poly: Is Medicine Healthy?

From: Robin Hanson <hanson@econ.berkeley.edu>
Date: Fri Apr 02 1999 - 16:54:58 PST

Amazingly enough, researchers have yet to measure a significant aggregate
effect of medicine (doctors, etc.) on health.

Sure there are lots of clinical studies purporting to show the benefits of
various treatments. But clinical trials tend to study best practice on
patients mostly likely to benefit. Negative results tend not to be
published, and the vast majority of medical practice has yet to be studied
with clinical trials.

The studies that have looked at the aggregate health effect of medicine as
typically practiced, averaging over all the things doctors do, have no
found a significant effect. Two decades ago ~$50 million was spent on a
randomized trial among 5000 people over 2-3 years, some of which got 1/3
more medical care than the rest. The only differences found were corrected
vision via glasses, more filled teeth cavities, and lowered blood pressure
among the poor (which may be a placebo effect). A recent analysis of 5
million Medicare patients, using regional spending variations of a factor
of two (controlling for lots of stuff), found that any mortality benefit of
spending in the last six months of life is less than a one part in a

Is this all from anal statisticians imposing too high a standard of
evidence? Consider that by these same standards we *have* seen significant
robust effects of age, exercise, fat, smoking, and social status on health.
(Exercise and social status reduce mortality by roughly a factor of three

How could medicine not be healthy? Surgery clearly has negative effects on
the body, drugs upset evolved balances, and hospitals are clearly great
places to spread contagious diseases. It is widely accepted that centuries
ago doctors did many harmful things like bleeding folks along with a few
useful things they did, and that the net effect may not have been positive.
The same may be true today.

Of course most of this data is about the "second half" of the money we
spend on medical care. Maybe we get more health value for the first half
of spending. Or maybe not.

The apparent low health value of medical care is all the more striking
given that we spend 14% of GDP on it in the U.S. And it must surely give
pause to those who hope that medicine will soon give us dramatically
expanded life spans.

Supporting cites:

Antibiotics little mortality effect AER 84(3)369-95 Jun94
Randomized health insure experiment AER 77(3):251-77 Jun87
Regional variations in spend at end http://nberws.nber.org/papers/W6513
Regional variations in spend NEJM 328(9):621-8 4Mar93
We can see other effects JAMA 279(21):1703-8 3Jun98
"Why are some people healthy and others not?" Evans et al 94

Contrary cites I'm not impressed by (I could explain why):

AER 88(2)132-6.
http://nberws.nber.org/papers/W5750 (now QJE 113(4):991-4 Nov98)
JAMA 277(7):535-42 19Feb97 (see critique JAMA 227(24):1931-4 25Jun97)

AER = American Economic Review, NEJM = New England Journal of Med.
JAMA = Journal of American Medical Association

Robin Hanson
hanson@econ.berkeley.edu http://hanson.berkeley.edu/
RWJF Health Policy Scholar FAX: 510-643-8614
140 Warren Hall, UC Berkeley, CA 94720-7360 510-643-1884
after 8/99: Assist. Prof. Economics, George Mason Univ.
Received on Fri Apr 2 16:59:40 1999

This archive was generated by hypermail 2.1.8 : Tue Mar 07 2006 - 14:45:30 PST