After publication of our recent article  we noticed a typographical error in the Results section. The sentence, 'Celecoxib, at both licensed and any dose always produced more endoscopic ulcers than NSAID' should read, 'Celecoxib, at both licensed and any dose always produced fewer endoscopic ulcers than NSAID'. The relevant section of our article with corrected text follows below:
Endoscopically detected ulcers
Seven trials were designed to detect the presence of endoscopic ulcers of 3 mm or more, in which celecoxib was compared with placebo and/or NSAID (additional file 4). Six reported at 12 weeks, and one at 24 weeks. Five trials also reported results according to the use of low dose aspirin of 325 mg or less daily. These results are shown in Table 8 and Figure 4, analysed across all patients and according to aspirin use. In no comparison was there any significant difference between celecoxib and placebo. For both celecoxib and NSAID there was the same 6% absolute increase in endoscopic ulcers with aspirin use. Celecoxib, at both licensed and any dose always produced fewer endoscopic ulcers than NSAID. The NNTp was the same at 7–8 both with and without concomitant aspirin use.
Moore RS, Derry S, Makinson GT, McQuay HJ: Tolerability and adverse events in clinical trials of celecoxib in osteoarthritis and rheumatoid arthritis: systematic review and meta-analysis of information from company clinical trial reports.