β Blockade after myocardial infarction: systematic review and meta regression analysis

Several errors crept into this paper by Freemantle and colleagues (26 June 1999, pp 1730-7). In figures 1 and 2 some of the references were incorrectly cited and several bars did not agree with the odds ratios and 95% confidence intervals. Readers may access the correct figures at www.bmj.com/cgi/content/full/318/7200/1730/DC2