Interpreting medical research

In one of my recent graduate school classes we spent a lot of time (A LOT of time) discussing sources of bias and confounders in medical research. And then we dissected a few selected articles — articles from really solid journals (and a few from the equivalent of Noelle's Journal of Oncology — I kinda though there might be problems from that starting point!). I was surprised with the weaknesses we were able to turn up with studies that at first read seemed extremely well done.

In any case, for about two months after that I spent a tremendous amount of time overanalyzing all the journal articles I read — and there were potential weaknesses in all of them. I thought of this moment in my life when I came across a recent PLoS Medicine editorial, entitled "Why most published research findings are false." Despite a rather alarming title, this is a great article outlining basic principles of study design and interpretation, and may be an eye opener for many clinicians, who tend not to have had a lot of emphasis on study interpretation in medical school. So, check it out.

Also interesting this week: From The New England Journal of Medicine via The New York Times, an article pointing out what I've known since, oh, 2000, when my med school debt amounted to the price of a nice house in the Chicago suburbs. Your debt influences the likelihood of wanting to go into a medical subspecialty. I've long wondered why reducing the cost of medical school doesn't seem to get a lot of press time in the discussion of reducing the costs of medical care. Less debt means more primary care doctors means more "medical homes," which may translate to lower costs, or at least, higher quality. Of course, also this week, an academic ER physician went on a total tirade against primary care (good luck to him getting a primary care partnership), so maybe we don't all agree about the importance of internists.