If you've got a high-quality health system with widespread access to its facilities, you've probably got pretty good population health, i.e. people in your nation are, on average, pretty healthy. If you have a high-quality system at a relatively low cost, you're getting good value for your health care dollar. The U.S. notoriously has very good care, though still comparable to other advanced nations along many metrics, but pays much more than anyone else to get it. So it's possible to have great care but get relatively little value for your health care dollar. And finally, if you've got a low-cost system in which access to its facilities is widespread, you've got what I'll call equity. It's possible to cut costs by cutting access but that doesn't make for a very equitable system.
Historically, the conventional wisdom about the cost-quality-access triangle has boiled down to two words: "pick two." You can cover everyone in a high-quality system, but it will be prohibitively expensive. Or you can have a low-cost, high-quality system but it won't be very equitable--not everyone gets to play. Or you can have a system with universal access at a low-cost but the care won't be any good.
Somewhere along the line that paradigm has shifted. I haven't been around long enough to know when that began but I do know the date the paradigm shift was enshrined in law: March 23, 2010. The philosophy underlying the health reform law is that we can have it all. Some, no doubt, would dismiss that as naive and mere wishful thinking. But it's a powerful thought.
Contrast that with the thinking underlying H.R. 4038, the "replace" in the Republican repeal-and-replace proposal. I don't want to sell it short--there is, after all, a decent amount of stuff in there--but for me it basically boils down to two things: 1) deregulation and 2) high-risk pools. The form of the deregulation is specific to health insurance and it's something we've looked at on here before: selling insurance across state lines (they literally just lifted the language for that section of their bill out of Shadegg's Health Choice Act). A complaint you'll often hear from the right is that we have too many mandates, insurers are required to provide too many benefits. At the same time, you'll sometimes hear that if only insurance were less expensive, coverage would increase because more people could afford to buy it.
But the limits of that possibility become clear when you realize that the plan is to lower costs by limiting access. What's the underlying principle here? Namely, that less is less: if we find a way to make it easier for insurers to jettison expensive customers and shed benefits (i.e. cover less), insurance will get cheaper. Get rid of consumer protections, sell potentially inferior products, and the products get cheaper. It's not wrong--the CBO score of their bill did find that it will lower average premiums in the individual and small group market. But it won't cover any more people; even with the high-risk pools they include to help the uninsurable gain access to coverage their bill leaves the percentage of the population without heath insurance unchanged. And I suspect it wouldn't be particularly beneficial to population health. But as I said, it's not wrong--that is a way to cut costs, I'm just not attracted to it philosophically. And, at a pragmatic level, I believe there are better options for getting better results. We need not take the Khan Noonien Singh approach, dismissing the uninsured with a helpless shrug and leaving them marooned for all eternity in the center of a dead planet, buried alive.
Most people agree that when it comes to our health care system we have too much: too much fraud, too much unnecessary treatment, too much inefficiency, too many preventable expenses. And that means we're going to have to cut if we're going to contain costs. So let's jump into the two broad philosophies of cost containment--what I'll call the Scalpel and the the Chainsaw--and then we can finally get to the punchline: how health reform promises to contain costs.
The Chainsaw
The Chainsaw is a blunt instrument. It accepts only that we've got to cut and it concedes that it's going to be bloody and imprecise. And, just maybe, employing the Chainsaw won't be good for the health of people in its path. Folks on the right tend to favor the Chainsaw when it comes to reeling in costs. We touched on this a bit above but let's look at exemplary statements from prominent conservatives and bloggers,
Several months ago, Becker argued:
The most important needed reform is an increase the fraction of total medical costs that come from out-of pocket expenses in the form of large deductibles and significant co-payments. [...]
The US health care market is over-regulated rather than under-regulated. One example is that families in one state are generally not allowed to buy their health insurance from companies located in other states. Another example is the mandates that states impose on insurance companies, such as coverage of the costs of normal birth deliveries. Such coverage has little to do with insurance against unexpected health costs, whereas coverage of extraordinary delivery costs is a desirable protection against unexpected health care risks.
