variation from the consequences these variations have on health. In both cases, we have worked hard to use rigorous obser vational methods to distinguish association from causation.

"Dr. Fahey is correct that there are two major competing explanations for the regional differences in intensity of usage: once we rule out differences in illness rates (addressed early in the course of this research), either the patients 'want' more or the providers are prescribing more. Data from two academic medical centers in Maryland reflect the kind of differences we see: Hospital A has a more intensive pattern of end-of-life practice than Hospital B, with more patients dying during an ICU-associated hospital stay (33.9% versus 21.7% of deaths) and fewer patients receiving hospice care (23.9% versus 31.5%). Research by Joanne Lynn and others found that similar differences across five major academic medical centers were not due to patients' preferences but to characteristics of the delivery system—most notably the local supply of hospital beds.

"We do not argue that physicians are venal. On the contrary, we believe that most physicians are trying to do the best they can to care for sick patients in systems that make it more efficient (from the physicians' perspective) to refer or to admit. Doctors' decisions are strongly influenced by how many specialists are on staff at their hospital (making it easier to refer) and how many beds there are relative to the population being served (making it easier to admit patients with complex problems to the hospital, where others can oversee their care).

"The second question—whether a higher-intensity practice pattern is beneficial to health—is just as important. And here the evidence is strong, developed over many years and based upon observational methods (instrumental variables approaches) that have convinced even our most skeptical reviewers. On average, higher-intensity systems have slightly worse technical quality, and higher-intensity care does not result in better health outcomes; if anything,

Be sure to tell us when you move! If your address changes and you want to keep getting Dartmouth Medicine, just tear off the address panel from the back of a recent issue, write your new address next to the old one, and mail it to: Dartmouth Medicine, 1 Medical Center Drive (HB 7070), Lebanon, NH 03756. It helps us greatly— since our mailing list is drawn from six separate databases—if you send the actual cover or a copy of it. If that's not possible, please include both your old and new addresses. Note, too, that if you receive more than one copy of the magazine, it's because of those six databases (which are in different formats, so they can't be automatically "de-duped"). We're happy to eliminate duplications, but it's a help to have the address panel on all the copies you get, not just the one(s) you'd like deleted.

higher-spending systems achieve slightly worse results. And physicians in higher-intensity systems are more likely to say they have difficulty providing high-quality care. This suggests that we have substantial opportunities to reduce costs while maintaining or even improving quality and outcomes.

"This brings us back to preferences. It seems unlikely that patients would demand high-intensity care if they knew they could receive care of higher quality at lower cost—perhaps pocketing the savings. But our delivery system does not offer that choice. We need more robust and convincing performance measures. We need new models of care. And we need to change the payment system.

"But achieving these goals will require change. We believe that academic

medical centers have both an opportunity and a responsibility to be leaders of that change by examining their own practices.

"On that score, readers may be interested to know that the higherintensity 'Hospital A' mentioned above is the University of Maryland Medical Center, while the lower-intensity 'Hospital B' is Johns Hopkins. But both lie at the higher end of U.S. practice patterns."

Foreign correspondent
I was enjoying the Winter 2006 issue of Dartmouth Medicine when I noticed the feature on the magazine's 30th anniversary. That significant milestone impelled me to write.

I have been receiving the magazine for more than 20 years, ever since I was class secretary for the Dartmouth College Class of '63. I have found the magazine consistently interesting and well done. Since retiring a year ago, I have been able to read it more carefully and have appreciated it even more. Before, it always left me wishing I could have spent more time with it.

The current issue is a good example. I read with interest the articles on chemoprevention and flu. (As a former magazine editor, I couldn't help but admire the intrepid way you got around the vexing lack of 1918 file photos.) The collection of past articles in the 30th anniversary feature startled me by bringing me face to face with a number of interesting pieces I had missed. I am trying to catch up now. The editorial remarks about the importance of "story" were very well put. And then there were the timely little insights one wouldn't get from any other source, such as the fact that Hillary Clinton cochaired the birthday dinner for Dr. Koop!

I am now living in Bolivia, where I taught high school history for four years before re-retiring, so I appreciate very much the tie back to Hanover. Best wishes for continued success.

David Boldt, DC '63
Santa Cruz de la Sierra, Bolivia

Boldt is the former editor of the