In health care, more is not necessarily better, but less may be even worse.

Posted by Nolan Miller on Oct 1, 2009

Filed Under (Health Care)

Time for another one of the big mysteries of the US health care system.  There are a number of studies that document this fact, but I also want to point you toward a couple of really cool tools. (1)  The Dartmouth Atlas of Health Care is an ongoing research project that documents regional variation in health care spending.  Its web site is  If you go to the site and poke around, you can actually generate a lot of tables and charts that will allow you to compare medical expenditure at various hospitals across the country.  (2)  The government has a web site, that will allow you to compare hospitals on various quality measures.

I went to these resources and pulled some information on four hospitals.  The first two are in Rochester MN, Rochester Methodist and St. Mary’s Hospital.  These two hospitals form part of the Mayo Clinic system.  The second two are in the McAllen, TX area, Edinburgh Regional Medical Center and McAllen Medical Center.  (I couldn’t find McAllen Medical Center in HospitalCompare, but Edinburgh is listed as South Texas Health System, 1102 W Trenton Road).  There are many other measures available on these sites, but I just selected a couple for illustration.


Total Medicare Spending in last 2 years of life

(per patient)

% patients given antibiotics before surgery

% heart attack patients given aspirin at discharge





McAllen Med.




Rochester Meth.




St. Mary’s.





The first column compares spending per Medicare patient during the last two years of their lives.  McAllen and Edinburgh clock in at almost 1/3 more than Rochester and St. Mary’s.  Now, one might argue that Edinburgh and McAllen treat sicker patients, in which case it is no surprise that they spend more on care.  However, the numbers here are spending in the last two years of life.  So, the numbers already look at the sickest patients, and it seems that McAllen Texas spends more on its sickest patients than Rochester Minnesota.  (Looking at raw spending per Medicare beneficiary, McAllen spends about twice the national average.)

Now, poor McAllen has gotten a lot of attention over this fact, almost all of it bad.  It turns out that it spends more on health care (at least for Medicare patients) than just about any other place in the US.  The reason, as far as we can tell, is that McAllen just does more procedures than they do in other parts of the country.  If you don’t believe me, check out this fascinating article by Atul Gawande that ran last summer in the New Yorker that takes a close look at McAllen.

Now, it wouldn’t be such a bad thing if McAllen were spending more on health care but also getting more for it.  But, that doesn’t seem to be true either.  Even if you think McAllen does a great job at providing quality care – the two quality measures in the table above are near 100%, which would seem to be pretty good – other areas do as well for significantly less money.  Take, for example, Rochester MN.  Few would argue that there is a better health system in the US than the Mayo Clinic.  It ranks highly across the board.  On top of delivering high quality, they appear to do so at a cost that is significantly lower than areas like McAllen.

This finding is not limited to the few facts I cherry-picked for the table above.  There have been a large number of academic studies, many of which are associated with the Dartmouth Atlas project (look for author names like Elliott Fisher, Jonathan Skinner, Amitabh Chandra, and Kate Baicker), that document “small area variation” in health care practice/cost that does not fully translate into differences in performance.

Which brings us to the $2.2 Trillion question.  Why are places like McAllen so much more expensive than places like Rochester?  And, if we can figure this out, is there some way to make McAllen more Rochester-like?  We could simply tell expensive locations to spend less.  But, merely forcing them to spend less would not necessarily leave quality unaffected.  Maybe, given the culture/technology/knowledge in these more expensive places, they need to spend 1/3 more than Rochester in order to achieve the same results.  Forcing them to spend less would only make things worse.  Certainly not all low-spending locations provide high quality care.

The success of the Mayo Clinic system, which drives the numbers in Rochester, has been attributed to it being a large, coordinated clinic that pays its doctors by salary, which removes their incentive to order care because they get paid more if they do.  Others have argued that Mayo’s success is an illusion arising from the fact that its patients are richer and healthier than other places.  (See the Washington Post Article “Is the Mayo Clinic a Model or a Mirage?  The Jury is Still Out.”)  Settling this question, and the Dartmouth Atlas’s more general finding that there is a lot of variation in US health expenditure that is unrelated to quality, should be one of our top priorities, because if we could somehow bring the most expensive locations in the US in line with ones like Rochester, that would go a long way toward solving this country’s health care cost crisis.