That second bit is the argument Republicans make: we need fewer consumer protections and less comprehensive insurance. The former is interesting, philosophically, because it assumes people just want extra care because it's cheap--or, rather, that pricing people out of care is the best way to allocate it. Or as blogmate Posner suggests, it's like going to a buffet!
A second demand-related cost effect will result from the fact that insurance,(even with deductibles and copayments, drives a wedge between the cost of a service and its price, and so increases demand. (It’s like a restaurant with a buffet: the marginal cost of eating all you want is zero.) Persons who are uninsured are deterred from consuming medical services in quantity— because of cost (they are billed for such services at very high prices and may be forced into bankruptcy if unable to pay), because of difficulty of obtaining quality service from charity hospitals or other “free” providers, or simply because, though they can “afford” insurance, they prefer to gamble on remaining healthy. These persons, when they become insured, will increase their utilization of medical services, because those services will now be cheaper to them.
I kid, of course, they're making valid points here. But I fundamentally disagree with the premise. Certainly there is an enormous amount of unnecessary care provided in our system (Thomson Reuters pegs the number at $250-$325 billion annually--that's up to 13% of all the money we spend on health care in a year). And I agree that in the absence of higher cost-sharing (an incarnation of the Chainsaw), people will tend to consume more care than they would if they were in the Chainsaw's path. But health care isn't equivalent to mashed potatoes in a buffet line. Gorging at a buffet is enjoyable, for a time anyway, and has no immediate costs (you have to wait until a bit after dinner to encounter the costs of your excess). Receiving care has many costs, be they in terms of time, anxiety, discomfort, and so on. If someone asked me if I'd rather spend an all-expenses-paid evening in a hospital or at a buffet, there'd be no contest. One is pleasurable; one is horrifying.
The reality is that people don't seek excess care because they just love soaking up additional health care resources and low cost-sharing allows them to indulge that default urge; rather, they don't know any better. Which if how you get incoherent responses like these in polls:
Let that sink in for a moment. Roughly half the respondents in this poll recognize that overtreatment--unnecessary care--is a problem in our system (though only 16% responded that they personally had gotten a test or treatment that was probably unnecessary). Simultaneously, 67% believe undertreatment is a serious problem. I suppose you could try and spin that as a recognition or expression of concern about inequality--perhaps some respondents believed lots of people are being overtreated and lots of other people, on some lower tier, are being undertreated. But it seems more likely to me that this is an indication that we, the consumers, don't know what unnecessary care is, we wouldn't recognize it if it was pounding on our chest and defibrillating us. We fear that someday we'll be denied the treatment we need (only 14% of people actually thought this had already happened to them), even as we recognize that other people are probably getting care they don't need. We don't want unnecessary care in the same way we want that unnecessary third helping of prime rib at the buffet, we simply don't know it's unnecessary.
Which is exactly why price signals alone, coupled with individual budget decisions, aren't the best way to cut through this mess. The health care system/market is not even close to being transparent, science-based, and patient-centered enough to allow people to make informed decisions about the best balance between their pocketbooks and their personal health. But the Chainsaw approach doesn't particularly care. More cost-sharing means we've placed a financial barrier in people's way that will discourage them from utilizing care--the Chainsaw lops off that spending and in the aggregate cost growth slows. But note that this individual decision isn't rooted in evidence-based, medical decision making. It's a blunt instrument that invariably cuts more closely the further down the income ladder you travel.
The Chainsaw isn't ideal for equity or population health concerns, but it probably will put the brakes on rising costs. And two of the primary long-term cost controls in the new reform law are forms of the Chainsaw: the excise tax on high-cost group insurance plans and the creation of an Independent Payment Advisory Board for Medicare. The excise tax goes into effect in 2018 and the threshold at which it kicks in will grow with the rate of inflation; however, since medical costs tend to rise faster than that, more and more plans will be affected by it with time. The idea is that it will cause employers to keep plan costs down (i.e. buy less expensive insurance plans) to avoid the tax and shift some of the costs to employees, something that's increasingly happening right now. The IPAB will be empowered to find ways to cut costs from Medicare, most likely by changing reimbursement rates. Luckily, these will be combined with cost control attempts with a bit more finesse.
