Testing Health-Care Providers' Threshold for Pain
Regulators have been playing a vast and costly game of whack-a-mole since the creation of Medicare and Medicaid 40 years ago. Regulators decide we’re spending too much on something, and reduce or disallow that spending. That budget line item duly drops, and yet health-care spending does not, because some other category has risen to compensate. Cut hospital spending, and doctors’ bills rise; disallow hospital readmissions, and hospitals start refusing to admit patients, instead putting them on “observation” status so that another crisis will not result in an unbillable “readmission.” This dynamic is a source of great frustration to health-care wonks, who are forever coming up with new policies to combat it. These methods are announced with great fanfare, rolled out … and gee, how come we’re still spending so much money on health care?
The latest hotness in cost control is called “all payer.” The idea is that without such a system, hospitals and doctors exploit pricing disparities between various categories of payer, with governments generally paying less and private insurers paying more. This causes various sorts of problems. 1 It also, people argue, makes it harder for anyone to get health-care prices under control. You cannot jam Medicare and Medicaid rates down to where technocrats would like them to be, because there’s still a lot of private insurance around, which means that if you set the government rates too low, doctors and hospitals have the option of saying “Thanks but no thanks, we’re not going to participate in your program any more.” Private insurers have similar problems: As long as there are other options in the market, the insurers' negotiating leverage with providers is limited.
Enter all-payer. Basically, the idea is that everyone gets charged similar rates, so providers cannot play insurers off against each other to extract higher payments. You band everyone together into one giant negotiating bloc that sets rates for procedures, rather than letting each payer and each provider set their own fees. In addition to reducing the negotiating leverage of providers, it could theoretically lower administrative costs, because you’re now only dealing with one set of prices.
The idea grew popular in the 1970s, but was largely abandoned, because the regulations required to make it work were so complex that they were, as one Massachusetts state official put it, “like Sanskrit -- no one could understand them, even the hospital people.” Of all the states that tried it, only Maryland’s system is still in operation.
Until now, that is. Vermont, having tried and failed to build a single-payer system, is now prepared to embark on a new experiment with all-payer. Going to this system requires a waiver from the federal government, which in Vermont’s case, wanted assurances that total federal spending on the state wouldn’t rise. (The state, in turn, wanted assurances that it wouldn’t fall, either.) Both parties now seem to be satisfied, and the project seems likely to go ahead, with an initial demonstration phase to run from 2017 to 2022.
Will it work? Maybe. But there's that whole thing about the devil and the details.
And unfortunately, some of those details are in short supply. I have read everything I could get my hands on about the proposed system, including the draft agreement with the federal government and the FAQ and other documents provided by the board in charge of setting the rates. Alas, I still had questions, and despite multiple phone calls, was unable to reach anyone at that board who could clear them up.
So I preface what follows with the note that it’s not entirely clear to me how the payments are going to work, beyond the fact that they are to be set per person ("capitated") and insurers (including Medicare) will pay into a large organization known as an accountable care organization, which will in turn be responsible for disbursing payments to providers. How those payments will be calculated and disbursed, however, remains unclear to me at this writing, and of course that will have a great deal of impact on success or failure.
With that caveat, let’s move on to explore what Vermont is doing, starting with explaining what “capitation” is. Basically, it means that instead of paying providers by the treatment, you pay them by the patient or the case, and make providers responsible for treating patients as cost-effectively as they can. This is a new model for an all-payer system, and it might alleviate some of the old problems -- notably that you ended up with absurdly complicated price schedules, which providers tried to game. For example, some studies suggested 2 that while the cost of hospital admissions tended to fall, the number of hospital admissions tended to rise, so the overall savings were unclear.
But capitation is not a new idea, and it also turns out to have some fairly serious problems. It’s more difficult to game, but not impossible, because you generally have to make some adjustments for things like how old and sick the patient population is, and determining what those adjustments should be can be as much art as science. Providers may no longer be have incentives to do a more expensive surgery instead of a cheaper one, but they might have incentives to discover more expensive patients, by deciding that a patient with a borderline condition definitely has it. So there’s a risk that you don’t end the gaming, just switch what game you’re playing.
Too, while paying providers per patient, rather than per treatment, gets rid of old incentives for overtreatment, it introduces new ones for undertreatment. Vermont’s FAQ offers the cheery reassurance that they will get around this problem “through a variety of means, including patient surveys, access reports, and transparent measures of quality and performance at an institutional, practice, and provider level”. I take the more jaundiced view that if you give doctors strong incentives to undertreat, you’re going to get quite a bit of undertreatment.
Before you tell me I’m being too hard on the secular saints who make up our medical profession, I should note that these plans can be quite risky for providers, especially smaller providers. If you have a small caseload, then under many forms of capitation, one or two unexpectedly expensive patients can be the difference between operating in the black and drowning in a sea of red ink.
I can’t exactly blame providers for undertreating some patients if the alternative is closing up shop and treating no patients at all. It’s not clear to me from Vermont’s documentation what sort of structures will be in place to mitigate this risk, or how effective those might be. Much, of course, will rest in the setting of the rates.
But there’s reason to think that Vermont will err on the side of setting those rates low. The state has agreed to meet some pretty aggressive targets -- to hold overall cost growth down to 3.5 percent, rather than the 6.6 percent that is projected, and to keep its Medicare spending from growing faster than the national trend. In order to combat any incentives for undertreatment, the state has also agreed to health outcome targets. Whether it can meet all of those targets at once, without bankrupting its providers, is an open question.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
One of the theories advanced for why Obamacare policies are becoming so expensive, for example, is that providers who care for very sick, very expensive patients are helping patients on Medicaid, which offers abysmally low reimbursements, to buy insurance policies that will reimburse the providers at higher private rates.
Full disclosure: The author of the linked article is witty, charming and handsome, and happens to be married to me.
To contact the author of this story:
Megan McArdle at email@example.com
To contact the editor responsible for this story:
Philip Gray at firstname.lastname@example.org