Are health care costs really slowing down?

Sarah Kliff writes:

The New England Journal of Medicine published a paper this week titled “When the Cost Curve Bent,” where researchers from the Center for Sustainable Health Costs suggest that the slowdown happened way before the recession. Their analysis shows — and you can see it in this chart — that excess health-care spending growth (any spending above and beyond potential gross domestic product) began to moderate in the early 2000s:

“Too soon to say” is a fair enough response, but this has become increasingly my view over the last year.

Addendum: Angus comments.

Comments

Tyler,
Did you see what is on the Y axis? "Potential GDP?" Please.

Get back to me when they can put actual GDP on the Y axis and say the same thing.

Read the paper rather than just throwing out a comment like this one. They reference the use of this approach here: http://healthpolicyforum.org/post/case-tracking-health-spending-share-potential-gdp You may not like it, that's fine but critique that one rather than the one off that Tyler put up on the website.

How far under PGDP are we at this point since 2003? The link you provided assumes, of course, that such dips below are temporary, balanced out by inflationary booms. Kling's point appears to be valid- until you can put real GDP on that graph, you don't know what was temporary, and what was not. Not only too early to say for sure, but way too early by all evidence present today.

My point is that the authors have provided "their" argument for why this approach should be used and that's where the critique should be. I find their approach to be somewhat persuasive.

You don't buy medical care with imagined output. A curve bend identified this way is utterly meaningless. We have the measured real GDP for the entire period of the curve. All this method seems designed to do is to find a cost bend that may, or may not really exist at all, and you won't know that until you can prove the value used for PGDP (was the CBO correct or not), but then you won't know that until the future has come and gone anyway, in which case you will already know what was spent on medical care out of what real income. Kling's point still stands.

"when they can put actual GDP on the Y axis and say the same thing"

Arnold, won't you ever get it that potential GDP IS the real actual GDP!!!?

(that might be some of my finest work)

It's clear that some areas of expenditures are going down such as spending on prescription drugs. Big therapeutic categories of drugs are now almost all generic and perhaps the lone exception going forward will be in the oncology area (assuming that the genomics approach pans out which is less than certain except for blood related cancers). If the ACA stands and the accountable care provisions work as expected (a big 'if') there should be some savings in primary and hospital associated care. Will this be enough to matter given we still spend almost twice as much per capita on health care as the next leading western economy? I don't know and it's really worth noting the money quote at the end of the article: "If lower growth rates continue, are they low enough? Since much health care is publicly funded, the answer depends on what we're willing to pay in taxes and what we spend on items not related to health care. It must also proceed from a clear notion of what Americans consider acceptable levels of access to and quality of care."

The fact that "big therapeutic categories of drugs" have gone generic would only reduce the growth of health care spending if there was a decline in the development and FDA approval of new therapeutic drugs. Which of course is what's happened.

I find it unconvincing that a moderation of health care costs that's partially the result of declined pharmaceutical and medical innovation and growing regulatory hurdles is actually a good thing. It's like saying low health care spending in poor countries is good, even if it comes at the expense of a lack of modern medicine.

But then, people like Robert Reich have favored health care reform as a regulatory battering ram to reduce medical progress and growth in life expectancy:

http://www.youtube.com/watch?v=IT7Y0TOBuG4

"The fact that “big therapeutic categories of drugs” have gone generic would only reduce the growth of health care spending if there was a decline in the development and FDA approval of new therapeutic drugs. Which of course is what’s happened."

You are drinking the wrong cool aid here. The fact is that there are no more advances to be made in cardiovascular drugs and probably arthritis treatments as well. Companies are leaving these therapeutic areas (and I was in the industry for 28 years) because of the practical R&D problems of finding new targets for drug design that will lead to better therapeutic alternatives than are presently available. Regulatory concerns are secondary.

The fact is that there are no more advances to be made in cardiovascular drugs and probably arthritis treatments as well. Companies are leaving these therapeutic areas (and I was in the industry for 28 years)

You made this exact same comment a few weeks back. People still die from heart disease, and they still suffer from things like arthritis, so there are still advances to be made, even in the pharmacopoeia. It may well be that the present paradigm of drug discovery and development needs to change, but I can tell you that one of the barriers to this needed change is the regulatory environment which is still geared for the discovery and development of single molecular interventions of a target.

Ranjit was completely correct in his observation that your claim might not be good news at all.

There is progress to be made. It is just a diminishing rate of progress toward the goal of "safe and effective" large doses of single small molecules as blockbusters.

I think there might be progress along the margins like the "drug cocktails" for HIV. Our regulation needs to evolve with the changing margins.

Maybe we need a nominal life expectancy target for the FDA.

I would note this blog has produced previous evidence of a more "absolute" and non-adjusted nature, also consistent with the trend of the health care cost curve seeing some real bending. For instance it's been a while since health care costs have been growing at 7-8% a year, and that change came before the cyclical collapse.

It seems as though the slowdown actually started shortly after the Medicare Part D law was passed, though before the Medicare Part D program started, which was in 2006 and increased drug spending.

The law also created HSAs, and encouraged high-deductible plans. Consumer-driven health plans have been discussed before here and elsewhere, but certainly that 2003 Medicare Part D law helped their spread.

So it may be that the real bill that combined increasing health care spending with bending the cost curve was the Medicare Part D law, after all. More study is needed, though.

This is going to be an esoteric point that noone gets, but this is an example of what I see is the problem with "consensus." I'd rather not be arguing about a particular research team's method. The "consensus" is about the big picture and what frame of reference do we make assumptions. It is about what we treat as constant for the future research questions. If the consensus is wrong, that will come out when our future research fails. It's not about convincing people of the finer details (and then using the so-called consensus to beat the out-group about the head and shoulders). The "consensus" certainly shouldn't be used as a bottleneck for grants and publications. It is used by individual researchers at their desks and benches to determine where are the highest returns for their research time. Should we go forward assuming healthcare spending is moderating? Likely, yes. But that is just the beginning of the process. What hypotheses does that assumption of truth suggest? How do we test them?

The consideration of 'potential' GDP actually appears helpful in that 1) healthcare does not readily respond to business cycles and 2) healthcare spending has to adjust at the end of spending periods to other financial realities. All of which, of course, make healthcare spending difficult to measure in real time.

What is this, an infographic for ants?

Well, if not, I guess we can always go to the third derivative...

As I have said before: Makes sense, how much more can they cut us.

Angus is right. Every industry has a process of disaggregating waste from good operations and potential investments and killing off investments that didn't work. As a continual process, doing it as crisis management probably isn't the right way to go about it.

Comments for this post are closed