Asleep at the Scalpel

I had an excuse for the greatest mistake of my medical career. I misdiagnosed a woman who was bleeding internally and she died; but I had been on duty for forty-eight hours with only a couple of hours of snatched sleep. It was not I, but tiredness, that killed her.

This happened in Africa, where mistakes have no consequences -- legal consequences, that is. It is otherwise in America, of course, and an editorial in the New England Journal of Medicine for December 30th draws attention, both clinical and legal, to the consequences of surgery performed by surgeons who have hardly slept the night before, or who suffer from chronic sleep deprivation.

The author quotes a figure that is alarming: serious and potentially fatal complications of elective surgery are 83 percent more frequent when the surgery is performed by a surgeon who has slept less than six hours the night before.

What is wrong with this figure as given? It is that the relative risk is quoted without the absolute risk which, in fact, is far more important to a person making a decision as whether or not to consent to surgery. An 83 percent increase in a negligible risk is likely to result in a still-negligible risk. A small increase in a serious risk will kill or maim far more people than a large increase in a slight risk. The habit of quoting relative risks with quoting absolute ones is therefore to be deplored, for it gives an air of propaganda even to the most respectable causes, and in some cases adds unjustifiably to the fears of the public.

Still, it is common sense that one would prefer one’s surgeon to be fresh as a daisy rather than have a mind befuddled by exhaustion. The author, a specialist in sleep medicine, suggests that hospitals should forbid surgeons from operating who have not slept the night before, and that that they, the surgeons, should inform their patients about the state of their sleep when asking for their consent to an elective (non-emergency) procedure. This sounds like common sense.

As usual in such proposals, however, there is difficulty in knowing where to draw the line. According to research, chronic as well as acute sleep deprivation affects performance, both intellectual and manual. Chronic sleep deprivation is caused by things other than long hours at the hospital: for example, marital unhappiness. Should the surgeon have to reveal the state of his marriage to his patients? "Are you sure you are happily married, doctor?" As further research is done (subjects for theses are always need, after all), it is inevitable that more factors affecting performance will be found, bad traffic on the way to the hospital, for example, or difficulty finding a parking space. The list is potentially endless.

Thus a request for informed consent could end up sounding a bit like the Confessions of St Augustine or, worse still, of Rousseau. And here is one interesting little snippet: when the hours that junior doctors in Britain were allowed to work were restricted by law, the sickness rate among them promptly doubled. Is it better to have a tired doctor, or no doctor at all? Probably a bit of both.