More on that Dartmouth health care study

Hank Aaron at the Brookings Institution, who knows a lot more about policy than I do, had some interesting comments on the recent New York Times article about problems with the Dartmouth health care atlas. which I discussed a few hours ago. Aaron writes that much of the criticism in that newspaper article was off-base, but that there are real difficulties in translating the Dartmouth results (finding little relation between spending and quality of care) to cost savings in the real world.

Aaron writes:

The Dartmouth research, showing huge variation in the use of various
medical procedures and large variations in per patient spending under
Medicare, has been a revelation and a useful one. There is no way to
explain such variation on medical grounds and it is problematic. But
readers, including my former colleague Orszag, have taken an
oversimplistic view of what the numbers mean and what to do about
them. There are three really big problems with the common
interpretations, none of which the Times authors explained clearly–
or, indeed, at all.

1. Spending variations on Medicare are inversely related to spending
variations by private payers. That means that what appear to be high
spending areas are, in many cases, high cost-shifting areas. Two
sights in Wisconsin, in cluding one that includes the justifiabley
praised Marshfield Clinic, have among the lowest per person Medicare
spending in the nation, but were found to have among the highest
private per person spending in the nation. What counts is total
medical spending if one wants to relate spending and quality.

2. Knowing that spending is high and poorly correlated with measures
of quality (if such a correlation exists!) is not the same as knowing
how to pare spending in a way that increases net social benefit from
health care. Repeated studies have shown that medical care spending
ON THE AVERAGE produces benefits far greater than its cost, but that
AT THE MARGIN there is a lot of waste. The waste is observed, for the
most part, ex post. To be sure that cutting spending improves well-
being one has to be able to predict ex ante who will derive benefits
that are less than cost. There are some egregious situations where ex
ante waste can be demonstrated. But for the most part, we just do not
have good enough research to predict with sufficient accuracy who will
benefit and who will not from doing more of some procedure. That will
take a lot of research. It hasn’t yet been done.

3. Some research has shown that even the most glaring cost
differences noted in the Dartmouth research are associated with
improved outcomes for those conditions where something close to random
assignment of patients to alternative patient regimes occurs. The guy
who is doing this work is Joseph Doyle at MIT. If the Times article
had described some of his work, readers would have learned something.
For example, you can be pretty sure that people who suffer coronaries
while on vacation do not choose in which county to have the heart
attack. Doyle studied such patients. He found that they did better
in Florida counties that spent most, the very counties that the
Dartmouth folks have held up for their poor average outcomes. He also
found that preemies who were a couple of grams under the threshold at
which they are described as in need of intensive care did materially
better than did preemies a couple of grams over that threshold and,
hence, treated as normal births. The weight difference was too small
to matter medically and went in the wrong direction; the intensity of
care mattered.

Doyle’s reasearch and the finding on the inverse relationship between
Medicare spending and private spending, together with the fact that we
can’t predict worth crap who will and who will not benefits from many
procedures means not that the Dartmouth people are wrong, but that any
savings will be very hard to achieve without doing more harm than good
and will be very slow in coming. The Administration, and Orszag in
particular, were wrong in neglecting the warning to forecasters of
Scottish economist, Alec Cairncross, ‘Give a number. Or a date.
Never both.’

2 thoughts on “More on that Dartmouth health care study

  1. I can't believe anyone would say this:

    Spending variations on Medicare are inversely related to spending
    variations by private payers. That means that what appear to be high
    spending areas are, in many cases, high cost-shifting areas … What counts is total
    medical spending if one wants to relate spending and quality.

    Does anyone really believe that? Do you really think that spending on extra knee replacements for Medicare patients will improve the quality of pre-natal care? That's absurd. If anything, the idea that high Medicare spending is due to cost shifting makes the situation bleaker and easier to deal with. It means that there is excess health services being provided everywhere and the Dartmouth group has just identified the places where the government is burdened by it. We should be able to lower total costs everywhere without a detrimental impact to health care.

Comments are closed.