The Scalpel
While the Chainsaw approach cuts costs in the goriest and most undiscerning manner possible ("less! less! damn the population health consequences!"), the Scalpel approach relies on more delicate instruments. Therein lies the fundamental philosophical shift that reform is institutionalizing: we can meet our cost containment goals by providing smarter and better care. Reform is partnering with efforts that began in last year's economic stimulus package to fundamentally shift our approach to health care. Right now it is too much like a buffet: a disconnected, uncoordinated smorgasboard of
But I digress. And the food analogy becomes incoherent remarkably quickly. Let's go back to the stimulus, because the nexus between the stimulus and the health reform law is hugely important (disclosure: I'm biased--that nexus is primarily where I live now, professionally). The stimulus made two health-related investments that stand to be very significant in the long-run: 1) comparative effectiveness research, and 2) health information exchange. The former refers to building a strong, scientific evidence base to guide clinical decision-making and help us understand what works best for who in which circumstances (notably, one-size-fits-all doesn't necessarily work, which is why understanding of this "treatment response heterogeneity" is important). The latter refers to not only storing health data electronically--on electronic health records--but having the means to transmit and use that data. I'll be writing about that in detail at some point because it's near and dear to my heart. But the moral here is that the stimulus set on us on the path toward embracing data-driven health system improvement. More science, less art. After all, medical knowledge with predictive power is an impressive thing to witness:
The reform law doubles down on comparative effectiveness research, creating a Patient-Centered Outcomes Research Institute to fund and consolidate this research, and it gets specific on ways to improve the way health care is delivered. It tackles the primary incentive in our system for unnecessary care--a payment structure that pays per service, prioritizing quantity over quality--by experimenting with payment schemes that reward quality and limit hospital readmissions. It supports relatively new models of coordinated care, known as medical homes and accountable care organizations. Bundled payment initiatives will further being launched to encourage providers to coordinate care and deliver it in a cost effective manner.
The essential thing to understand here is that while there are dozens of initiatives in the law aimed at determining what works to lower costs and improve quality, they are experiments. That is, they're generally pilot (demonstration) projects operating on some limited scale (e.g. in a handful of states), though, to be clear, there are some things in there that aren't merely demonstrations. These changes, demonstrations or not, occur within the programs over which the government has primary authority: Medicare and Medicaid. But, excitingly, the reform law establishes a Center for Medicare and Medicaid Innovation that will be empowered to rapidly roll out successful experiments on a wider scale within Medicare and Medicaid (the assumption is that successful models will also spread to use in the private sector). We don't know exactly which mechanisms are going to work best, which is why the law takes an FDR-esque philosophy: "It is common sense to take a method and try it; if it fails, admit it frankly and try another. But above all, try something."
The shift here--and this goes back to the point I was making about the increasing emphasis on data-driven improvement--stands to be pretty fundamental: we're stumbling toward a future in which health is priced as an output good. Right now, we generally don't pay for health, we pay for procedures; if they don't work and you end up needing yet another procedure, so much the better for the provider accepting your payments (that's one of the volume-based incentives that drives up unnecessary care). One would think that this sort of shift is a necessary pre-condition for treating the health care market as a well-functioning market, as conservatives like Messrs Becker and Posner might like. But before that can work, we need some big changes in place. We need to start pricing health as an output good, we need transparent indicators (like reliable and accurate quality reporting) to aid shoppers, and we need readily available information on clinical effectiveness to better inform doctor and patient-as-customer alike when it comes to making medical decisions.
And those changes require laying down the appropriate infrastructure. If we're going to pay providers based on what works, we need to know what "working" is, we need the means to evaluate whether the criteria for that (whatever they may be) have been met, and we need doctors themselves to know what's most likely to work for a given patient in a given situation. And that requires an incredible capacity for processing information, whether that be collecting quality indicators from physicians via electronic health records or evaluating the comparative effectiveness of various treatments. But we're about to set off down that path. Exciting. But this law is the beginning of the journey, not the end. We'll need to continually tinker with it, building on its successes and frankly evaluating its failures, tweaking it, adding to it. Vigilance is the price of long-term cost containment.
