Gawande scoured the literature, actually showing a mild protective effect for full moons: fewer overdoses. More recent studies have shown no full-moon correlation in ER admissions for cardiopulmonary arrest or maxillofacial emergencies, at least in Europe (Eur J Emerg Med. 2003 Sep;10(3):225-8 and Br J Oral Maxillofac Surg. 2003 Jun;41(3):170-2, respectively). But that 'protective' effect may be real, though: new moon cardiac arrest is slightly higher, as shown by a couple of other studies. Finally, an old (1978) study posits that psychiatric emergencies in Dade County do in fact cluster around full moons (J Clin Psychiatry. 1978 May;39(5):385-92).
As for Friday the 13th, there are some interesting stats. In Britain and in Finland, there were measurably more traffic accidents over several years. The British study went so far as to observe fewer cars on the road (compared to Friday the 6th), but more auto-related ED admissions. Thus, a large attributable risk (52%). The Finnish study, published in an American journal, found that all the extra traffic deaths on the 13th were coming from women, and blame their anxiety on the accidents. I'm not making this up. The references are: BMJ. 1993 Dec 18-25;307(6919):1584-6 and Am J Psychiatry. 2002 Dec;159(12):2110-1. Or just enter "Friday the 13th" at www.pubmed.gov.
An amusing survey, but Gawande misses a point: The surgery residents trying to avoid the full moon shift weren't quoting the above data. Nor were they observing pagan superstitions. What they were doing, I think, was participating in some groupthink, or bowing to peer pressure.
How many call nights have I commented "it's been quiet so far" -- only to be mildly reprimanded by a resident that my remarks will jinx the team? It happens often. It happens in the midst of discussions on physiology, or evidence-based guidelines.
This is, I think, a consequence of the OR mentality. Surgeons don't maintain a sterile field in the OR by reminding themselves of the risk factors for bacterial contamination. They do it by committing a regimented series of behaviors to memory. There are some aspects to the scrubbing and gowning that are clearly not evidence-based, but based on tradition and repetition. Deviating from this would be unsettling. So, too, is jinxing the team or signing up for a full moon shift.
Easterbrook invokes something similar in the Tuesday Morning Quarterback columns. In the midst of his play analyses, in which he often invokes probability and arcane NFL statistics, he'll also talk about offending ye Football Gods with poor etiquette or selfish behavior.
It seems the more complex the tasks, the thought processes, the more likely some kind of rigid codes of conduct appear. Actually, no, these codes are everywhere, regardless of the complexity of the job. It's just that, in the world of medicine and surgery, I'm surprised they're still there at all.
Full disclosure: I've yet to work in the ED on a Friday the 13th, but there was that memorable call last Halloween, when I put in my first chest tube. We were very busy. And Gawande got slammed on his shift, by the way. Tons of admissions.