Wednesday, June 22, 2011


Peter Orszag worries that new health care technology will promote inequality; that the well-off will be able to afford the newer stuff and the not-so-well-off won’t be. He wisely chooses to focus on the gap in available information—the new brand of health care technologies collect much more information, record much more information, and thereby make much more information available for patients and doctors to act upon. This seems plausible enough on first glance, and it might actually turn out to be true. But I think there are solid reasons for skepticism: I’m not sure such technology would work, and if it does work I see few reasons why it wouldn’t be widespread.

Let’s take the latter critique first. It’s a commonplace that Moore’s Law delivers more computing power at a lower price every year; that’s often mistakenly extended to other, inapplicable fields—not all problems are computing problems, believe it or not—but this is not such an instance. Orszag’s entire case rests on the availability of information, but we know that more information is available at a cheaper price than ever before and I see no reason that this wouldn’t continue.

The more interesting point is one Orszag only hints at but doesn’t get at further, and ties into my idea that I’m not sure such technology would work. I’m using “work” in an unconventional way here—instead of its actual technical use “working,” I mean its social function “working.” The electric car works technically but may not work socially. Same thing for the type of information technology Orszag envisions—I’m not sure it will work socially unless the superstructure of the health care system is changed.

More information isn’t necessarily better. Technocrats like Orszag are very comfortable with integrating data into their lives—with knowing the context and nuance to the thing; they are comfortable with the fact that data is just another language and as such has difficulties—and so they mistakenly extend their mental model to everyone. But we know that most people don’t necessarily work that way. They did a study of calorie counts on restaurant menus in New York City a while back and found that it actually caused customers to order more calories. This, I should note, is for a fairly intuitive statistical measure. There are more obscure and hence more ignorable statistics out there for people to misunderstand.

And if human nature is one thing, how are we to interact with doctors who might be able to explain such subtleties? These days the American doctor is a harried individual packing her day full of appointments and consultations, with nary a moment for explanation. It certainly doesn’t pay at the moment to be so patient about your relationships; medicine is all about volume. And with the amount of information Orszag anticipates, volume is the enemy and context the ally—you can’t boil things down to a specific, headline number.

That’s the social stuff. I’m not even sure that the technical benefits are all that large. Most of the scans they give these days aren’t particularly useful; if you give the same doctor the same scan on two different occasions he will often give you two different diagnoses. This is to say nothing of the various state-of-the-art joint replacements that work no better than their ancestors, or the pharmaceuticals with little clinical benefit but much marketing muscle. Technologically, a lot of the new stuff is not particularly huge.

All this is a way of saying that like many prophecies it sounds good but the reality is far more muddled.

No comments:

Post a Comment