Problems of stopping trials early

This Analysis article by Guyatt and colleagues contains a typographical error (BMJ 2012;344:e3863, doi:10.1136/bmj.e3863) in the second paragraph of the section “β blockers in non-cardiac surgery.” The odds ratio documented for the “twofold increase in non-fatal strokes” has a 95% confidence interval of 1.27 to 3.68 (not “27 to 3.68” as was published).