Indirect comparison meta-analysis of aspirin therapy after coronary surgery

A few errors crept into both versions of this paper by Eric Lim and colleagues (BMJ 2003;327:1309-11). In the full version of the article (on bmj.com) the printed equations in the statistical methods section should be on separate lines as follows:

In the abridged version, the table still contains the errors that were corrected at proof stage in the full version (where it is table 2): the numbers of patients with events and the event rates for Lorenz (both regimens) and Sanz (aspirin regimen) and the relative risk (95% confidence interval) for Hockings should therefore read as in the full version (on bmj.com). Finally, an extraneous “2” slipped into Eric Lim's email address; eric.lim{at}cvsnet.orgis his correct address.