Today’s post starts with a summary of another study showing the importance of physician-patient communication, this time in the key and “massive” area of weight loss. Then we veer into what may be the topic du jour in health policy: the very credibility of research results. I think you’ll find the discussion interesting–and I’ll be interested in your reactions.
One of our readers, Trish, noted a study report in the October issue of the American Journal of Preventive Medicine that ties weight loss in overweight patients to the physician’s style of communication. (Note, BTW, use of the word “preventive,” not the mouth-contorting “preventative.”) Patients in the study tended to lose a significant amount of weight if their doctor used “motivational interviewing” that was tailored to the patient’s circumstances and readiness to change behavior.
But, patients tended not to lose weight if the doctor avoided the topic, or if the doctor spoke only in a general, prescriptive, or confrontational way about the importance of losing weight. It seems we don’t respond well to robotic advice or to nannying. Surprise.
Nicholas Bakalar reported on the study in the New York Times on October 25. Bakalar emphasized that “cajoling and coaxing, scolding and reproach are all ineffective.” What works is “collaborative discussion.” The latter is exactly the type of conversation promoted in this blog and in our book.
The study, by Kathryn I. Pollak of Duke University and colleagues, involved 40 primary care physicians and 461 patients.
But, here’s the question: Can we trust the results of studies like this? New findings show that shockingly large numbers of medical studies reflect the researchers’ biases and self-interests and produce inaccurate results.
For example, as a professor of community and family medicine, Pollak might be altogether willing to find fault with the performance of most primary care physicians. She also might be willing to design a study in the way most likely to show problems with the current system of care, since that’s the kind of study most likely to be published in a major journal or earn the attention of a national reporter. (I don’t know her and am not making an accusation. I’m just using this article to illustrate the potential problem.)
Sound far-fetched? An article entitled “Lies, Damned Lies, and Medical Science” in the November issue of the Atlantic focuses on a Greek medical researcher, John Ioannidis. According to the article, Ioannidis has found that more than two-fifths of 49 widely referenced articles presented results that were contradicted, or disproved, in later studies.
Using a statistical model, Ioannidis further estimates that one-fourth of randomized trials are wrong. This is a particularly damning finding, since randomized trials are considered the gold standard in drug research and comparative effectiveness. And, other statisticians back his claims.
David H. Freedman, writer of the Atlantic piece, suggests that obesity studies are especially prone to error and misleading conclusions. For example, they typically don’t show whether obesity, or a weight-loss intervention, affects mortality. Mortality is not mentioned in the American Journal of Preventive Medicine report.
Commenting on the Atlantic article in e-Patients.net on October 17, the great patient advocate “e-Patient Dave” suggests patients now should be especially wary of physicians who flaunt their knowledge of research results and trivialize patients’ concerns. Journals are recklessly publishing studies that are:
- skewed toward positive results (who’s interested in a study that says an intervention didn’t work?)
- influenced by commercial funders
- methodologically weak
- based solely on patients who are considered “average,” and
- buttressed by big-name professors.
Do you have a sense that your health care has been affected by this problem, or a thought about whether the field of medical research–or health policy research–needs reforming? We’re listening.