Thursday, June 3, 2010

More on the Dartmouth Atlas, Courtesy of the New York Times

The Disease Management Care Blog has written lots on the good, the bad and the ugly of the Dartmouth Atlas (here, here and here). Now we have an excellent article in the New York Times by Reed Abelson and Gardiner Harris that reviews the strengths and weaknesses of their data. As the DMCB has suspected, it appears that the heady excitement from being quoted by Presidents and Senators may have lead the researchers to sometimes portray their findings as more than they really are. Kudos to these reporters for actually reading the research for themselves.

6 comments:

pheski said...

While the NYTimes article is generally well written, it is not without flaws:

http://www.stat.columbia.edu/~cook/movabletype/archives/2010/06/how_can_news_re.html

son2 said...

Kudos to these reporters for actually reading the research for themselves.

How can you say that? These reporters' seem to have come to the conclusion that the Dartmouth Atlas researchers main conclusion is that "that the nation’s best hospitals tend to be among the least expensive."

This is totally wrong. This is not the main assertion of the Dartmouth researchers, and it is not a widely held view (as the reporters claim).

Jaan Sidorov said...

Pheski has a link to an excellent and neutral examination of the Times article and the arguments that underlie it. For that, the DMCB says thanks! That being said, the Dartmouth Atlas data is also not free of any criticisms. Does DA warrant not being questioned at all? Hardly.

Son2 points out that there may be a disconnect between what people THINK the DA is saying and what it really says. That is also true, but one question is: to what degree have the DA researchers themselves facilitated the perspective that cheap = best?

Phil 314 said...

I wouldn't give the authors kudos. I've always seen the Dartmouth Atlas as forcing us to ask ourselves difficult questions ("Why didn't more money lead to better outcomes?")

And most important it has forced organized medicine to address the whole issue of variation in care. Why does zip code X with similar demographics to zip code Y have twice the number of cardiac cath's?

I believe we've finally gotten past the well that's the art of medicine response. Unfortunately we haven't given up on the my patients are sicker excuse even when asked to "prove it".

Its the data wars.

However, when all is said and done the data wars stop with:

50% more per capita than the next highest country with no better outcomes and 1 in 7 still not covered

son2 said...

...to what degree have the DA researchers themselves facilitated the perspective that cheap = best?

Okay, you're right. But I also think that the ones who stand to gain from Abelson and Harris's distorted picture of the Dartmouth work ("cheap = better") are those using scaremongering tactics of healthcare rationing to block health reform.

Wennberg and Fisher don't stand to gain from that idea, and so that is why they tend not to frame their work that way, and instead to express it as, "more != better."

That is, after all, the strength of their work, right? That it provides a counter-argument to the claim that health care is good, and so more health care is better?

For disclosure, I was a Dartmouth undergrad, and so I can't promise that I don't feel some irrational bias for something called the "Dartmouth Atlas," even if I'm not personally invested in the work that they do... :)

Jaan Sidorov said...

c3 points out that the Dartmouth Atlas numbers have a lot of 'face validity,' i.e., it just makes too much sense to discard their troubling results.

As I recall, the central tenent of the DA is that the discrepancy in the cost of preference sensitive care cannot be explained by any observable variant. That's OK. However, they and others have speculated that the difference correlates with the availability of the speciality care AND fee for service payment schedules. In the meantime, other academics have pointed out that there may be other explanations, such as the percent of Medicare or the surrounding poverty. I personally think that it's generally healthy to continuously seek the truth. Until the New York Times pointed out that there may be chinks in the DA armour, many - including the President - seemed to treat the findings as orthodoxy.

son2 wonders if there are more nefarious incentives at play. I don't see that as likely - I mean this is the New York Times. Then there is the highly academic Dr. Cooper, the gadfly of the DA who also has little economic incentive to purposefully lead folks astray. And.... I'm a Dartmouth alum also - it's where I did my internship and residency.