In this article by K McPherson and colleagues (9 September, pp 624-8) the fourth sentence of the final paragraph on page 627 should have started: “The NSABP study randomised 13 338 [not 3338] women with a risk equal to that of a 60 year old woman. …”