• Health care is confusing

    The following is based on a long response to some very good questions in the comments to a prior post.

    In the realm of health care, there is much confusion around price vs. volume, levels vs. rates of change, and US vs. other nations. A terrific paper would sort all this out in a rational way. I’m not aware of one. I am also (still and somewhat) among the confused. Below is a little bit about what I (think I) know and (know) I don’t know. I clearly know more than I can put in a short post and I don’t know vastly more than that. This is a start, meant to raise questions, possibly answering none.

    US prices far exceed those of other OECD countries. In many key respects our volume is lower. The product of the two, costs, are higher, even controlling for our higher GDP, as is the rate of change of cost. Is this because of the relative rate of change in prices or volume or both?

    Now, sticking with the US and focusing on Medicare for the moment, that program has implemented systems to control price but done very poorly on volume control. That is, the Medicare’s cost problem is largely a volume one, which is not to say anything about whether that volume is too high or low in some objective sense. And, as we know, volume levels, and rates of change, vary geographically, itself raising a number of important and (to me, at the moment) confusing questions. Our national experts in that area (Chernew, Berenson, Fisher and colleagues, among others) have written many papers in this area. But I am not expert in all aspects of this (some yes, others, I’m learning). I haven’t sorted it all out for myself. I’m not sure anyone has, but I don’t know right now.

    That was Medicare. What about the rest of the market? Chernew and colleagues have published a recent paper on this topic finding positive correlation in commercial market and Medicare volume and negative correlation in prices. Thus, if Medicare suffers a volume problem, so might the rest of the market. If volume is “too low” in Medicare, perhaps the same can be said of the rest of the market. (Keep the distinction between levels and rates of change in mind. Volume can be both too low in one sense and rising too fast in some other sense.)

    The negative price correlation suggests that where it’s “the prices, stupid” in general, it’s less so for Medicare. Meanwhile, there (may be) problems of over-use or inappropriate use for some populations and for some procedures or specialties while, at the same time, there is an under use problem for some. I’ll write more about this in early November, summarizing Cutler’s take.

    I find all of this very confusing. And we haven’t even begun to get into what policy would or should do to all of this. If you’re expecting me to sort all this out right now, you’ve got the wrong guy. I’d love to be that guy but I’d have to do vastly more reading and thinking. If you are aware of any papers or books that do exactly this–explain all of the above differences, what they mean, and what to do about it–I’d love to know about them. Based on what I’ve read it seems various authors focus on one or another element of these issues, ignoring the others. Thus you get stuff that suggests overuse is the main problem, other stuff that says it’s prices, others that say it’s lack of access, and a lot of confusion about analysis of US with respect to itself (over time) and international comparisons, as well as a lot of misunderstanding about levels vs. changes.

    I’m guilty of this too, focusing on one aspect of “the” problem while putting others aside. In general, it’s the only way to get to any depth on anything. One doesn’t get funding to study everything at once, typically. Clearly some synthesis is needed. I’ll keep at it.

    • I certainly dont know of any papers that sort this out. Sorry on that. What I think I see in the literature, is that people start with a bias on the issue, then publish that which confirms that bias. How many times have we read about a system which has worked well for two years, so that proves the author’s assertion. Heck, for two years, you can do almost anything with a system to get the results you want. It takes a while to work out that selection bias.

      Translating solutions into policy is difficult right now because of our politics. I keep hoping we can use that to our advantage and set up competing models in different parts of the country. Does not look like that will happen. In the end, I think we need to put all health care in the same kind of system, whatever that might be, in order to make our system functional and reduce costs. Having a separate system for the elderly, in retrospect, was probably a bad idea.

      I think you are correct in that fixing health care is difficult. My hope is that people like you, Aaron and other data driven people will at least get us going in the right direction. Finally, I am disappointed that so few physicians are participating in this general discussion. I firmly believe that we need input from the provider side to point our problems with proposed policy.