Wednesday, July 29, 2009

Neither necessary nor unnecessary to whom? 

While I was thinking about QALYs last night, I remembered an excellent summary video of an Arnold Kling presentation to Cato from June. It's less than ten minutes but the second part of this makes a major point I think gets lost here. You'd do well to watch the whole thing. I can wait...

At one point he says roughly this: around age 50 men are usually told to get routine colonoscopies. They're not fun (given my health issues of the last four years, I've had three procedures, including one in hospital when I really was in no position to argue.) They're also not given in Canada; you only get a colonoscopy if you've got family history of colon cancer or some symptoms. In three of the four cases where they'd be recommended, waiting times exceed what gastroenterologists would recommend. (Lucky for hypothetical me, my symptom was the one that got you in the fastest.)

But actually catching colon cancer at the earliest stage, before the polyps turn cancerous, is best. You probably can wait, though, and maybe you will be fine -- there's likely to be other symptoms that present to tell you waiting time is over. So it isn't really a necessary procedure. Kling's analysis is that there's a middle ground between treatments that are necessary and those that are simply unnecessary and wasteful.

But nobody chooses to spend money on something they see as wasteful. If I tell you "you don't need that colonoscopy" you probably would go without it, because they're really no fun. But you accept it -- I accepted it twice in three years -- because there's a chance there's a polyp that has a chance to become cancerous. It's really in that middle category in the continuum that Kling lays out.

So for those, you really have to work the cost-benefit analysis. You weigh the probability that the colonoscopy will find something times the benefit of early detection, and put that against the cost. Guess what? Your health insurance provider has made that calculation too, which informs its guidelines of how often it will cover your colonoscopies. Since it may face the cost of non-detection in the form of higher insurance claims later, it has some incentive to get it right. (I will acknowledge the possibility that they don't pay for colonoscopies often enough and then drop your coverage later when you get colon cancer. It would seem we have laws about this already; if that's really the problem, you need to enforce the laws you have rather than radically change the health care system.)

Suppose we thought there were too many colonoscopies in the US. How could we cut that back? We could have people pay a higher share of the cost, and have that work through the cost-benefit analysis to reduce the quantity demanded. If it's really a case of neither necessary nor unnecessary, that seems reasonable.

But that, critics would say, means the rich get colonoscopies and the poor don't. That's unfair, so we should use a different rationing system than price. OK, but that means you move the cost-benefit analysis away from the patient and to government. And to do it, it has to take the cost of the treatment and compare it to the benefit as it sees it rather than as the patient sees it.

And that's where you get the QALY.

So what does it mean to use a QALY calculation for a treatment that is neither necessary nor unnecessary? The political economy of that is difficult. No politician will want to be seen as funding some unnecessary procedures -- that supposedly is why they want to have health care reform. But if colon cancer mortality rates start to reach British or European levels, those who pass Obamacare will not see power again for a very long time. Knowing that, they are likely instead to keep funding many of those middle-category procedures, just as they do now. Electoral outcomes are part of the cost-benefit analysis when government chooses your medical procedures.

This is what makes Keith Hennessey's analysis all the more real.

Under the current bill, as scored by CBO, spending rises faster than taxes increase each and every year after the ten year window, ad infinitum.The point I would make is that the only way to make those two lines Hennessey has drawn converge, is by taxes. Now it could show up as spending cuts by continuing to decrease reimbursements to doctors and hospitals, but that's just a tax on those providers by a different name. (It makes no difference to you if I pay you $50 for a $100 procedure, or pay you $100 and then charge you a 50% surtax because you're a doctor. It just gets scored differently.)

Shorter version: Government grows, massively. If output grows 3% a year and taxes have to grow 8% a year to keep funding this beast, government's share of GDP gets larger.

Labels: ,