SuperFreakonomics and the hospital epidemiologist

In their new book, SuperFreakonomics, Steven Levitt and Stephen Dubner give us their analysis of why hand hygiene compliance is poor among doctors. They cite the following reasons: the large number of patients that may be seen in a day and how busy doctors are; inaccessibility of sinks, though they note that conveniently placed alcohol-product dispensers are often ignored; perception deficit, that is, doctors believe their compliance is much better than it actually is; and arrogance. Then they put on their economist hats and talk about negative externalities. By this they mean that the doctor bears little risk personally when he or she is noncompliant with hand hygiene and thus the doctor has little incentive to comply. So far, so good. But then they downplay trying to change behavior in favor of solutions that bypass the need for change in behavior (e.g., antimicrobial impregnated products). I think the answer is both behavior change and non-behavioral solutions are required since there aren't enough of the latter to overcome dirty hands. I was particularly happy that they advised doctors to stop wearing neckties to improve infection control (though my happiness is not just infection control related). Also of note, they profile our colleague, Dr. Rekha Murthy, hospital epidemiologist at Cedars-Sinai Medical Center, and her successful program on hand hygiene.

So what's the importance of this? The economists haven't really added any new insights or solutions to the problem of healthcare associated infections. But they clearly will have impact by offering their analysis to the general public. Their previous book, Freakonomics, has sold over 4 million copies, and led to the launching of a blog with a full time editor, and a movie is in the works. How did I learn about their interest in infection control? I saw Stephen Dubner on CNN talking about why doctors shouldn't wear neckties. Interestingly, if you do a Google image search of him, you'll note that he rarely wears a tie.