The Demi-Decade of Coverage and the Long March Toward Cost Containment
And so, at long last, we arrive at the heart of the matter: are my premiums going to go down, are costs contained? As I intimated above, the reform law is extremely ambitious. It goes after the trifecta on the notion that cost, quality, and access can all be addressed without the need to sacrifice one. Access--which here I'm somewhat imprecisely using as a synonym for insurance coverage--is often the odd man out in our country but it's also the low-hanging fruit. Coverage is easier to address than are cost and quality, particularly when the primary strategy for reigning in long-term costs is based on improving the quality of our woefully inefficient system. And in the first decade of the reform law's existence--or, really, the first roughly half-decade, since the law doesn't fully kick in fully until 2014 and we tend to only think as far as 2019--will address coverage in a big way. This period is what I'd call the Demi-Decade of Coverage, because its most visible achievement will be insuring another 30-odd million people, fully ten percent of the American population. There will still be uninsured folks at the end of the decade--some by choice--but substantially fewer than there are today.
Meanwhile, the tinkering will begin. The Chainsaw of the excise tax will begin revving up at the end of the decade (2018) and the Scalpels of various payment and delivery system reforms aimed at constructing a smarter, leaner, more coordinated health system will make their first incisions over the next few years. It will take time for these programs to operate and be evaluated for effectiveness. The stimulus programs for encouraging adoption and meaningful use of electronic health records (i.e. health information) won't wrap up until 2021. All of this means that we can't really expect real cost curve bending to happen until the second decade of the law's existence (the 2020s). The primary cost control items operating over the next few years will be the insurance market reforms--which get all the press anyway--like the construction of the health insurance exchanges mentioned in my last post, which bring transparency and cost competition into state marketplaces. But those gains will be somewhat offset by the influx of newly insured people and the increasing quality of insurance plans available in the marketplace (read: benefit mandates). Which is exactly what the paper from the CMS actuaries that I opened this post with is saying:
For 2015–19, national health spending is now projected to increase 6.7 percent per year, on average—slightly less than the 6.8 percent average annual growth rate projected in February 2010. However, the year-by-year pattern of growth is anticipated to be different (Exhibit 5). Enrollment shifts associated with the Affordable Care Act coverage expansions are projected to continue, contributing to continuing relatively faster spending growth rates through 2016. Thereafter, spending growth is projected to decelerate more substantially as a result of Affordable Care Act–mandated reductions to Medicare provider payment updates and the excise tax on high-cost insurance plans starting in 2018.
The Demi-Decade of Coverage won't see total national spending change substantially, which is actually pretty impressive given the number of people who will be entering the system. Premiums will continue to rise faster than inflation year after year. It won't look like we've done much, except provide care to more people, which in some corners is considered laudable all by itself. But the Scalpels will be behind the scenes, finding the most precise ways to cut. And then the Chainsaw will begin at the end of the decade, giving even more urgency to the mission of the Scalpels--finding ways to ensure that the cuts the Chainsaw will force are targeted, smart, and don't harm population health. Frankly, I'm excited about these two very different cutting tools working together. I love the idea of the Scalpels and I believe that in the long run a number of them will be found to work pretty well, transforming our system into something much better in the process. But I also like that the threat of a buzzing Chainsaw will be lighting a fire under the Scalpels' asses.
But in the end, we'll have to have patient. Going for the trifecta is inherently a long-term process. We could lower insurance costs tomorrow by throwing as many sick people out of insurance pools as we can but I don't think that fits with the vision of what our health care system should--and can--be that many of us have. Too much Chainsaw, not enough Scalpel. So hang in there, the journey's just beginning. And if you've actually read all the way down to this point, I know you do have patience. This much-too-long post is now mercifully at an end.
I borrowed some of your graphics for a class presentation on health economics paradigm shifts. this was a great post!! thanks for the post and the graphics. :)
ReplyDeleteThanks.
Delete