The following video is of the NBC Nightly News report of the Dartmouth study of the variation of medicare costs by location.

Here is the email that I sent to NBC Nightly News at Nightly@NBC.com.

Brian Williams and Lester Holt,

Your report on the Dartmouth Medicare Health Care Cost Study was very naive. Below is the email I sent to Dartmouth cecs.atlas@dartmouth.edu about the serious flaws in their study. I wish your chief science correspondent understood statistical studies enough to have realized the questions that should have been asked before swallowing this crock hook, line, and sinker.

In the future, Robert Bazell needs to recognize his own limitations so that he can ask for help in formulating crucial questions to ask before reporting so gullibly these kinds of stories.

/Steven Greenberg

Steven Greenberg wrote:

Your health care cost study appears to me to be seriously flawed.

In the press release there is no indication that the variations you measured from location to location were adjusted for the age of the population, the poverty rate, urban versus suburban or rural, or a host of other things that might explain the discrepancy.

For instance in comparing Miami to Ft. Lauderdale, you did not account for the density of population in Miami compared to Ft. Lauderdale. You didn’t indicate the average age of the populations in the two cities. You did not account for the fact that the Cuban American population make-up between the two cities might be vastly different. You did not account for the fact that Miami may be the medical hub and therefore attracts the hard cases. Perhaps Miami has a higher concentration of sophisticated diagnostic tools and thus attracts patients who need them. In other words, if you know you are getting to an age where you will need intensive medical care, you might decide to live in Miami rather than in Ft. Lauderdale.

In your publicity, why talk about the assumed causes of the disparities until you have done some analysis as to which ones might be the actual causes? Your discussion of the possible causes is highly unbalanced as you didn’t mention the obvious possibilities that I mention above.

It is unfair of you to take advantage of the abysmal lack of sophistication in the media for being able to understand the meaning of statistical studies. You should know that reporters and editors are not able to judge any stories that involve numbers. Did any of these reporters even raise any issues like the ones I have raised in this email?

/Steven Greenberg