Several errors crept unseen into this editorial by Paul Little (BMJ 2007;334:909, 5 May, doi: 10.1136/bmj.39184.617049.80) because of a loophole in a minor part of our submission and editorial process; we are improving this bit of the process to ensure such errors can't infiltrate again. Firstly, in the opening paragraph, 8/34 is 24% [not 21%], and the 21% cited as an adjusted relative risk is in fact a value for absolute risk reduction. Secondly, in the version published in the printed journal (and in the pdf), the third sentence of the final paragraph should read: “Until such evidence is available, I would advise patients who have had four episodes of sore throat in one year or three in six months that they are likely to have on average two episodes (12 days) of sore throat in the next six months and two or three days of fever if they decide not to have the operation; if they decide to have the operation they are likely to have about 13 days of severe pain immediately after surgery, and then on average half an episode (3 days) of sore throat in the next six months and half a day of fever.”